<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD with MathML3 v1.2 20190208//EN" "JATS-archivearticle1-mathml3.dtd"> <article xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" article-type="research-article" dtd-version="1.2"><front><journal-meta><journal-id journal-id-type="nlm-ta">elife</journal-id><journal-id journal-id-type="publisher-id">eLife</journal-id><journal-title-group><journal-title>eLife</journal-title></journal-title-group><issn pub-type="epub" publication-format="electronic">2050-084X</issn><publisher><publisher-name>eLife Sciences Publications, Ltd</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="publisher-id">72904</article-id><article-id pub-id-type="doi">10.7554/eLife.72904</article-id><article-categories><subj-group subj-group-type="display-channel"><subject>Short Report</subject></subj-group><subj-group subj-group-type="heading"><subject>Neuroscience</subject></subj-group></article-categories><title-group><article-title>Charting brain growth and aging at high spatial precision</article-title></title-group><contrib-group><contrib contrib-type="author" corresp="yes" id="author-249127"><name><surname>Rutherford</surname><given-names>Saige</given-names></name><contrib-id authenticated="true" contrib-id-type="orcid">https://orcid.org/0000-0003-3006-9044</contrib-id><email>saige.rutherford@donders.ru.nl</email><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="aff" rid="aff3">3</xref><xref ref-type="fn" rid="con1"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-249128"><name><surname>Fraza</surname><given-names>Charlotte</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="fn" rid="con2"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249129"><name><surname>Dinga</surname><given-names>Richard</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="fn" rid="con3"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249130"><name><surname>Kia</surname><given-names>Seyed Mostafa</given-names></name><contrib-id authenticated="true" contrib-id-type="orcid">https://orcid.org/0000-0002-7128-814X</contrib-id><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="aff" rid="aff4">4</xref><xref ref-type="fn" rid="con4"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249131"><name><surname>Wolfers</surname><given-names>Thomas</given-names></name><xref ref-type="aff" rid="aff5">5</xref><xref ref-type="aff" rid="aff6">6</xref><xref ref-type="other" rid="fund9"/><xref ref-type="fn" rid="con5"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249132"><name><surname>Zabihi</surname><given-names>Mariam</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="fn" rid="con6"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249133"><name><surname>Berthet</surname><given-names>Pierre</given-names></name><contrib-id authenticated="true" contrib-id-type="orcid">https://orcid.org/0000-0002-6878-6842</contrib-id><xref ref-type="aff" rid="aff5">5</xref><xref ref-type="aff" rid="aff6">6</xref><xref ref-type="fn" rid="con7"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249134"><name><surname>Worker</surname><given-names>Amanda</given-names></name><xref ref-type="aff" rid="aff7">7</xref><xref ref-type="fn" rid="con8"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249135"><name><surname>Verdi</surname><given-names>Serena</given-names></name><xref ref-type="aff" rid="aff8">8</xref><xref ref-type="aff" rid="aff9">9</xref><xref ref-type="fn" rid="con9"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249136"><name><surname>Andrews</surname><given-names>Derek</given-names></name><xref ref-type="aff" rid="aff10">10</xref><xref ref-type="fn" rid="con10"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249137"><name><surname>Han</surname><given-names>Laura KM</given-names></name><xref ref-type="aff" rid="aff11">11</xref><xref ref-type="aff" rid="aff12">12</xref><xref ref-type="fn" rid="con11"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-249138"><name><surname>Bayer</surname><given-names>Johanna MM</given-names></name><contrib-id authenticated="true" contrib-id-type="orcid">https://orcid.org/0000-0003-4891-6256</contrib-id><xref ref-type="aff" rid="aff13">13</xref><xref ref-type="aff" rid="aff14">14</xref><xref ref-type="fn" rid="con12"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249139"><name><surname>Dazzan</surname><given-names>Paola</given-names></name><xref ref-type="aff" rid="aff7">7</xref><xref ref-type="aff" rid="aff15">15</xref><xref ref-type="fn" rid="con13"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-249140"><name><surname>McGuire</surname><given-names>Phillip</given-names></name><xref ref-type="aff" rid="aff16">16</xref><xref ref-type="fn" rid="con14"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249141"><name><surname>Mocking</surname><given-names>Roel T</given-names></name><xref ref-type="aff" rid="aff17">17</xref><xref ref-type="fn" rid="con15"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249142"><name><surname>Schene</surname><given-names>Aart</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff18">18</xref><xref ref-type="fn" rid="con16"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-140083"><name><surname>Sripada</surname><given-names>Chandra</given-names></name><contrib-id authenticated="true" contrib-id-type="orcid">https://orcid.org/0000-0001-9025-6453</contrib-id><xref ref-type="aff" rid="aff3">3</xref><xref ref-type="other" rid="fund13"/><xref ref-type="fn" rid="con17"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249143"><name><surname>Tso</surname><given-names>Ivy F</given-names></name><xref ref-type="aff" rid="aff3">3</xref><xref ref-type="other" rid="fund11"/><xref ref-type="fn" rid="con18"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249144"><name><surname>Duval</surname><given-names>Elizabeth R</given-names></name><xref ref-type="aff" rid="aff3">3</xref><xref ref-type="other" rid="fund14"/><xref ref-type="fn" rid="con19"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249145"><name><surname>Chang</surname><given-names>Soo-Eun</given-names></name><xref ref-type="aff" rid="aff3">3</xref><xref ref-type="other" rid="fund12"/><xref ref-type="fn" rid="con20"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249146"><name><surname>Penninx</surname><given-names>Brenda WJH</given-names></name><xref ref-type="aff" rid="aff11">11</xref><xref ref-type="aff" rid="aff12">12</xref><xref ref-type="fn" rid="con21"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-249147"><name><surname>Heitzeg</surname><given-names>Mary M</given-names></name><xref ref-type="aff" rid="aff3">3</xref><xref ref-type="fn" rid="con22"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-64832"><name><surname>Burt</surname><given-names>S Alexandra</given-names></name><xref ref-type="aff" rid="aff19">19</xref><xref ref-type="other" rid="fund15"/><xref ref-type="other" rid="fund16"/><xref ref-type="fn" rid="con23"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-128219"><name><surname>Hyde</surname><given-names>Luke W</given-names></name><xref ref-type="aff" rid="aff20">20</xref><xref ref-type="other" rid="fund15"/><xref ref-type="other" rid="fund16"/><xref ref-type="fn" rid="con24"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-47303"><name><surname>Amaral</surname><given-names>David</given-names></name><xref ref-type="aff" rid="aff10">10</xref><xref ref-type="other" rid="fund6"/><xref ref-type="other" rid="fund7"/><xref ref-type="other" rid="fund8"/><xref ref-type="fn" rid="con25"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-166975"><name><surname>Wu Nordahl</surname><given-names>Christine</given-names></name><xref ref-type="aff" rid="aff10">10</xref><xref ref-type="other" rid="fund6"/><xref ref-type="other" rid="fund7"/><xref ref-type="other" rid="fund8"/><xref ref-type="fn" rid="con26"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249148"><name><surname>Andreasssen</surname><given-names>Ole A</given-names></name><xref ref-type="aff" rid="aff6">6</xref><xref ref-type="aff" rid="aff21">21</xref><xref ref-type="fn" rid="con27"/><xref ref-type="fn" rid="conf3"/></contrib><contrib contrib-type="author" id="author-55257"><name><surname>Westlye</surname><given-names>Lars T</given-names></name><xref ref-type="aff" rid="aff5">5</xref><xref ref-type="aff" rid="aff6">6</xref><xref ref-type="aff" rid="aff21">21</xref><xref ref-type="other" rid="fund2"/><xref ref-type="fn" rid="con28"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-70730"><name><surname>Zahn</surname><given-names>Roland</given-names></name><xref ref-type="aff" rid="aff22">22</xref><xref ref-type="other" rid="fund10"/><xref ref-type="fn" rid="con29"/><xref ref-type="fn" rid="conf2"/></contrib><contrib contrib-type="author" id="author-249149"><name><surname>Ruhe</surname><given-names>Henricus G</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff18">18</xref><xref ref-type="fn" rid="con30"/><xref ref-type="fn" rid="conf4"/></contrib><contrib contrib-type="author" equal-contrib="yes" id="author-249150"><name><surname>Beckmann</surname><given-names>Christian</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="aff" rid="aff23">23</xref><xref ref-type="fn" rid="equal-contrib1">†</xref><xref ref-type="other" rid="fund4"/><xref ref-type="fn" rid="con31"/><xref ref-type="fn" rid="conf5"/></contrib><contrib contrib-type="author" equal-contrib="yes" id="author-248983"><name><surname>Marquand</surname><given-names>Andre F</given-names></name><contrib-id authenticated="true" contrib-id-type="orcid">https://orcid.org/0000-0001-5903-203X</contrib-id><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="fn" rid="equal-contrib1">†</xref><xref ref-type="other" rid="fund1"/><xref ref-type="other" rid="fund3"/><xref ref-type="other" rid="fund5"/><xref ref-type="fn" rid="con32"/><xref ref-type="fn" rid="conf1"/></contrib><aff id="aff1"><label>1</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/016xsfp80</institution-id><institution>Donders Institute for Brain, Cognition, and Behavior, Radboud University</institution></institution-wrap><addr-line><named-content content-type="city">Nijmegen</named-content></addr-line><country>Netherlands</country></aff><aff id="aff2"><label>2</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/05wg1m734</institution-id><institution>Department of Cognitive Neuroscience, Radboud University Medical Center</institution></institution-wrap><addr-line><named-content content-type="city">Nijmegen</named-content></addr-line><country>Netherlands</country></aff><aff id="aff3"><label>3</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/00jmfr291</institution-id><institution>Department of Psychiatry, University of Michigan</institution></institution-wrap><addr-line><named-content content-type="city">Ann Arbor</named-content></addr-line><country>United States</country></aff><aff id="aff4"><label>4</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/04pp8hn57</institution-id><institution>Department of Psychiatry, Utrecht University Medical Center</institution></institution-wrap><addr-line><named-content content-type="city">Utrecht</named-content></addr-line><country>Netherlands</country></aff><aff id="aff5"><label>5</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/01xtthb56</institution-id><institution>Department of Psychology, University of Oslo</institution></institution-wrap><addr-line><named-content content-type="city">Oslo</named-content></addr-line><country>Norway</country></aff><aff id="aff6"><label>6</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/00j9c2840</institution-id><institution>Norwegian Center for Mental Disorders Research (NORMENT), University of Oslo, and Oslo University Hospital</institution></institution-wrap><addr-line><named-content content-type="city">Oslo</named-content></addr-line><country>Norway</country></aff><aff id="aff7"><label>7</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/0220mzb33</institution-id><institution>Department of Psychological Medicine, Institute of Psychiatry, Psychology and Neuroscience, King’s College London</institution></institution-wrap><addr-line><named-content content-type="city">London</named-content></addr-line><country>United Kingdom</country></aff><aff id="aff8"><label>8</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/02jx3x895</institution-id><institution>Centre for Medical Image Computing, Medical Physics and Biomedical Engineering, University College London</institution></institution-wrap><addr-line><named-content content-type="city">London</named-content></addr-line><country>United Kingdom</country></aff><aff id="aff9"><label>9</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/02jx3x895</institution-id><institution>Dementia Research Centre, UCL Queen Square Institute of Neurology</institution></institution-wrap><addr-line><named-content content-type="city">London</named-content></addr-line><country>United Kingdom</country></aff><aff id="aff10"><label>10</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/05rrcem69</institution-id><institution>The Medical Investigation of Neurodevelopmental Disorders (MIND) Institute and Department of Psychiatry and Behavioral Sciences, UC Davis School of Medicine, University of California, Davis</institution></institution-wrap><addr-line><named-content content-type="city">Sacramento</named-content></addr-line><country>United States</country></aff><aff id="aff11"><label>11</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/05grdyy37</institution-id><institution>Amsterdam UMC, Vrije Universiteit, Psychiatry, Amsterdam Public Health Research Institute</institution></institution-wrap><addr-line><named-content content-type="city">Amsterdam</named-content></addr-line><country>Netherlands</country></aff><aff id="aff12"><label>12</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/01x2d9f70</institution-id><institution>GGZ inGeest, Amsterdam Neuroscience</institution></institution-wrap><addr-line><named-content content-type="city">Amsterdam</named-content></addr-line><country>Netherlands</country></aff><aff id="aff13"><label>13</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/01ej9dk98</institution-id><institution>Centre for Youth Mental Health, University of Melbourne</institution></institution-wrap><addr-line><named-content content-type="city">Melbourne</named-content></addr-line><country>Australia</country></aff><aff id="aff14"><label>14</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/00ej70p06</institution-id><institution>Orygen Youth Health</institution></institution-wrap><addr-line><named-content content-type="city">Melbourne</named-content></addr-line><country>Australia</country></aff><aff id="aff15"><label>15</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/0220mzb33</institution-id><institution>National Institute for Health Research Mental Health Biomedical Research Centre, South London and Maudsley National Health Service Foundation Trust and King’s College London</institution></institution-wrap><addr-line><named-content content-type="city">London</named-content></addr-line><country>United Kingdom</country></aff><aff id="aff16"><label>16</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/0220mzb33</institution-id><institution>Department of Psychosis Studies, Institute of Psychiatry, King’s College London</institution></institution-wrap><addr-line><named-content content-type="city">London</named-content></addr-line><country>United Kingdom</country></aff><aff id="aff17"><label>17</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/05grdyy37</institution-id><institution>Department of Psychiatry, Amsterdam UMC, Location AMC</institution></institution-wrap><addr-line><named-content content-type="city">Amsterdam</named-content></addr-line><country>Netherlands</country></aff><aff id="aff18"><label>18</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/05wg1m734</institution-id><institution>Department of Psychiatry, Radboud University Medical Center</institution></institution-wrap><addr-line><named-content content-type="city">Nijmegen</named-content></addr-line><country>Netherlands</country></aff><aff id="aff19"><label>19</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/05hs6h993</institution-id><institution>Department of Psychology, Michigan State University</institution></institution-wrap><addr-line><named-content content-type="city">East Lansing</named-content></addr-line><country>United States</country></aff><aff id="aff20"><label>20</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/00jmfr291</institution-id><institution>Department of Psychology, University of Michigan</institution></institution-wrap><addr-line><named-content content-type="city">Ann Arbor</named-content></addr-line><country>United States</country></aff><aff id="aff21"><label>21</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/01xtthb56</institution-id><institution>KG Jebsen Centre for Neurodevelopmental Disorders Research, Institute of Clinical Medicine, University of Oslo</institution></institution-wrap><addr-line><named-content content-type="city">Oslo</named-content></addr-line><country>Norway</country></aff><aff id="aff22"><label>22</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/0220mzb33</institution-id><institution>Centre for Affective Disorders at the Institute of Psychiatry, King's College London</institution></institution-wrap><addr-line><named-content content-type="city">London</named-content></addr-line><country>United Kingdom</country></aff><aff id="aff23"><label>23</label><institution-wrap><institution-id institution-id-type="ror">https://ror.org/052gg0110</institution-id><institution>Centre for Functional MRI of the Brain (FMRIB), Nuffield Department of Clinical Neurosciences, Wellcome Centre for Integrative Neuroimaging, University of Oxford</institution></institution-wrap><addr-line><named-content content-type="city">Oxford</named-content></addr-line><country>United Kingdom</country></aff></contrib-group><contrib-group content-type="section"><contrib contrib-type="editor"><name><surname>Baker</surname><given-names>Chris I</given-names></name><role>Reviewing Editor</role><aff><institution>National Institute of Mental Health, National Institutes of Health</institution><country>United States</country></aff></contrib><contrib contrib-type="senior_editor"><name><surname>Baker</surname><given-names>Chris I</given-names></name><role>Senior Editor</role><aff><institution>National Institute of Mental Health, National Institutes of Health</institution><country>United States</country></aff></contrib></contrib-group><author-notes><fn fn-type="con" id="equal-contrib1"><label>†</label><p>These authors contributed equally to this work</p></fn></author-notes><pub-date date-type="publication" publication-format="electronic"><day>01</day><month>02</month><year>2022</year></pub-date><pub-date pub-type="collection"><year>2022</year></pub-date><volume>11</volume><elocation-id>e72904</elocation-id><history><date date-type="received" iso-8601-date="2021-08-19"><day>19</day><month>08</month><year>2021</year></date><date date-type="accepted" iso-8601-date="2022-01-26"><day>26</day><month>01</month><year>2022</year></date></history><pub-history><event><event-desc>This manuscript was published as a preprint at .</event-desc><date date-type="preprint" iso-8601-date="2021-08-08"><day>08</day><month>08</month><year>2021</year></date><self-uri content-type="preprint" xlink:href="https://doi.org/10.1101/2021.08.08.455487"/></event></pub-history><permissions><copyright-statement>© 2022, Rutherford et al</copyright-statement><copyright-year>2022</copyright-year><copyright-holder>Rutherford et al</copyright-holder><ali:free_to_read/><license xlink:href="http://creativecommons.org/licenses/by/4.0/"><ali:license_ref>http://creativecommons.org/licenses/by/4.0/</ali:license_ref><license-p>This article is distributed under the terms of the <ext-link ext-link-type="uri" xlink:href="http://creativecommons.org/licenses/by/4.0/">Creative Commons Attribution License</ext-link>, which permits unrestricted use and redistribution provided that the original author and source are credited.</license-p></license></permissions><self-uri content-type="pdf" xlink:href="elife-72904-v2.pdf"/><self-uri content-type="figures-pdf" xlink:href="elife-72904-figures-v2.pdf"/><abstract><p>Defining reference models for population variation, and the ability to study individual deviations is essential for understanding inter-individual variability and its relation to the onset and progression of medical conditions. In this work, we assembled a reference cohort of neuroimaging data from 82 sites (N=58,836; ages 2–100) and used normative modeling to characterize lifespan trajectories of cortical thickness and subcortical volume. Models are validated against a manually quality checked subset (N=24,354) and we provide an interface for transferring to new data sources. We showcase the clinical value by applying the models to a transdiagnostic psychiatric sample (N=1985), showing they can be used to quantify variability underlying multiple disorders whilst also refining case-control inferences. These models will be augmented with additional samples and imaging modalities as they become available. This provides a common reference platform to bind results from different studies and ultimately paves the way for personalized clinical decision-making.</p></abstract><kwd-group kwd-group-type="author-keywords"><kwd>normative model</kwd><kwd>lifespan</kwd><kwd>growth chart</kwd><kwd>brain chart</kwd><kwd>big data</kwd><kwd>individual prediction</kwd></kwd-group><kwd-group kwd-group-type="research-organism"><title>Research organism</title><kwd>Human</kwd></kwd-group><funding-group><award-group id="fund1"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100010663</institution-id><institution>H2020 European Research Council</institution></institution-wrap></funding-source><award-id>10100118</award-id><principal-award-recipient><name><surname>Marquand</surname><given-names>Andre F</given-names></name></principal-award-recipient></award-group><award-group id="fund2"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100010663</institution-id><institution>H2020 European Research Council</institution></institution-wrap></funding-source><award-id>802998</award-id><principal-award-recipient><name><surname>Westlye</surname><given-names>Lars T</given-names></name></principal-award-recipient></award-group><award-group id="fund3"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100004440</institution-id><institution>Wellcome Trust</institution></institution-wrap></funding-source><award-id>215698/Z/19/Z</award-id><principal-award-recipient><name><surname>Marquand</surname><given-names>Andre F</given-names></name></principal-award-recipient></award-group><award-group id="fund4"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100004440</institution-id><institution>Wellcome Trust</institution></institution-wrap></funding-source><award-id>098369/Z/12/Z</award-id><principal-award-recipient><name><surname>Beckmann</surname><given-names>Christian</given-names></name></principal-award-recipient></award-group><award-group id="fund5"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/501100003246</institution-id><institution>Nederlandse Organisatie voor Wetenschappelijk Onderzoek</institution></institution-wrap></funding-source><award-id>VIDI grant 016.156.415</award-id><principal-award-recipient><name><surname>Marquand</surname><given-names>Andre F</given-names></name></principal-award-recipient></award-group><award-group id="fund6"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000025</institution-id><institution>National Institute of Mental Health</institution></institution-wrap></funding-source><award-id>R01MH104438</award-id><principal-award-recipient><name><surname>Amaral</surname><given-names>David</given-names></name><name><surname>Wu Nordahl</surname><given-names>Christine</given-names></name></principal-award-recipient></award-group><award-group id="fund7"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000025</institution-id><institution>National Institute of Mental Health</institution></institution-wrap></funding-source><award-id>R01MH103371</award-id><principal-award-recipient><name><surname>Amaral</surname><given-names>David</given-names></name><name><surname>Wu Nordahl</surname><given-names>Christine</given-names></name></principal-award-recipient></award-group><award-group id="fund8"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100009633</institution-id><institution>Eunice Kennedy Shriver National Institute of Child Health and Human Development</institution></institution-wrap></funding-source><award-id>P50 HD093079</award-id><principal-award-recipient><name><surname>Amaral</surname><given-names>David</given-names></name><name><surname>Wu Nordahl</surname><given-names>Christine</given-names></name></principal-award-recipient></award-group><award-group id="fund9"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100010665</institution-id><institution>H2020 Marie Skłodowska-Curie Actions</institution></institution-wrap></funding-source><award-id>895011</award-id><principal-award-recipient><name><surname>Wolfers</surname><given-names>Thomas</given-names></name></principal-award-recipient></award-group><award-group id="fund10"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/501100000265</institution-id><institution>Medical Research Council</institution></institution-wrap></funding-source><award-id>G0902304</award-id><principal-award-recipient><name><surname>Zahn</surname><given-names>Roland</given-names></name></principal-award-recipient></award-group><award-group id="fund11"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000025</institution-id><institution>National Institute of Mental Health</institution></institution-wrap></funding-source><award-id>K23MH108823</award-id><principal-award-recipient><name><surname>Tso</surname><given-names>Ivy F</given-names></name></principal-award-recipient></award-group><award-group id="fund12"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000055</institution-id><institution>National Institute on Deafness and Other Communication Disorders</institution></institution-wrap></funding-source><award-id>R01DC011277</award-id><principal-award-recipient><name><surname>Chang</surname><given-names>Soo-Eun</given-names></name></principal-award-recipient></award-group><award-group id="fund13"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000025</institution-id><institution>National Institute of Mental Health</institution></institution-wrap></funding-source><award-id>R01MH107741</award-id><principal-award-recipient><name><surname>Sripada</surname><given-names>Chandra</given-names></name></principal-award-recipient></award-group><award-group id="fund14"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100008269</institution-id><institution>Michigan Institute for Clinical and Health Research</institution></institution-wrap></funding-source><award-id>UL1TR002240</award-id><principal-award-recipient><name><surname>Duval</surname><given-names>Elizabeth R</given-names></name></principal-award-recipient></award-group><award-group id="fund15"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100000025</institution-id><institution>National Institute of Mental Health</institution></institution-wrap></funding-source><award-id>UG3MH114249</award-id><principal-award-recipient><name><surname>Burt</surname><given-names>S Alexandra</given-names></name><name><surname>Hyde</surname><given-names>Luke</given-names></name></principal-award-recipient></award-group><award-group id="fund16"><funding-source><institution-wrap><institution-id institution-id-type="FundRef">http://dx.doi.org/10.13039/100009633</institution-id><institution>Eunice Kennedy Shriver National Institute of Child Health and Human Development</institution></institution-wrap></funding-source><award-id>R01HD093334</award-id><principal-award-recipient><name><surname>Burt</surname><given-names>S Alexandra</given-names></name><name><surname>Hyde</surname><given-names>Luke</given-names></name></principal-award-recipient></award-group><funding-statement>The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.</funding-statement></funding-group><custom-meta-group><custom-meta specific-use="meta-only"><meta-name>Author impact statement</meta-name><meta-value>During an individual's lifetime, their brain undergoes a unique transformation that we have precisely mapped across thousands of people to create and share reference growth charts of the human brain's lifespan which allow for single subject exploration and inference.</meta-value></custom-meta></custom-meta-group></article-meta></front><body><sec id="s1" sec-type="intro"><title>Introduction</title><p>Since their introduction more than a century ago, normative growth charts have become fundamental tools in pediatric medicine and also in many other areas of anthropometry (<xref ref-type="bibr" rid="bib5">Cole, 2012</xref>). They provide the ability to quantify individual variation against centiles of variation in a reference population, which shifts focus away from group-level (e.g., case-control) inferences to the level of the individual. This idea has been adopted and generalized in clinical neuroimaging, and normative modeling is now established as an effective technique for providing inferences at the level of the individual in neuroimaging studies (<xref ref-type="bibr" rid="bib27">Marquand et al., 2016</xref>; <xref ref-type="bibr" rid="bib28">Marquand et al., 2019</xref>).</p><p>Although normative modeling can be used to estimate many different kinds of mappings—for example between behavioral scores and neurobiological readouts—normative models of brain development and aging are appealing considering that many brain disorders are grounded in atypical trajectories of brain development (<xref ref-type="bibr" rid="bib19">Insel, 2014</xref>) and the association between cognitive decline and brain tissue in aging and neurodegenerative diseases (<xref ref-type="bibr" rid="bib20">Jack et al., 2010</xref>; <xref ref-type="bibr" rid="bib22">Karas et al., 2004</xref>). Indeed, normative modeling has been applied in many different clinical contexts, including charting the development of infants born pre-term (<xref ref-type="bibr" rid="bib8">Dimitrova et al., 2020</xref>) and dissecting the biological heterogeneity across cohorts of individuals with different brain disorders, including schizophrenia, bipolar disorder, autism, and attention-deficit/hyperactivity disorder (<xref ref-type="bibr" rid="bib3">Bethlehem et al., 2020</xref>; <xref ref-type="bibr" rid="bib46">Wolfers et al., 2021</xref>; <xref ref-type="bibr" rid="bib47">Zabihi et al., 2019</xref>).</p><p>A hurdle to the widespread application of normative modeling is a lack of well-defined reference models to quantify variability across the lifespan and to compare results from different studies. Such models should: (1) accurately model population variation across large samples; (2) be derived from widely accessible measures; (3) provide the ability to be updated as additional data come online, (4) be supported by easy-to-use software tools, and (5) should quantify brain development and aging at a high spatial resolution, so that different patterns of atypicality can be used to stratify cohorts and predict clinical outcomes with maximum spatial precision. Prior work on building normative modeling reference cohorts (<xref ref-type="bibr" rid="bib4">Bethlehem et al., 2021</xref>) has achieved some of these aims (1–4), but has modeled only global features (i.e., total brain volume), which is useful for quantifying brain growth but has limited utility for the purpose of stratifying clinical cohorts (aim 5). The purpose of this paper is to introduce a set of reference models that satisfy all these criteria.</p><p>To this end, we assemble a large neuroimaging data set (<xref ref-type="table" rid="table1">Table 1</xref>) from 58,836 individuals across 82 scan sites covering the human lifespan (aged 2–100, <xref ref-type="fig" rid="fig1">Figure 1A</xref>) and fit normative models for cortical thickness and subcortical volumes derived from Freesurfer (version 6.0). We show the clinical utility of these models in a large transdiagnostic psychiatric sample (N=1985, <xref ref-type="fig" rid="fig2">Figure 2</xref>). To maximize the utility of this contribution, we distribute model coefficients freely along with a <ext-link ext-link-type="uri" xlink:href="https://github.com/predictive-clinical-neuroscience/braincharts">set of software tools</ext-link> to enable researchers to derive subject-level predictions for new data sets against a set of common reference models.</p><table-wrap id="table1" position="float"><label>Table 1.</label><caption><title>Sample description and demographics.</title><p>mQC refers to the manual quality checked subset of the full sample. ‘All’ rows=Train+Test. Clinical refers to the transdiagnostic psychiatric sample (diagnostic details in <xref ref-type="fig" rid="fig2">Figure 2A</xref>).</p></caption><table frame="hsides" rules="groups"><thead><tr><th align="left" valign="bottom"/><th align="left" valign="bottom"/><th align="left" valign="bottom">N (subjects)</th><th align="left" valign="bottom">N (sites)</th><th align="left" valign="bottom">Sex (%F/%M)</th><th align="left" valign="bottom">Age (Mean, S.D)</th></tr></thead><tbody><tr><td align="left" valign="bottom">Full</td><td align="left" valign="bottom">All</td><td align="char" char="." valign="bottom">58,836</td><td align="char" char="." valign="bottom">82</td><td align="left" valign="bottom"/><td align="left" valign="bottom"/></tr><tr><td align="left" valign="bottom"> </td><td align="left" valign="bottom">Training set</td><td align="char" char="." valign="bottom">29,418</td><td align="char" char="." valign="bottom">82</td><td align="char" char="." valign="bottom">51.1/48.9</td><td align="char" char="." valign="bottom">46.9, 24.4</td></tr><tr><td align="left" valign="bottom"> </td><td align="left" valign="bottom"> Test set</td><td align="char" char="." valign="bottom">29,418</td><td align="char" char="." valign="bottom">82</td><td align="char" char="." valign="bottom">50.9/49.1</td><td align="char" char="." valign="bottom">46.9, 24.4</td></tr><tr><td align="left" valign="bottom">mQC</td><td align="left" valign="bottom">All</td><td align="char" char="." valign="bottom">24,354</td><td align="char" char="." valign="bottom">59</td><td align="left" valign="bottom"/><td align="left" valign="bottom"/></tr><tr><td align="left" valign="bottom"> </td><td align="left" valign="bottom"> Training set</td><td align="char" char="." valign="bottom">12,177</td><td align="char" char="." valign="bottom">59</td><td align="char" char="." valign="bottom">50.2/49.8</td><td align="char" char="." valign="bottom">30.2, 24.1</td></tr><tr><td align="left" valign="bottom"> </td><td align="left" valign="bottom">Test set</td><td align="char" char="." valign="bottom">12,177</td><td align="char" char="." valign="bottom">59</td><td align="char" char="." valign="bottom">50.4/49.4</td><td align="char" char="." valign="bottom">30.1, 24.2</td></tr><tr><td align="left" valign="bottom">Clinical</td><td align="left" valign="bottom">Test set</td><td align="char" char="." valign="bottom">1985</td><td align="char" char="." valign="bottom">24</td><td align="char" char="." valign="bottom">38.9/61.1</td><td align="char" char="." valign="bottom">30.5, 14.1</td></tr><tr><td align="left" valign="bottom">Transfer</td><td align="left" valign="bottom">Test set</td><td align="char" char="." valign="bottom">546</td><td align="char" char="." valign="bottom">6</td><td align="char" char="." valign="bottom">44.5/55.5</td><td align="char" char="." valign="bottom">24.8, 13.7</td></tr></tbody></table></table-wrap><fig id="fig1" position="float"><label>Figure 1.</label><caption><title>Normative model overview.</title><p>(<bold>A</bold>) Age density distribution (x-axis) of each site (y-axis) in the full model train and test, clinical, and transfer validation set. (<bold>B</bold>) Age count distribution of the full sample (N=58,836). (<bold>C, D</bold>) Examples of lifespan trajectories of brain regions. Age is shown on x-axis and predicted thickness (or volume) values are on the y-axis. Centiles of variation are plotted for each region. In (<bold>C</bold>), we show that sex differences between females (red) and males (blue) are most pronounced when modeling large-scale features such as mean cortical thickness across the entire cortex or total gray matter volume. These sex differences manifest as a shift in the mean in that the shape of these trajectories is the same for both sexes, as determined by sensitivity analyses where separate normative models were estimated for each sex. The explained variance (in the full test set) of the whole cortex and subcortex is highlighted inside the circle of (<bold>D</bold>). All plots within the circle share the same color scale. Visualizations for all ROI trajectories modeled are shared on GitHub for users that wish to explore regions not shown in this figure.</p></caption><graphic mime-subtype="jpeg" mimetype="image" xlink:href="elife-72904.xml.media/fig1.jpg"/></fig><fig id="fig2" position="float"><label>Figure 2.</label><caption><title>Normative modeling in clinical cohorts.</title><p>Reference brain charts were transferred to several clinical samples (described in (<bold>A</bold>)). Patterns of extreme deviations were summarized for each clinical group and compared to matched control groups (from the same sites). (<bold>B</bold>) Shows extreme positive deviations (thicker/larger than expected) and (<bold>C</bold>) shows the extreme negative deviation (thinner/smaller than expected) patterns. (<bold>D</bold>) Shows the significant (FDR corrected p<0.05) results of classical case-control methods (mass-univariate t-tests) on the true cortical thickness data (top row) and on the deviations scores (bottom row). There is unique information added by each approach which becomes evident when noticing the maps in (<bold>B–D</bold>) are not identical. ADHD, attention-deficit hyperactive disorder; ASD, autism spectrum disorder; BD, bipolar disorder; EP, early psychosis; FDR, false discovery rate; MDD, major depressive disorder; SZ, schizophrenia.</p></caption><graphic mime-subtype="jpeg" mimetype="image" xlink:href="elife-72904.xml.media/fig2.jpg"/></fig></sec><sec id="s2" sec-type="results"><title>Results</title><p>We split the available data into training and test sets, stratifying by site (<xref ref-type="table" rid="table1">Table 1</xref>, <xref ref-type="supplementary-material" rid="supp1 supp2">Supplementary files 1 and 2</xref>). After careful quality checking procedures, we fit normative models using a set of covariates (age, sex, and fixed effects for site) to predict cortical thickness and subcortical volume for each parcel in a high-resolution atlas (<xref ref-type="bibr" rid="bib7">Destrieux et al., 2010</xref>). We employed a warped Bayesian linear regression model to accurately model non-linear and non-Gaussian effects (<xref ref-type="bibr" rid="bib14">Fraza et al., 2021</xref>), whilst accounting for scanner effects (<xref ref-type="bibr" rid="bib2">Bayer et al., 2021</xref>; <xref ref-type="bibr" rid="bib23">Kia et al., 2021</xref>). These models are summarized in <xref ref-type="fig" rid="fig1">Figure 1</xref> and <xref ref-type="fig" rid="fig3">Figure 3</xref>, <xref ref-type="fig" rid="fig3s1">Figure 3—figure supplements 1</xref>–<xref ref-type="fig" rid="fig3s3">3</xref>, and with an online interactive visualization tool for exploring the evaluation metrics across different test sets (overview of this tool shown in <xref ref-type="video" rid="video1">Video 1</xref>). The raw data used in these visualizations are available on <ext-link ext-link-type="uri" xlink:href="https://github.com/predictive-clinical-neuroscience/braincharts/tree/master/metrics">GitHub</ext-link> (<xref ref-type="bibr" rid="bib36">Rutherford, 2022a</xref>).</p><fig-group><fig id="fig3" position="float"><label>Figure 3.</label><caption><title>Evaluation metrics across all test sets.</title><p>The distribution of evaluation metrics in four different test sets (full, mQC, patients, and transfer, see Materials and methods) separated into left and right hemispheres and subcortical regions, with the skew and excess kurtosis being measures that depict the accuracy of the estimated shape of the model, ideally both would be around zero. Note that kurtosis is highly sensitive to outlying samples. Overall, these models show that the models fit well in term of central tendency and variance (explained variance and MSLL) and model the shape of the distribution well in most regions (skew and kurtosis). Code and sample data for transferring these models to new sites not included in training is shared.</p></caption><graphic mime-subtype="jpeg" mimetype="image" xlink:href="elife-72904.xml.media/fig3.jpg"/></fig><fig id="fig3s1" position="float" specific-use="child-fig"><label>Figure 3—figure supplement 1.</label><caption><title>Comparison of the explained variance in cortical thickness across the different test sets.</title><p>The patterns appear to be robust and consistent across the different test sets.</p></caption><graphic mime-subtype="jpeg" mimetype="image" xlink:href="elife-72904.xml.media/fig3-figsupp1.jpg"/></fig><fig id="fig3s2" position="float" specific-use="child-fig"><label>Figure 3—figure supplement 2.</label><caption><title>Showing the explained variance for each brain region across 10 randomized resampling of the full control test set.</title></caption><graphic mime-subtype="jpeg" mimetype="image" xlink:href="elife-72904.xml.media/fig3-figsupp2.jpg"/></fig><fig id="fig3s3" position="float" specific-use="child-fig"><label>Figure 3—figure supplement 3.</label><caption><title>Per site explained variance across the different test sets.</title></caption><graphic mime-subtype="jpeg" mimetype="image" xlink:href="elife-72904.xml.media/fig3-figsupp3.jpg"/></fig></fig-group><media id="video1" mime-subtype="mp4" mimetype="video" xlink:href="elife-72904-video1.mp4"><label>Video 1.</label><caption><title>"Demonstation of the functionality of our interactive online visualization tool (<ext-link ext-link-type="uri" xlink:href="https://brainviz-app.herokuapp.com/">https://brainviz-app.herokuapp.com/</ext-link>) that is available for all evaluation metrics across all test sets.</title><p>The code for creating this website can be found on GitHub (<ext-link ext-link-type="uri" xlink:href="https://github.com/saigerutherford/brainviz-app">https://github.com/saigerutherford/brainviz-app</ext-link>; copy archived at <ext-link ext-link-type="uri" xlink:href="https://archive.softwareheritage.org/swh:1:dir:5172634bcf0ed341052462df7780d000c5bd4f9e;origin=https://github.com/saigerutherford/brainviz-app;visit=swh:1:snp:70e2db995b8f63e9bdcb233217089d22000d2147;anchor=swh:1:rev:021fff9a48b26f2d07bbb4b3fb92cd5202418905">swh:1:rev:021fff9a48b26f2d07bbb4b3fb92cd5202418905</ext-link>; <xref ref-type="bibr" rid="bib37">Rutherford, 2022b</xref>).</p></caption></media><p>We validate our models with several careful procedures: first, we report out of sample metrics; second, we perform a supplementary analysis on a subset of participants for whom input data had undergone manual quality checking by an expert rater (<xref ref-type="table" rid="table1">Table 1</xref> – mQC). Third, each model fit was evaluated using metrics (<xref ref-type="fig" rid="fig3">Figure 3</xref>, <xref ref-type="fig" rid="fig3s1">Figure 3—figure supplements 1</xref>–<xref ref-type="fig" rid="fig3s3">3</xref>) that quantify central tendency and distributional accuracy (<xref ref-type="bibr" rid="bib9">Dinga et al., 2021</xref>; <xref ref-type="bibr" rid="bib14">Fraza et al., 2021</xref>). We also estimated separate models for males and females, which indicate that sex effects are adequately modeled using a global offset. Finally, to facilitate independent validation, we packaged pretrained models and code for transferring to new samples into an <ext-link ext-link-type="uri" xlink:href="https://github.com/predictive-clinical-neuroscience/braincharts">open resource</ext-link> for use by the community and demonstrated <ext-link ext-link-type="uri" xlink:href="https://colab.research.google.com/github/predictive-clinical-neuroscience/braincharts/blob/master/scripts/apply_normative_models.ipynb">how to transfer</ext-link> the models to new samples (i.e., data not present in the initial training set).</p><p>Our models provide the opportunity for mapping the diverse trajectories of different brain areas. Several examples are shown in <xref ref-type="fig" rid="fig1">Figure 1C and D</xref> which align with known patterns of development and aging (<xref ref-type="bibr" rid="bib10">Ducharme et al., 2016</xref>; <xref ref-type="bibr" rid="bib16">Gogtay et al., 2004</xref>; <xref ref-type="bibr" rid="bib40">Tamnes et al., 2010</xref>). Moreover, across the cortex and subcortex our model fits well, explaining up to 80% of the variance out of sample (<xref ref-type="fig" rid="fig3">Figure 3</xref>, <xref ref-type="fig" rid="fig3s1">Figure 3—figure supplements 1</xref>–<xref ref-type="fig" rid="fig3s3">3</xref>).</p><p>A goal of this work is to develop normative models that can be applied to many different clinical conditions. To showcase this, we apply the model to a transdiagnostic psychiatric cohort (<xref ref-type="table" rid="table1">Table 1</xref> – Clinical; <xref ref-type="fig" rid="fig2">Figure 2</xref>) resulting in personalized, whole-brain deviation maps that can be used to understand inter-individual variability (e.g., for stratification) and to quantify group separation (e.g., case-control effects). To demonstrate this, for each clinical group, we summarized the individual deviations within that group by computing the proportion of subjects that have deviations in each region and comparing to matched (same sites) controls in the test set (<xref ref-type="fig" rid="fig2">Figure 2B–C</xref>). Additionally, we performed case-control comparisons on the raw cortical thickness and subcortical volumes, and on the deviation maps (<xref ref-type="fig" rid="fig2">Figure 2D</xref>), again against a matched sample from the test set. This demonstrates the advantages of using normative models for investigating individual differences in psychiatry, that is, quantifying clinically relevant information at the level of each individual. For most diagnostic groups, the z-statistics derived from the normative deviations also provided stronger case-control effects than the raw data. This shows the importance of accurate modeling of population variance across multiple clinically relevant dimensions. The individual-level deviations provide complimentary information to the group effects, which aligns with previous work (<xref ref-type="bibr" rid="bib44">Wolfers et al., 2018</xref>; <xref ref-type="bibr" rid="bib45">Wolfers et al., 2020</xref>; <xref ref-type="bibr" rid="bib48">Zabihi et al., 2020</xref>). We note that a detailed description of the clinical significance of our findings is beyond the scope of this work and will be presented separately.</p></sec><sec id="s3" sec-type="discussion"><title>Discussion</title><p>In this work, we create lifespan brain charts of cortical thickness and subcortical volume derived from structural MRI, to serve as reference models. Multiple data sets were joined to build a mega-site lifespan reference cohort to provide good coverage of the lifespan. We applied the reference cohort models to clinical data sets and demonstrated the benefits of normative modeling in addition to standard case-control comparisons. All models, including <ext-link ext-link-type="uri" xlink:href="https://pcntoolkit.readthedocs.io/en/latest/pages/tutorial_braincharts_apply_nm.html">documentation</ext-link> and <ext-link ext-link-type="uri" xlink:href="https://github.com/predictive-clinical-neuroscience/braincharts">code</ext-link>, are made available to the research community. We also provide an example data set (that includes data from sites not in the training sample) along with the code to demonstrate how well our models can adapt to new sites, and how easy it is to transfer our pretrained models to users’ own data sets.</p><p>We identify three main strengths of our approach. First, our large lifespan data set provides high anatomical specificity, necessary for discriminating between conditions, predicting outcomes, and stratifying subtypes. Second, our models are flexible in that they can model non-Gaussian distributions, can easily be transferred to new sites, and are built on validated analytical techniques and software tools (<xref ref-type="bibr" rid="bib14">Fraza et al., 2021</xref>; <xref ref-type="bibr" rid="bib23">Kia et al., 2021</xref>; <xref ref-type="bibr" rid="bib28">Marquand et al., 2019</xref>). Third, we show the general utility of this work in that it provides the ability to map individual variation whilst also improving case-control inferences across multiple disorders.</p><p>In recent work, a large consortium established lifespan brain charts that are complementary to our approach (<xref ref-type="bibr" rid="bib4">Bethlehem et al., 2021</xref>). Benefits of their work include precisely quantifying brain growth using a large cohort, but they only provide estimates of four coarse global measures (e.g., total brain volume). While this can precisely quantify brain growth and aging this does not provide the ability to generate individualized fingerprints or to stratify clinical cohorts. In contrast, in this work, we focus on providing spatially specific estimates (188 different brain regions) across the post-natal lifespan which provides fine-grained anatomical estimates of deviation, offering an individualized perspective that can be used for clinical stratification. We demonstrate the transdiagnostic clinical value of our models (<xref ref-type="fig" rid="fig2">Figure 2</xref>) by showing how clinical variation is widespread in a fine-grain manner (e.g., not all individuals deviate in the same regions and not all disorders have the same characteristic patterns) and we facilitate clinical applications of our models by sharing tutorial code notebooks with sample data that can be run locally or online in a web browser.</p><p>We also identify the limitations of this work. We view the word ‘normative’ as problematic. This language implies that there are normal and abnormal brains, a potentially problematic assumption. As indicated in <xref ref-type="fig" rid="fig2">Figure 2</xref>, there is considerable individual variability and heterogeneity among trajectories. We encourage the use of the phrase ‘reference cohort’ over ‘normative model’. In order to provide coverage of the lifespan the curated data set is based on aggregating existing data, meaning there is unavoidable sampling bias. Race, education, and socioeconomic variables were not fully available for all included data sets, however, given that data were compiled from research studies, they are likely samples drawn predominantly from Western, Educated, Industrialized, Rich, and Democratic (WEIRD) societies (<xref ref-type="bibr" rid="bib18">Henrich et al., 2010</xref>) and future work should account for these factors. The sampling bias of UKBiobank (<xref ref-type="bibr" rid="bib15">Fry et al., 2017</xref>) is especially important for users to consider as UKBiobank data contributes 59% of the full sample. By sampling both healthy population samples and case-control studies, we achieve a reasonable estimate of variation across individuals, however, downstream analyses should consider the nature of the reference cohort and whether it is appropriate for the target sample. Second, we have relied on semi-automated quality control (QC) for the full sample which—despite a conservative choice of inclusion threshold—does not guarantee either that low-quality data were excluded or that the data were excluded are definitively excluded because of artifacts. We addressed this by comparing our full test set to a manually quality check data set and observed similar model performance. Also, Freesurfer was not adjusted for the very young age ranges (2–7 yo) thus caution should be used when interpreting the model on new data in this age range. Finally, although the models presented in this study are comprehensive, they are only the first step, and we will augment our <ext-link ext-link-type="uri" xlink:href="https://github.com/predictive-clinical-neuroscience/braincharts/models">repository</ext-link> with more diverse data, different features, and modeling advances as these become available.</p></sec><sec id="s4" sec-type="materials|methods"><title>Materials and methods</title><p>Data from 82 sites were combined to create the initial full sample. These sites are described in detail in <xref ref-type="supplementary-material" rid="supp1 supp2">Supplementary files 1-2</xref>, including the sample size, age (mean and standard deviation), and sex distribution of each site. Many sites were pulled from publicly available data sets including <ext-link ext-link-type="uri" xlink:href="https://abcdstudy.org/">ABCD</ext-link>, <ext-link ext-link-type="uri" xlink:href="http://fcon_1000.projects.nitrc.org/indi/abide/">ABIDE</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://fcon_1000.projects.nitrc.org/indi/adhd200/">ADHD200</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://www.cam-can.org/index.php?content=dataset">CAMCAN</ext-link>, <ext-link ext-link-type="uri" xlink:href="http://fcon_1000.projects.nitrc.org/indi/cmi_healthy_brain_network/">CMI-HBN</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://www.humanconnectome.org/study/hcp-lifespan-aging">HCP-Aging</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://www.humanconnectome.org/study/hcp-lifespan-development">HCP-Development</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://www.humanconnectome.org/study/human-connectome-project-for-early-psychosis">HCP-Early Psychosis</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://www.humanconnectome.org/study/hcp-young-adult">HCP-Young Adult</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://brain-development.org/ixi-dataset/">IXI</ext-link>, <ext-link ext-link-type="uri" xlink:href="http://fcon_1000.projects.nitrc.org/indi/enhanced/">NKI-RS</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://www.oasis-brains.org/">Oasis</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://openneuro.org/">OpenNeuro</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://www.med.upenn.edu/bbl/philadelphianeurodevelopmentalcohort.html">PNC</ext-link>, <ext-link ext-link-type="uri" xlink:href="https://bicr-resource.atr.jp/srpbsopen/">SRPBS</ext-link>, and <ext-link ext-link-type="uri" xlink:href="https://www.ukbiobank.ac.uk/">UKBiobank</ext-link>. For data sets that include repeated visits (i.e., ABCD and UKBiobank), only the first visit was included. Other included data come from studies conducted at the University of Michigan (<xref ref-type="bibr" rid="bib11">Duval et al., 2018</xref>; <xref ref-type="bibr" rid="bib35">Rutherford et al., 2020</xref>; <xref ref-type="bibr" rid="bib41">Tomlinson et al., 2020</xref>; <xref ref-type="bibr" rid="bib42">Tso et al., 2021</xref>; <xref ref-type="bibr" rid="bib43">Weigard et al., 2021</xref>; <xref ref-type="bibr" rid="bib49">Zucker et al., 2009</xref>), University of California Davis (<xref ref-type="bibr" rid="bib32">Nordahl et al., 2020</xref>), University of Oslo (<xref ref-type="bibr" rid="bib31">Nesvåg et al., 2017</xref>), King’s College London (<xref ref-type="bibr" rid="bib17">Green et al., 2012</xref>; <xref ref-type="bibr" rid="bib25">Lythe et al., 2015</xref>), and Amsterdam University Medical Center (<xref ref-type="bibr" rid="bib29">Mocking et al., 2016</xref>). Full details regarding sample characteristics, diagnostic procedures, and acquisition protocols can be found in the publications associated with each of the studies. Equal sized training and testing data sets (split half) were created using scikit-learn’s train_test_split function, stratifying on the site variable. It is important to stratify based on site, not only study (<xref ref-type="bibr" rid="bib4">Bethlehem et al., 2021</xref>), as many of the public studies (i.e., ABCD) include several sites, thus modeling study does not adequately address MRI scanner confounds. To test stability of the model performance, the full test set was randomly resampled 10 times and evaluation metrics were re-calculated on each split of the full test set (<xref ref-type="fig" rid="fig3s2">Figure 3—figure supplement 2</xref>). To show generalizability of the models to new data not included in training, we leveraged data from <ext-link ext-link-type="uri" xlink:href="https://openneuro.org/">OpenNeuro.org</ext-link> (<xref ref-type="bibr" rid="bib26">Markiewicz et al., 2021</xref>) to create a transfer data set (six sites, N=546, <xref ref-type="supplementary-material" rid="supp3">Supplementary file 3</xref>). This data are provided along with the code for transferring to walk users through how to apply these models to their own data.</p><p>The clinical validation sample consisted of a subset of the full data set (described in detail in <xref ref-type="fig" rid="fig1">Figure 1A</xref>, <xref ref-type="fig" rid="fig2">Figure 2A</xref> and <xref ref-type="supplementary-material" rid="supp1">Supplementary file 1</xref>). Studies (sites) contributing clinical data included: Autism Brain Imaging Database Exchange (ABIDE GU, KKI, NYU, USM), ADHD200 (KKI, NYU), CNP, SRPBS (CIN, COI, KTT, KUT, HKH, HRC, HUH, SWA, UTO), Delta (AmsterdamUMC), Human Connectome Project Early Psychosis (HCP-EP BWH, IU, McL, MGH), KCL, University of Michigan Children Who Stutter (UMich_CWS), University of Michigan Social Anxiety Disorder (UMich_SAD), University of Michigan Schizophrenia Gaze Processing (UMich_SZG), and TOP (University of Oslo).</p><p>In addition to the sample-specific inclusion criteria, inclusion criteria for the full sample were based on participants having basic demographic information (age and sex), a T1-weighted MRI volume, and Freesurfer output directories that include summary files that represent left and right hemisphere cortical thickness values of the Destrieux parcellation and subcortical volumetric values (aseg.stats, lh.aparc.a2009s.stats, and rh.aparc.a2009s.stats). Freesurfer image analysis suite (version 6.0) was used for cortical reconstruction and volumetric segmentation for all studies. The technical details of these procedures are described in prior publications (<xref ref-type="bibr" rid="bib6">Dale et al., 1999</xref>; <xref ref-type="bibr" rid="bib13">Fischl et al., 2002</xref>; <xref ref-type="bibr" rid="bib12">Fischl and Dale, 2000</xref>). UK Biobank was the only study for which Freesurfer was not run by the authors. Freesurfer functions <italic>aparcstats2table</italic> and <italic>asegstats2table</italic> were run to extract cortical thickness from the Destrieux parcellation (<xref ref-type="bibr" rid="bib7">Destrieux et al., 2010</xref>) and subcortical volume for all participants into CSV files. These files were inner merged with the demographic files, using Pandas, and NaN rows were dropped.</p><p>QC is an important consideration for large samples and is an active research area (<xref ref-type="bibr" rid="bib1">Alfaro-Almagro et al., 2018</xref>; <xref ref-type="bibr" rid="bib24">Klapwijk et al., 2019</xref>; <xref ref-type="bibr" rid="bib34">Rosen et al., 2018</xref>). We consider manual quality checking of images both prior to and after preprocessing to be the gold standard. However, this is labor intensive and prohibitive for very large samples. Therefore, in this work, we adopt a pragmatic and multi-pronged approach to QC. First, a subset of the full data set underwent manual quality checking (mQC) by author S.R. <ext-link ext-link-type="uri" xlink:href="https://rii-mango.github.io/Papaya/">Papaya</ext-link>, a JavaScript-based image viewer. Manual quality checking was performed during December 2020 when the Netherlands was in full lockdown due to COVID-19 and S.R. was living alone in a new country with a lot of free time. Data included in this manual QC step was based on what was available at the time (<xref ref-type="supplementary-material" rid="supp2">Supplementary file 2</xref>). Later data sets that were included were not manually QC’d due to resource and time constraints. Scripts were used to initialize a manual QC session and track progress and organize ratings. All images (T1w volume and Freesurfer brain.finalsurfs) were put into JSON files that the mQC script would call when loading Papaya. Images were rated using a ‘pass/fail/flag’ scale and the rating was tracked in an automated manner using keyboard inputs (up arrow=pass, down arrow=fail, F key=flag, and left/right arrows were used to move through subjects). Each subject’s T1w volume was viewed in 3D volumetric space, with the Freesurfer brain.finalsurfs file as an overlay, to check for obvious quality issues such as excessive motion, ghosting or ringing artifacts. Example scripts used for quality checking and further instructions for using the manual QC environment can be found on <ext-link ext-link-type="uri" xlink:href="https://github.com/saigerutherford/lifespan_qc_scripts">GitHub</ext-link>(<xref ref-type="bibr" rid="bib38">Rutherford, 2022c</xref> copy archived at <ext-link ext-link-type="uri" xlink:href="https://archive.softwareheritage.org/swh:1:dir:9c98ca93b3fb3b463607286eec7dfc9c4c3e97db;origin=https://github.com/saigerutherford/lifespan_qc_scripts;visit=swh:1:snp:84918033541e80549e91c96e85a29d191321d0a3;anchor=swh:1:rev:70894691c74febe2a4d40ab0c84c50094b9e99ce">swh:1:rev:70894691c74febe2a4d40ab0c84c50094b9e99ce</ext-link>). We relied on ABCD consortium QC procedures for the QC for this sample. The ABCD study data distributes a variable (freesqc01.txt; fsqc_qc = = 1/0) that represents manual quality checking (pass/fail) of the T1w volume and Freesurfer data, thus this data set was added into our manual quality checked data set bringing the sample size to 24,354 individuals passing manual quality checks. Note that QC was performed on the data prior to splitting of the data to assess generalizability. Although this has a reduced sample, we consider this to be a gold-standard sample in that every single scan has been checked manually. All inferences reported in this manuscript were validated against this sample. Second, for the full sample, we adopted an automated QC procedure that quantifies image quality based on the Freesurfer Euler Characteristic (EC), which has been shown to be an excellent proxy for manual labeling of scan quality (<xref ref-type="bibr" rid="bib30">Monereo-Sánchez et al., 2021</xref>; <xref ref-type="bibr" rid="bib34">Rosen et al., 2018</xref>) and is the most important feature in automated scan quality classifiers (<xref ref-type="bibr" rid="bib24">Klapwijk et al., 2019</xref>). Since the distribution of the EC varies across sites, we adopt a simple approach that involves scaling and centering the distribution over the EC across sites and removing samples in the tail of the distribution (see <xref ref-type="bibr" rid="bib23">Kia et al., 2021</xref> for details). While any automated QC heuristic is by definition imperfect, we note that this is based on a conservative inclusion threshold such that only samples well into the tail of the EC distribution are excluded, which are likely to be caused by true topological defects rather than abnormalities due to any underlying pathology. We separated the evaluation metrics into full test set (relying on automated QC) and mQC test set in order to compare model performance between the two QC approaches and were pleased to notice that the evaluation metrics were nearly identical across the two methods.</p><p>Normative modeling was run using python 3.8 and the <ext-link ext-link-type="uri" xlink:href="https://pcntoolkit.readthedocs.io/">PCNtoolkit package</ext-link> (version 0.20). Bayesian Linear Regression (BLR) with likelihood warping was used to predict cortical thickness and subcortical volume from a vector of covariates (age, sex, and site). For a complete mathematical description and explanation of this implementation, see <xref ref-type="bibr" rid="bib14">Fraza et al., 2021</xref>. Briefly, for each brain region of interest (cortical thickness or subcortical volume), <inline-formula><mml:math id="inf1"><mml:mi>y</mml:mi></mml:math></inline-formula> is predicted as:<disp-formula id="equ1"><label>(1)</label><mml:math id="m1"><mml:mi>y</mml:mi><mml:mo>=</mml:mo><mml:msup><mml:mrow><mml:mi>w</mml:mi></mml:mrow><mml:mrow><mml:mi>T</mml:mi></mml:mrow></mml:msup><mml:mi>ϕ</mml:mi><mml:mfenced separators="|"><mml:mrow><mml:mi>x</mml:mi></mml:mrow></mml:mfenced><mml:mo>+</mml:mo><mml:mi>ϵ</mml:mi></mml:math></disp-formula></p><p>where <inline-formula><mml:math id="inf2"><mml:msup><mml:mrow><mml:mi>w</mml:mi></mml:mrow><mml:mrow><mml:mi>T</mml:mi></mml:mrow></mml:msup></mml:math></inline-formula> is the estimated weight vector, <inline-formula><mml:math id="inf3"><mml:mi>ϕ</mml:mi><mml:mfenced separators="|"><mml:mrow><mml:mi>x</mml:mi></mml:mrow></mml:mfenced></mml:math></inline-formula> is a basis expansion of the of covariate vector <bold>x,</bold> consisting of a B-spline basis expansion (cubic spline with five evenly spaced knots) to model non-linear effects of age, and <inline-formula><mml:math id="inf4"><mml:mi>ϵ</mml:mi><mml:mo>=</mml:mo><mml:mi>η</mml:mi><mml:mfenced separators="|"><mml:mrow><mml:mn>0</mml:mn><mml:mo>,</mml:mo><mml:mi>β</mml:mi></mml:mrow></mml:mfenced></mml:math></inline-formula> a Gaussian noise distribution with mean zero and noise precision term β (the inverse variance). A likelihood warping approach (<xref ref-type="bibr" rid="bib33">Rios and Tobar, 2019</xref>; <xref ref-type="bibr" rid="bib39">Snelson et al., 2003</xref>) was used to model non-Gaussian effects. This involves applying a bijective non-linear warping function to the non-Gaussian response variables to map them to a Gaussian latent space where inference can be performed in closed form. We employed a ‘sinarcsinsh’ warping function, which is equivalent to the SHASH distribution commonly used in the generalized additive modeling literature (<xref ref-type="bibr" rid="bib21">Jones and Pewsey, 2009</xref>) and which we have found to perform well in prior work (<xref ref-type="bibr" rid="bib9">Dinga et al., 2021</xref>; <xref ref-type="bibr" rid="bib14">Fraza et al., 2021</xref>). Site variation was modeled using fixed effects, which we have shown in prior work provides relatively good performance (<xref ref-type="bibr" rid="bib23">Kia et al., 2021</xref>), although random effects for site may provide additional flexibility at higher computational cost. A fast numerical optimization algorithm was used to optimize hyperparameters (L-BFGS). Computational complexity of hyperparameter optimization was controlled by minimizing the negative log-likelihood. Deviation scores (Z-scores) are calculated for the n-th subject, and d-th brain area, in the test set as:<disp-formula id="equ2"><label>(2)</label><mml:math id="m2"><mml:mrow><mml:msub><mml:mi>Z</mml:mi><mml:mrow><mml:mi>n</mml:mi><mml:mi>d</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mfrac><mml:mrow><mml:msub><mml:mi>y</mml:mi><mml:mrow><mml:mi>n</mml:mi><mml:mi>d</mml:mi></mml:mrow></mml:msub><mml:mo>−</mml:mo><mml:msub><mml:mrow><mml:mover><mml:mi>y</mml:mi><mml:mo stretchy="false">^</mml:mo></mml:mover></mml:mrow><mml:mrow><mml:mi>n</mml:mi><mml:mi>d</mml:mi></mml:mrow></mml:msub></mml:mrow><mml:msqrt><mml:msubsup><mml:mi>σ</mml:mi><mml:mrow><mml:mi>d</mml:mi></mml:mrow><mml:mrow><mml:mn>2</mml:mn></mml:mrow></mml:msubsup><mml:mo>+</mml:mo><mml:mo stretchy="false">(</mml:mo><mml:msubsup><mml:mi>σ</mml:mi><mml:mrow><mml:mo>∗</mml:mo></mml:mrow><mml:mrow><mml:mn>2</mml:mn></mml:mrow></mml:msubsup><mml:msub><mml:mo stretchy="false">)</mml:mo><mml:mrow><mml:mi>d</mml:mi></mml:mrow></mml:msub></mml:msqrt></mml:mfrac></mml:mrow></mml:math></disp-formula></p><p>Where <inline-formula><mml:math id="inf5"><mml:msub><mml:mrow><mml:mi>y</mml:mi></mml:mrow><mml:mrow><mml:mi>n</mml:mi><mml:mi>d</mml:mi></mml:mrow></mml:msub></mml:math></inline-formula> is the true response, <inline-formula><mml:math id="inf6"><mml:mstyle displaystyle="true" scriptlevel="0"><mml:mrow><mml:msub><mml:mrow><mml:mover><mml:mi>y</mml:mi><mml:mo stretchy="false">^</mml:mo></mml:mover></mml:mrow><mml:mrow><mml:mi>n</mml:mi><mml:mi>d</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:mstyle></mml:math></inline-formula> is the predicted mean, <inline-formula><mml:math id="inf7"><mml:msubsup><mml:mrow><mml:mi>σ</mml:mi></mml:mrow><mml:mrow><mml:mi>d</mml:mi></mml:mrow><mml:mrow><mml:mn>2</mml:mn></mml:mrow></mml:msubsup></mml:math></inline-formula> is the estimated noise variance (reflecting uncertainty in the data), and <inline-formula><mml:math id="inf8"><mml:msub><mml:mrow><mml:mfenced separators="|"><mml:mrow><mml:msubsup><mml:mrow><mml:mi>σ</mml:mi></mml:mrow><mml:mrow/><mml:mrow><mml:mn>2</mml:mn></mml:mrow></mml:msubsup></mml:mrow></mml:mfenced></mml:mrow><mml:mrow><mml:mi>d</mml:mi></mml:mrow></mml:msub></mml:math></inline-formula> is the variance attributed to modeling uncertainty. Model fit for each brain region was evaluated by calculating the explained variance (which measures central tendency), the mean squared log-loss (MSLL, central tendency, and variance) plus skew and kurtosis of the deviation scores (2) which measures how well the shape of the regression function matches the data (<xref ref-type="bibr" rid="bib9">Dinga et al., 2021</xref>). Note that for all models, we report out of sample metrics.</p><p>To provide a summary of individual variation within each clinical group, deviation scores were summarized for each clinical group (<xref ref-type="fig" rid="fig2">Figure 2B–C</xref>) by first separating them into positive and negative deviations, counting how many subjects had an extreme deviation (positive extreme deviation defined as Z>2, negative extreme deviation as Z<−2) at a given ROI, and then dividing by the group size to show the percentage of individuals with extreme deviations at that brain area. Controls from the same sites as the patient groups were summarized in the same manner for comparison. We also performed classical case versus control group difference testing on the true data and on the deviation scores (<xref ref-type="fig" rid="fig2">Figure 2D</xref>) and thresholded results at a Benjamini-Hochberg false discovery rate of p<0.05. Note that in both cases, we directly contrast each patient group to their matched controls to avoid nuisance variation confounding any reported effects (e.g., sampling characteristics and demographic differences).</p><p>All pretrained models and code are shared online with straightforward directions for transferring to new sites and including an example transfer data set derived from several <ext-link ext-link-type="uri" xlink:href="https://openneuro.org/">OpenNeuro.org</ext-link> data sets. Given a new set of data (e.g., sites not present in the training set), this is done by first applying the warp parameters estimating on the training data to the new data set, adjusting the mean and variance in the latent Gaussian space, then (if necessary) warping the adjusted data back to the original space, which is similar to the approach outlined in <xref ref-type="bibr" rid="bib9">Dinga et al., 2021</xref>. Note that to remain unbiased, this should be done on a held-out calibration data set. To illustrate this procedure, we apply this approach to predicting a subset of data that was not used during the model estimation step. We leveraged data from <ext-link ext-link-type="uri" xlink:href="https://openneuro.org/">OpenNeuro.org</ext-link> (<xref ref-type="bibr" rid="bib26">Markiewicz et al., 2021</xref>) to create a transfer data set (six sites, N=546, <xref ref-type="supplementary-material" rid="supp3">Supplementary file 3</xref>). This data are provided along with the code for transferring to walk users through how to apply these models to their own data. These results are reported in <xref ref-type="fig" rid="fig3">Figure 3</xref> (transfer) and <xref ref-type="supplementary-material" rid="supp3">Supplementary file 3</xref>. We also distribute scripts for this purpose in the GitHub Repository associated with this manuscript. Furthermore, to promote the use of these models and remove barriers to using them, we have set up access to the pretrained models and code for transferring to users’ own data, using <ext-link ext-link-type="uri" xlink:href="https://colab.research.google.com/github/predictive-clinical-neuroscience/braincharts/blob/master/scripts/apply_normative_models.ipynb">Google Colab</ext-link>, a free, cloud-based platform for running python notebooks. This eliminates the need to install python/manage package versions and only requires users to have a personal computer with stable internet connection.</p></sec></body><back><sec id="s5" sec-type="additional-information"><title>Additional information</title><fn-group content-type="competing-interest"><title>Competing interests</title><fn fn-type="COI-statement" id="conf1"><p>No competing interests declared</p></fn><fn fn-type="COI-statement" id="conf2"><p>No competing interests declared</p></fn><fn fn-type="COI-statement" id="conf3"><p>is a consultant for HealthLytix and received speaker's honorarium from Lundbeck and Sunovion</p></fn><fn fn-type="COI-statement" id="conf4"><p>received speaker's honorarium from Lundbeck and Janssen</p></fn><fn fn-type="COI-statement" id="conf5"><p>is director and shareholder of SBGNeuro Ltd</p></fn></fn-group><fn-group content-type="author-contribution"><title>Author contributions</title><fn fn-type="con" id="con1"><p>Conceptualization, Data curation, Formal analysis, Methodology, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review and editing</p></fn><fn fn-type="con" id="con2"><p>Methodology, Software, Writing – review and editing</p></fn><fn fn-type="con" id="con3"><p>Methodology</p></fn><fn fn-type="con" id="con4"><p>Methodology, Software, Writing – review and editing</p></fn><fn fn-type="con" id="con5"><p>Methodology, Software, Writing – review and editing</p></fn><fn fn-type="con" id="con6"><p>Methodology, Software, Writing – review and editing</p></fn><fn fn-type="con" id="con7"><p>Methodology, Software, Writing – review and editing</p></fn><fn fn-type="con" id="con8"><p>Methodology, Writing – review and editing</p></fn><fn fn-type="con" id="con9"><p>Methodology, Writing – review and editing</p></fn><fn fn-type="con" id="con10"><p>Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con11"><p>Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con12"><p>Methodology, Writing – review and editing</p></fn><fn fn-type="con" id="con13"><p>Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con14"><p>Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con15"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con16"><p>Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con17"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con18"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con19"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con20"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con21"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con22"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con23"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con24"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con25"><p>Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con26"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con27"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con28"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con29"><p>Funding acquisition, Resources, Writing – review and editing</p></fn><fn fn-type="con" id="con30"><p>Conceptualization, Funding acquisition, Resources, Supervision, Writing – original draft, Writing – review and editing</p></fn><fn fn-type="con" id="con31"><p>Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review and editing</p></fn><fn fn-type="con" id="con32"><p>Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Visualization, Writing – original draft, Writing – review and editing</p></fn></fn-group><fn-group content-type="ethics-information"><title>Ethics</title><fn fn-type="other"><p>Human subjects: Ethical approval for the public data were provided by the relevant local research authorities for the studies contributing data. For full details see the main study publications given in the main text. For all clinical studies, approval was obtained via the local ethical review authorities, i.e.: TOP: Regional Committee for Medical & Health Research Ethics South East Norway. Approval number: 2009/2485- C, KCL: South Manchester NHS National Research Ethics Service. Approval number: 07/H1003/194. Delta: The local ethics committee of the Academic Medical Center of the University of Amsterdam (AMC-METC) Nr.:11/050, UMich_IMPS: University of Michigan Institution Review Board HUM00088188, UMich_SZG: University of Michigan Institution Review Board HUM00080457, UMich_MLS: University of Michigan Institution Review Board HUM00040642, UMich_CWS: MSU Biomedical and Health Institutional Review Board (BIIRB) IRB#09-810, UMich_MTWiNS: University of Michigan Institution Review HUM00163965, UCDavis: University of California Davis Institutional Review Board IRB ID: 220915, 592866, 1097084.</p></fn></fn-group></sec><sec id="s6" sec-type="supplementary-material"><title>Additional files</title><supplementary-material id="supp1"><label>Supplementary file 1.</label><caption><title>Full Sample Description, per site, in the train, test, and clinical validation sets.</title></caption><media mime-subtype="docx" mimetype="application" xlink:href="elife-72904-supp1-v2.docx"/></supplementary-material><supplementary-material id="supp2"><label>Supplementary file 2.</label><caption><title>Manual Quality Checked (mQC) Sample Description, per site, train and test sets.</title></caption><media mime-subtype="docx" mimetype="application" xlink:href="elife-72904-supp2-v2.docx"/></supplementary-material><supplementary-material id="supp3"><label>Supplementary file 3.</label><caption><title>Transfer Test Set Sample Descriptions, per site.</title><p>Data were downloaded from <ext-link ext-link-type="uri" xlink:href="https://openneuro.org/">https://openneuro.org/</ext-link> (<xref ref-type="bibr" rid="bib26">Markiewicz et al., 2021</xref>) and are shared on GitHub along with code to transfer to demonstrate how to re-use models on new data.</p></caption><media mime-subtype="docx" mimetype="application" xlink:href="elife-72904-supp3-v2.docx"/></supplementary-material><supplementary-material id="transrepform"><label>Transparent reporting form</label><media mime-subtype="pdf" mimetype="application" xlink:href="elife-72904-transrepform1-v2.pdf"/></supplementary-material></sec><sec id="s7" sec-type="data-availability"><title>Data availability</title><p>All pre-trained models and code for transferring to new sites are shared online via GitHub (<ext-link ext-link-type="uri" xlink:href="https://github.com/predictive-clinical-neuroscience/braincharts">https://github.com/predictive-clinical-neuroscience/braincharts</ext-link>, copy archived at <ext-link ext-link-type="uri" xlink:href="https://archive.softwareheritage.org/swh:1:dir:cd28f07707409fed9c4e3bcfefaa08b9638dbccb;origin=https://github.com/predictive-clinical-neuroscience/braincharts;visit=swh:1:snp:ad9fb4794977fbb77f867431c767d66d31781b3f;anchor=swh:1:rev:ee2b7ebcb46bab0f302f73f8d6fc913f63fccda5">swh:1:rev:ee2b7ebcb46bab0f302f73f8d6fc913f63fccda5</ext-link>). We have also shared the models on Zenodo (<ext-link ext-link-type="uri" xlink:href="https://zenodo.org/record/5535467#.YVRECmYzZhF">https://zenodo.org/record/5535467#.YVRECmYzZhF</ext-link>).</p><p>The following dataset was generated:</p><p><element-citation id="dataset1" publication-type="data" specific-use="isSupplementedBy"><person-group person-group-type="author"><name><surname>Rutherford</surname><given-names>S</given-names></name><name><surname>Andre</surname><given-names>M</given-names></name></person-group><year iso-8601-date="2021">2021</year><data-title>Braincharts</data-title><source>Zenodo</source><pub-id pub-id-type="doi">10.5281/zenodo.5535467</pub-id></element-citation></p><p>The following previously published datasets were used:</p><p><element-citation id="dataset2" publication-type="data" specific-use="references"><person-group person-group-type="author"><collab>National Institute of Mental Health</collab></person-group><year iso-8601-date="2020">2020</year><data-title>ABCD</data-title><source>nih</source><pub-id pub-id-type="doi">10.15154/1503209</pub-id></element-citation></p><p><element-citation id="dataset3" publication-type="data" specific-use="references"><person-group person-group-type="author"><collab>Cambridge Centre for Ageing and Neuroscience</collab></person-group><year iso-8601-date="2010">2010</year><data-title>CAMCAN</data-title><source>NCBI</source><pub-id pub-id-type="accession" xlink:href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4219118/">PMC4219118</pub-id></element-citation></p></sec><ack id="ack"><title>Acknowledgements</title><p>This research was supported by grants from the European Research Council (ERC, grant ‘MENTALPRECISION’ 10100118 and ‘BRAINMINT’ 802998), the Wellcome Trust under an Innovator award (‘BRAINCHART,’ 215698/Z/19/Z) and a Strategic Award (098369/Z/12/Z), the Dutch Organisation for Scientific Research (VIDI grant 016.156.415) the Research Council of Norway (223273, 249795, 298646, 300768, and 276082), the South-Eastern Norway Regional Health Authority (2014097, 2015073, 2016083, and 2019101), the KG Jebsen Stiftelsen, an Autism Center of Excellence grant awarded by the National Institute of Child Health and Development (NICHD) (P50 HD093079) as well as the National Institute of Mental Health (R01MH104438 and R01MH103371). TW also gratefully acknowledges the Niels Stensen Fellowship as well as the European Union’s Horizon 2020 Research and Innovation Program under the Marie Sklodowska-Curie Grant Agreement no. 895011. RZ was funded by Medical Research Council grant (G0902304). IFT was funded by National Institute of Mental Health K23MH108823. SC was funded by National Institute on Deafness and other Communication Disorders (NIDCD/NIH) grant R01DC011277. CS was funded by the National Institute of Mental Health R01MH107741. LD was funded by Michigan Institute for Clinical Health Research (MICHR) Pilot Grant Program (UL1TR002240) through an NIH Clinical and Translational Science Award (CTSA). MTwiNS was supported by the National Institute of Mental Health and the Office of the Director National Institute of Health, under Award Number UG3MH114249 and the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health under Award number R01HD093334 to SAB and LWH. RJTM was funded by an ABC Talent Grant. The ABCD Study is supported by the National Institutes of Health (NIH) and additional federal partners under Award numbers U01DA041022, U01DA041028, U01DA041048, U01DA041089, U01DA041106, U01DA041117, U01DA041120, U01DA041134, U01DA041148, U01DA041156, U01DA041174, U24DA041123, and U24DA041147.</p></ack><ref-list><title>References</title><ref id="bib1"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Alfaro-Almagro</surname><given-names>F</given-names></name><name><surname>Jenkinson</surname><given-names>M</given-names></name><name><surname>Bangerter</surname><given-names>NK</given-names></name><name><surname>Andersson</surname><given-names>JLR</given-names></name><name><surname>Griffanti</surname><given-names>L</given-names></name><name><surname>Douaud</surname><given-names>G</given-names></name><name><surname>Sotiropoulos</surname><given-names>SN</given-names></name><name><surname>Jbabdi</surname><given-names>S</given-names></name><name><surname>Hernandez-Fernandez</surname><given-names>M</given-names></name><name><surname>Vallee</surname><given-names>E</given-names></name><name><surname>Vidaurre</surname><given-names>D</given-names></name><name><surname>Webster</surname><given-names>M</given-names></name><name><surname>McCarthy</surname><given-names>P</given-names></name><name><surname>Rorden</surname><given-names>C</given-names></name><name><surname>Daducci</surname><given-names>A</given-names></name><name><surname>Alexander</surname><given-names>DC</given-names></name><name><surname>Zhang</surname><given-names>H</given-names></name><name><surname>Dragonu</surname><given-names>I</given-names></name><name><surname>Matthews</surname><given-names>PM</given-names></name><name><surname>Miller</surname><given-names>KL</given-names></name><name><surname>Smith</surname><given-names>SM</given-names></name></person-group><year iso-8601-date="2018">2018</year><article-title>Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank</article-title><source>NeuroImage</source><volume>166</volume><fpage>400</fpage><lpage>424</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2017.10.034</pub-id><pub-id pub-id-type="pmid">29079522</pub-id></element-citation></ref><ref id="bib2"><element-citation publication-type="preprint"><person-group person-group-type="author"><name><surname>Bayer</surname><given-names>JMM</given-names></name><name><surname>Dinga</surname><given-names>R</given-names></name><name><surname>Kia</surname><given-names>SM</given-names></name><name><surname>Kottaram</surname><given-names>AR</given-names></name><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Lv</surname><given-names>J</given-names></name><name><surname>Zalesky</surname><given-names>A</given-names></name><name><surname>Schmaal</surname><given-names>L</given-names></name><name><surname>Marquand</surname><given-names>A</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Accommodating Site Variation in Neuroimaging Data Using Normative and Hierarchical Bayesian Models</article-title><source>bioRxiv</source><pub-id pub-id-type="doi">10.1101/2021.02.09.430363</pub-id></element-citation></ref><ref id="bib3"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Bethlehem</surname><given-names>RAI</given-names></name><name><surname>Seidlitz</surname><given-names>J</given-names></name><name><surname>Romero-Garcia</surname><given-names>R</given-names></name><name><surname>Trakoshis</surname><given-names>S</given-names></name><name><surname>Dumas</surname><given-names>G</given-names></name><name><surname>Lombardo</surname><given-names>MV</given-names></name></person-group><year iso-8601-date="2020">2020</year><article-title>A normative modelling approach reveals age-atypical cortical thickness in a subgroup of males with autism spectrum disorder</article-title><source>Communications Biology</source><volume>3</volume><elocation-id>486</elocation-id><pub-id pub-id-type="doi">10.1038/s42003-020-01212-9</pub-id><pub-id pub-id-type="pmid">32887930</pub-id></element-citation></ref><ref id="bib4"><element-citation publication-type="preprint"><person-group person-group-type="author"><name><surname>Bethlehem</surname><given-names>R</given-names></name><name><surname>Seidlitz</surname><given-names>J</given-names></name><name><surname>White</surname><given-names>SR</given-names></name><name><surname>Vogel</surname><given-names>JW</given-names></name><name><surname>Anderson</surname><given-names>KM</given-names></name><name><surname>Adamson</surname><given-names>C</given-names></name><name><surname>Adler</surname><given-names>S</given-names></name><name><surname>Alexopoulos</surname><given-names>GS</given-names></name><name><surname>Anagnostou</surname><given-names>E</given-names></name><name><surname>Areces-Gonzalez</surname><given-names>A</given-names></name><name><surname>Astle</surname><given-names>DE</given-names></name><name><surname>Auyeung</surname><given-names>B</given-names></name><name><surname>Ayub</surname><given-names>M</given-names></name><name><surname>Ball</surname><given-names>G</given-names></name><name><surname>Baron-Cohen</surname><given-names>S</given-names></name><name><surname>Beare</surname><given-names>R</given-names></name><name><surname>Bedford</surname><given-names>SA</given-names></name><name><surname>Benegal</surname><given-names>V</given-names></name><name><surname>Beyer</surname><given-names>F</given-names></name><name><surname>Alexander-Bloch</surname><given-names>AF</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Brain Charts for the Human Lifespan</article-title><source>bioRxiv</source><pub-id pub-id-type="doi">10.1101/2021.06.08.447489</pub-id></element-citation></ref><ref id="bib5"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cole</surname><given-names>TJ</given-names></name></person-group><year iso-8601-date="2012">2012</year><article-title>The development of growth references and growth charts</article-title><source>Annals of Human Biology</source><volume>39</volume><fpage>382</fpage><lpage>394</lpage><pub-id pub-id-type="doi">10.3109/03014460.2012.694475</pub-id><pub-id pub-id-type="pmid">22780429</pub-id></element-citation></ref><ref id="bib6"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Dale</surname><given-names>AM</given-names></name><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Sereno</surname><given-names>MI</given-names></name></person-group><year iso-8601-date="1999">1999</year><article-title>Cortical Surface-Based Analysis</article-title><source>NeuroImage</source><volume>9</volume><fpage>179</fpage><lpage>194</lpage><pub-id pub-id-type="doi">10.1006/nimg.1998.0395</pub-id><pub-id pub-id-type="pmid">9931268</pub-id></element-citation></ref><ref id="bib7"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Destrieux</surname><given-names>C</given-names></name><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Dale</surname><given-names>A</given-names></name><name><surname>Halgren</surname><given-names>E</given-names></name></person-group><year iso-8601-date="2010">2010</year><article-title>Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature</article-title><source>NeuroImage</source><volume>53</volume><fpage>1</fpage><lpage>15</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2010.06.010</pub-id><pub-id pub-id-type="pmid">20547229</pub-id></element-citation></ref><ref id="bib8"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Dimitrova</surname><given-names>R</given-names></name><name><surname>Pietsch</surname><given-names>M</given-names></name><name><surname>Christiaens</surname><given-names>D</given-names></name><name><surname>Ciarrusta</surname><given-names>J</given-names></name><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Batalle</surname><given-names>D</given-names></name><name><surname>Hughes</surname><given-names>E</given-names></name><name><surname>Hutter</surname><given-names>J</given-names></name><name><surname>Cordero-Grande</surname><given-names>L</given-names></name><name><surname>Price</surname><given-names>AN</given-names></name><name><surname>Chew</surname><given-names>A</given-names></name><name><surname>Falconer</surname><given-names>S</given-names></name><name><surname>Vecchiato</surname><given-names>K</given-names></name><name><surname>Steinweg</surname><given-names>JK</given-names></name><name><surname>Carney</surname><given-names>O</given-names></name><name><surname>Rutherford</surname><given-names>MA</given-names></name><name><surname>Tournier</surname><given-names>J-D</given-names></name><name><surname>Counsell</surname><given-names>SJ</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name><name><surname>Rueckert</surname><given-names>D</given-names></name><name><surname>Hajnal</surname><given-names>JV</given-names></name><name><surname>McAlonan</surname><given-names>G</given-names></name><name><surname>Edwards</surname><given-names>AD</given-names></name><name><surname>O’Muircheartaigh</surname><given-names>J</given-names></name></person-group><year iso-8601-date="2020">2020</year><article-title>Heterogeneity in Brain Microstructural Development Following Preterm Birth</article-title><source>Cerebral Cortex (New York, N.Y</source><volume>30</volume><fpage>4800</fpage><lpage>4810</lpage><pub-id pub-id-type="doi">10.1093/cercor/bhaa069</pub-id><pub-id pub-id-type="pmid">32306044</pub-id></element-citation></ref><ref id="bib9"><element-citation publication-type="preprint"><person-group person-group-type="author"><name><surname>Dinga</surname><given-names>R</given-names></name><name><surname>Fraza</surname><given-names>CJ</given-names></name><name><surname>Bayer</surname><given-names>JMM</given-names></name><name><surname>Kia</surname><given-names>SM</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Normative Modeling of Neuroimaging Data Using Generalized Additive Models of Location Scale and Shape</article-title><source>bioRxiv</source><pub-id pub-id-type="doi">10.1101/2021.06.14.448106</pub-id></element-citation></ref><ref id="bib10"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ducharme</surname><given-names>S</given-names></name><name><surname>Albaugh</surname><given-names>MD</given-names></name><name><surname>Nguyen</surname><given-names>TV</given-names></name><name><surname>Hudziak</surname><given-names>JJ</given-names></name><name><surname>Mateos-Pérez</surname><given-names>JM</given-names></name><name><surname>Labbe</surname><given-names>A</given-names></name><name><surname>Evans</surname><given-names>AC</given-names></name><name><surname>Karama</surname><given-names>S</given-names></name><collab>Brain Development Cooperative Group</collab></person-group><year iso-8601-date="2016">2016</year><article-title>Trajectories of cortical thickness maturation in normal brain development--The importance of quality control procedures</article-title><source>NeuroImage</source><volume>125</volume><fpage>267</fpage><lpage>279</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2015.10.010</pub-id><pub-id pub-id-type="pmid">26463175</pub-id></element-citation></ref><ref id="bib11"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Duval</surname><given-names>ER</given-names></name><name><surname>Joshi</surname><given-names>SA</given-names></name><name><surname>Russman Block</surname><given-names>S</given-names></name><name><surname>Abelson</surname><given-names>JL</given-names></name><name><surname>Liberzon</surname><given-names>I</given-names></name></person-group><year iso-8601-date="2018">2018</year><article-title>Insula activation is modulated by attention shifting in social anxiety disorder</article-title><source>Journal of Anxiety Disorders</source><volume>56</volume><fpage>56</fpage><lpage>62</lpage><pub-id pub-id-type="doi">10.1016/j.janxdis.2018.04.004</pub-id><pub-id pub-id-type="pmid">29729828</pub-id></element-citation></ref><ref id="bib12"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Dale</surname><given-names>AM</given-names></name></person-group><year iso-8601-date="2000">2000</year><article-title>Measuring the thickness of the human cerebral cortex from magnetic resonance images</article-title><source>PNAS</source><volume>97</volume><fpage>11050</fpage><lpage>11055</lpage><pub-id pub-id-type="doi">10.1073/pnas.200033797</pub-id><pub-id pub-id-type="pmid">10984517</pub-id></element-citation></ref><ref id="bib13"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fischl</surname><given-names>B</given-names></name><name><surname>Salat</surname><given-names>DH</given-names></name><name><surname>Busa</surname><given-names>E</given-names></name><name><surname>Albert</surname><given-names>M</given-names></name><name><surname>Dieterich</surname><given-names>M</given-names></name><name><surname>Haselgrove</surname><given-names>C</given-names></name><name><surname>van der Kouwe</surname><given-names>A</given-names></name><name><surname>Killiany</surname><given-names>R</given-names></name><name><surname>Kennedy</surname><given-names>D</given-names></name><name><surname>Klaveness</surname><given-names>S</given-names></name><name><surname>Montillo</surname><given-names>A</given-names></name><name><surname>Makris</surname><given-names>N</given-names></name><name><surname>Rosen</surname><given-names>B</given-names></name><name><surname>Dale</surname><given-names>AM</given-names></name></person-group><year iso-8601-date="2002">2002</year><article-title>Whole Brain Segmentation</article-title><source>Neuron</source><volume>33</volume><fpage>341</fpage><lpage>355</lpage><pub-id pub-id-type="doi">10.1016/S0896-6273(02)00569-X</pub-id><pub-id pub-id-type="pmid">11832223</pub-id></element-citation></ref><ref id="bib14"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fraza</surname><given-names>CJ</given-names></name><name><surname>Dinga</surname><given-names>R</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Warped Bayesian linear regression for normative modelling of big data</article-title><source>NeuroImage</source><volume>245</volume><elocation-id>118715</elocation-id><pub-id pub-id-type="doi">10.1016/j.neuroimage.2021.118715</pub-id><pub-id pub-id-type="pmid">34798518</pub-id></element-citation></ref><ref id="bib15"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fry</surname><given-names>A</given-names></name><name><surname>Littlejohns</surname><given-names>TJ</given-names></name><name><surname>Sudlow</surname><given-names>C</given-names></name><name><surname>Doherty</surname><given-names>N</given-names></name><name><surname>Adamska</surname><given-names>L</given-names></name><name><surname>Sprosen</surname><given-names>T</given-names></name><name><surname>Collins</surname><given-names>R</given-names></name><name><surname>Allen</surname><given-names>NE</given-names></name></person-group><year iso-8601-date="2017">2017</year><article-title>Comparison of Sociodemographic and Health-Related Characteristics of UK Biobank Participants With Those of the General Population</article-title><source>American Journal of Epidemiology</source><volume>186</volume><fpage>1026</fpage><lpage>1034</lpage><pub-id pub-id-type="doi">10.1093/aje/kwx246</pub-id><pub-id pub-id-type="pmid">28641372</pub-id></element-citation></ref><ref id="bib16"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Gogtay</surname><given-names>N</given-names></name><name><surname>Giedd</surname><given-names>JN</given-names></name><name><surname>Lusk</surname><given-names>L</given-names></name><name><surname>Hayashi</surname><given-names>KM</given-names></name><name><surname>Greenstein</surname><given-names>D</given-names></name><name><surname>Vaituzis</surname><given-names>AC</given-names></name><name><surname>Nugent</surname><given-names>TF</given-names></name><name><surname>Herman</surname><given-names>DH</given-names></name><name><surname>Clasen</surname><given-names>LS</given-names></name><name><surname>Toga</surname><given-names>AW</given-names></name><name><surname>Rapoport</surname><given-names>JL</given-names></name><name><surname>Thompson</surname><given-names>PM</given-names></name></person-group><year iso-8601-date="2004">2004</year><article-title>Dynamic mapping of human cortical development during childhood through early adulthood</article-title><source>PNAS</source><volume>101</volume><fpage>8174</fpage><lpage>8179</lpage><pub-id pub-id-type="doi">10.1073/pnas.0402680101</pub-id><pub-id pub-id-type="pmid">15148381</pub-id></element-citation></ref><ref id="bib17"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Green</surname><given-names>S</given-names></name><name><surname>Lambon Ralph</surname><given-names>MA</given-names></name><name><surname>Moll</surname><given-names>J</given-names></name><name><surname>Deakin</surname><given-names>JFW</given-names></name><name><surname>Zahn</surname><given-names>R</given-names></name></person-group><year iso-8601-date="2012">2012</year><article-title>Guilt-selective functional disconnection of anterior temporal and subgenual cortices in major depressive disorder</article-title><source>Archives of General Psychiatry</source><volume>69</volume><fpage>1014</fpage><lpage>1021</lpage><pub-id pub-id-type="doi">10.1001/archgenpsychiatry.2012.135</pub-id><pub-id pub-id-type="pmid">22638494</pub-id></element-citation></ref><ref id="bib18"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Henrich</surname><given-names>J</given-names></name><name><surname>Heine</surname><given-names>SJ</given-names></name><name><surname>Norenzayan</surname><given-names>A</given-names></name></person-group><year iso-8601-date="2010">2010</year><article-title>The weirdest people in the world?</article-title><source>The Behavioral and Brain Sciences</source><volume>33</volume><fpage>61</fpage><lpage>83</lpage><pub-id pub-id-type="doi">10.1017/S0140525X0999152X</pub-id><pub-id pub-id-type="pmid">20550733</pub-id></element-citation></ref><ref id="bib19"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Insel</surname><given-names>TR</given-names></name></person-group><year iso-8601-date="2014">2014</year><article-title>Mental disorders in childhood: shifting the focus from behavioral symptoms to neurodevelopmental trajectories</article-title><source>JAMA</source><volume>311</volume><fpage>1727</fpage><lpage>1728</lpage><pub-id pub-id-type="doi">10.1001/jama.2014.1193</pub-id><pub-id pub-id-type="pmid">24794359</pub-id></element-citation></ref><ref id="bib20"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jack</surname><given-names>CR</given-names></name><name><surname>Knopman</surname><given-names>DS</given-names></name><name><surname>Jagust</surname><given-names>WJ</given-names></name><name><surname>Shaw</surname><given-names>LM</given-names></name><name><surname>Aisen</surname><given-names>PS</given-names></name><name><surname>Weiner</surname><given-names>MW</given-names></name><name><surname>Petersen</surname><given-names>RC</given-names></name><name><surname>Trojanowski</surname><given-names>JQ</given-names></name></person-group><year iso-8601-date="2010">2010</year><article-title>Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade</article-title><source>The Lancet. Neurology</source><volume>9</volume><fpage>119</fpage><lpage>128</lpage><pub-id pub-id-type="doi">10.1016/S1474-4422(09)70299-6</pub-id><pub-id pub-id-type="pmid">20083042</pub-id></element-citation></ref><ref id="bib21"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jones</surname><given-names>MC</given-names></name><name><surname>Pewsey</surname><given-names>A</given-names></name></person-group><year iso-8601-date="2009">2009</year><article-title>Sinh-arcsinh distributions</article-title><source>Biometrika</source><volume>96</volume><fpage>761</fpage><lpage>780</lpage><pub-id pub-id-type="doi">10.1093/biomet/asp053</pub-id></element-citation></ref><ref id="bib22"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Karas</surname><given-names>GB</given-names></name><name><surname>Scheltens</surname><given-names>P</given-names></name><name><surname>Rombouts</surname><given-names>SARB</given-names></name><name><surname>Visser</surname><given-names>PJ</given-names></name><name><surname>van Schijndel</surname><given-names>RA</given-names></name><name><surname>Fox</surname><given-names>NC</given-names></name><name><surname>Barkhof</surname><given-names>F</given-names></name></person-group><year iso-8601-date="2004">2004</year><article-title>Global and local gray matter loss in mild cognitive impairment and Alzheimer’s disease</article-title><source>NeuroImage</source><volume>23</volume><fpage>708</fpage><lpage>716</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2004.07.006</pub-id><pub-id pub-id-type="pmid">15488420</pub-id></element-citation></ref><ref id="bib23"><element-citation publication-type="preprint"><person-group person-group-type="author"><name><surname>Kia</surname><given-names>SM</given-names></name><name><surname>Huijsdens</surname><given-names>H</given-names></name><name><surname>Rutherford</surname><given-names>S</given-names></name><name><surname>Dinga</surname><given-names>R</given-names></name><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Mennes</surname><given-names>M</given-names></name><name><surname>Andreassen</surname><given-names>OA</given-names></name><name><surname>Westlye</surname><given-names>LT</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Federated Multi-Site Normative Modeling Using Hierarchical Bayesian Regression</article-title><source>bioRxiv</source><pub-id pub-id-type="doi">10.1101/2021.05.28.446120</pub-id></element-citation></ref><ref id="bib24"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Klapwijk</surname><given-names>ET</given-names></name><name><surname>van de Kamp</surname><given-names>F</given-names></name><name><surname>van der Meulen</surname><given-names>M</given-names></name><name><surname>Peters</surname><given-names>S</given-names></name><name><surname>Wierenga</surname><given-names>LM</given-names></name></person-group><year iso-8601-date="2019">2019</year><article-title>Qoala-T: A supervised-learning tool for quality control of FreeSurfer segmented MRI data</article-title><source>NeuroImage</source><volume>189</volume><fpage>116</fpage><lpage>129</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2019.01.014</pub-id><pub-id pub-id-type="pmid">30633965</pub-id></element-citation></ref><ref id="bib25"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Lythe</surname><given-names>KE</given-names></name><name><surname>Moll</surname><given-names>J</given-names></name><name><surname>Gethin</surname><given-names>JA</given-names></name><name><surname>Workman</surname><given-names>CI</given-names></name><name><surname>Green</surname><given-names>S</given-names></name><name><surname>Lambon Ralph</surname><given-names>MA</given-names></name><name><surname>Deakin</surname><given-names>JFW</given-names></name><name><surname>Zahn</surname><given-names>R</given-names></name></person-group><year iso-8601-date="2015">2015</year><article-title>Self-blame-Selective Hyperconnectivity Between Anterior Temporal and Subgenual Cortices and Prediction of Recurrent Depressive Episodes</article-title><source>JAMA Psychiatry</source><volume>72</volume><fpage>1119</fpage><lpage>1126</lpage><pub-id pub-id-type="doi">10.1001/jamapsychiatry.2015.1813</pub-id><pub-id pub-id-type="pmid">26445229</pub-id></element-citation></ref><ref id="bib26"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Markiewicz</surname><given-names>CJ</given-names></name><name><surname>Gorgolewski</surname><given-names>KJ</given-names></name><name><surname>Feingold</surname><given-names>F</given-names></name><name><surname>Blair</surname><given-names>R</given-names></name><name><surname>Halchenko</surname><given-names>YO</given-names></name><name><surname>Miller</surname><given-names>E</given-names></name><name><surname>Hardcastle</surname><given-names>N</given-names></name><name><surname>Wexler</surname><given-names>J</given-names></name><name><surname>Esteban</surname><given-names>O</given-names></name><name><surname>Goncavles</surname><given-names>M</given-names></name><name><surname>Jwa</surname><given-names>A</given-names></name><name><surname>Poldrack</surname><given-names>R</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>The OpenNeuro resource for sharing of neuroscience data</article-title><source>eLife</source><volume>10</volume><elocation-id>e71774</elocation-id><pub-id pub-id-type="doi">10.7554/eLife.71774</pub-id><pub-id pub-id-type="pmid">34658334</pub-id></element-citation></ref><ref id="bib27"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Marquand</surname><given-names>AF</given-names></name><name><surname>Rezek</surname><given-names>I</given-names></name><name><surname>Buitelaar</surname><given-names>J</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name></person-group><year iso-8601-date="2016">2016</year><article-title>Understanding Heterogeneity in Clinical Cohorts Using Normative Models: Beyond Case-Control Studies</article-title><source>Biological Psychiatry</source><volume>80</volume><fpage>552</fpage><lpage>561</lpage><pub-id pub-id-type="doi">10.1016/j.biopsych.2015.12.023</pub-id><pub-id pub-id-type="pmid">26927419</pub-id></element-citation></ref><ref id="bib28"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Marquand</surname><given-names>AF</given-names></name><name><surname>Kia</surname><given-names>SM</given-names></name><name><surname>Zabihi</surname><given-names>M</given-names></name><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Buitelaar</surname><given-names>JK</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name></person-group><year iso-8601-date="2019">2019</year><article-title>Conceptualizing mental disorders as deviations from normative functioning</article-title><source>Molecular Psychiatry</source><volume>24</volume><fpage>1415</fpage><lpage>1424</lpage><pub-id pub-id-type="doi">10.1038/s41380-019-0441-1</pub-id><pub-id pub-id-type="pmid">31201374</pub-id></element-citation></ref><ref id="bib29"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mocking</surname><given-names>RJT</given-names></name><name><surname>Figueroa</surname><given-names>CA</given-names></name><name><surname>Rive</surname><given-names>MM</given-names></name><name><surname>Geugies</surname><given-names>H</given-names></name><name><surname>Servaas</surname><given-names>MN</given-names></name><name><surname>Assies</surname><given-names>J</given-names></name><name><surname>Koeter</surname><given-names>MWJ</given-names></name><name><surname>Vaz</surname><given-names>FM</given-names></name><name><surname>Wichers</surname><given-names>M</given-names></name><name><surname>van Straalen</surname><given-names>JP</given-names></name><name><surname>de Raedt</surname><given-names>R</given-names></name><name><surname>Bockting</surname><given-names>CLH</given-names></name><name><surname>Harmer</surname><given-names>CJ</given-names></name><name><surname>Schene</surname><given-names>AH</given-names></name><name><surname>Ruhé</surname><given-names>HG</given-names></name></person-group><year iso-8601-date="2016">2016</year><article-title>Vulnerability for new episodes in recurrent major depressive disorder: protocol for the longitudinal DELTA-neuroimaging cohort study</article-title><source>BMJ Open</source><volume>6</volume><elocation-id>e009510</elocation-id><pub-id pub-id-type="doi">10.1136/bmjopen-2015-009510</pub-id><pub-id pub-id-type="pmid">26932139</pub-id></element-citation></ref><ref id="bib30"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Monereo-Sánchez</surname><given-names>J</given-names></name><name><surname>de Jong</surname><given-names>JJA</given-names></name><name><surname>Drenthen</surname><given-names>GS</given-names></name><name><surname>Beran</surname><given-names>M</given-names></name><name><surname>Backes</surname><given-names>WH</given-names></name><name><surname>Stehouwer</surname><given-names>CDA</given-names></name><name><surname>Schram</surname><given-names>MT</given-names></name><name><surname>Linden</surname><given-names>DEJ</given-names></name><name><surname>Jansen</surname><given-names>JFA</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Quality control strategies for brain MRI segmentation and parcellation: Practical approaches and recommendations - insights from the Maastricht study</article-title><source>NeuroImage</source><volume>237</volume><elocation-id>118174</elocation-id><pub-id pub-id-type="doi">10.1016/j.neuroimage.2021.118174</pub-id><pub-id pub-id-type="pmid">34000406</pub-id></element-citation></ref><ref id="bib31"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nesvåg</surname><given-names>R</given-names></name><name><surname>Jönsson</surname><given-names>EG</given-names></name><name><surname>Bakken</surname><given-names>IJ</given-names></name><name><surname>Knudsen</surname><given-names>GP</given-names></name><name><surname>Bjella</surname><given-names>TD</given-names></name><name><surname>Reichborn-Kjennerud</surname><given-names>T</given-names></name><name><surname>Melle</surname><given-names>I</given-names></name><name><surname>Andreassen</surname><given-names>OA</given-names></name></person-group><year iso-8601-date="2017">2017</year><article-title>The quality of severe mental disorder diagnoses in a national health registry as compared to research diagnoses based on structured interview</article-title><source>BMC Psychiatry</source><volume>17</volume><elocation-id>93</elocation-id><pub-id pub-id-type="doi">10.1186/s12888-017-1256-8</pub-id><pub-id pub-id-type="pmid">28292279</pub-id></element-citation></ref><ref id="bib32"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nordahl</surname><given-names>CW</given-names></name><name><surname>Iosif</surname><given-names>AM</given-names></name><name><surname>Young</surname><given-names>GS</given-names></name><name><surname>Hechtman</surname><given-names>A</given-names></name><name><surname>Heath</surname><given-names>B</given-names></name><name><surname>Lee</surname><given-names>JK</given-names></name><name><surname>Libero</surname><given-names>L</given-names></name><name><surname>Reinhardt</surname><given-names>VP</given-names></name><name><surname>Winder-Patel</surname><given-names>B</given-names></name><name><surname>Amaral</surname><given-names>DG</given-names></name><name><surname>Rogers</surname><given-names>S</given-names></name><name><surname>Solomon</surname><given-names>M</given-names></name><name><surname>Ozonoff</surname><given-names>S</given-names></name></person-group><year iso-8601-date="2020">2020</year><article-title>High Psychopathology Subgroup in Young Children With Autism: Associations With Biological Sex and Amygdala Volume</article-title><source>Journal of the American Academy of Child and Adolescent Psychiatry</source><volume>59</volume><fpage>1353</fpage><lpage>1363</lpage><pub-id pub-id-type="doi">10.1016/j.jaac.2019.11.022</pub-id><pub-id pub-id-type="pmid">31972262</pub-id></element-citation></ref><ref id="bib33"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Rios</surname><given-names>G</given-names></name><name><surname>Tobar</surname><given-names>F</given-names></name></person-group><year iso-8601-date="2019">2019</year><article-title>Compositionally-warped Gaussian processes</article-title><source>Neural Networks</source><volume>118</volume><fpage>235</fpage><lpage>246</lpage><pub-id pub-id-type="doi">10.1016/j.neunet.2019.06.012</pub-id><pub-id pub-id-type="pmid">31319321</pub-id></element-citation></ref><ref id="bib34"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Rosen</surname><given-names>AFG</given-names></name><name><surname>Roalf</surname><given-names>DR</given-names></name><name><surname>Ruparel</surname><given-names>K</given-names></name><name><surname>Blake</surname><given-names>J</given-names></name><name><surname>Seelaus</surname><given-names>K</given-names></name><name><surname>Villa</surname><given-names>LP</given-names></name><name><surname>Ciric</surname><given-names>R</given-names></name><name><surname>Cook</surname><given-names>PA</given-names></name><name><surname>Davatzikos</surname><given-names>C</given-names></name><name><surname>Elliott</surname><given-names>MA</given-names></name><name><surname>Garcia de La Garza</surname><given-names>A</given-names></name><name><surname>Gennatas</surname><given-names>ED</given-names></name><name><surname>Quarmley</surname><given-names>M</given-names></name><name><surname>Schmitt</surname><given-names>JE</given-names></name><name><surname>Shinohara</surname><given-names>RT</given-names></name><name><surname>Tisdall</surname><given-names>MD</given-names></name><name><surname>Craddock</surname><given-names>RC</given-names></name><name><surname>Gur</surname><given-names>RE</given-names></name><name><surname>Gur</surname><given-names>RC</given-names></name><name><surname>Satterthwaite</surname><given-names>TD</given-names></name></person-group><year iso-8601-date="2018">2018</year><article-title>Quantitative assessment of structural image quality</article-title><source>NeuroImage</source><volume>169</volume><fpage>407</fpage><lpage>418</lpage><pub-id pub-id-type="doi">10.1016/j.neuroimage.2017.12.059</pub-id><pub-id pub-id-type="pmid">29278774</pub-id></element-citation></ref><ref id="bib35"><element-citation publication-type="preprint"><person-group person-group-type="author"><name><surname>Rutherford</surname><given-names>S</given-names></name><name><surname>Angstadt</surname><given-names>M</given-names></name><name><surname>Sripada</surname><given-names>C</given-names></name><name><surname>Chang</surname><given-names>SE</given-names></name></person-group><year iso-8601-date="2020">2020</year><article-title>Leveraging Big Data for Classification of Children Who Stutter from Fluent Peers</article-title><source>bioRxiv</source><pub-id pub-id-type="doi">10.1101/2020.10.28.359711</pub-id></element-citation></ref><ref id="bib36"><element-citation publication-type="software"><person-group person-group-type="author"><name><surname>Rutherford</surname><given-names>S</given-names></name></person-group><year iso-8601-date="2022">2022a</year><data-title>braincharts</data-title><version designator="swh:1:rev:ee2b7ebcb46bab0f302f73f8d6fc913f63fccda5">swh:1:rev:ee2b7ebcb46bab0f302f73f8d6fc913f63fccda5</version><source>Software Heritage</source><ext-link ext-link-type="uri" xlink:href="https://archive.softwareheritage.org/swh:1:dir:cd28f07707409fed9c4e3bcfefaa08b9638dbccb;origin=https://github.com/predictive-clinical-neuroscience/braincharts;visit=swh:1:snp:ad9fb4794977fbb77f867431c767d66d31781b3f;anchor=swh:1:rev:ee2b7ebcb46bab0f302f73f8d6fc913f63fccda5">https://archive.softwareheritage.org/swh:1:dir:cd28f07707409fed9c4e3bcfefaa08b9638dbccb;origin=https://github.com/predictive-clinical-neuroscience/braincharts;visit=swh:1:snp:ad9fb4794977fbb77f867431c767d66d31781b3f;anchor=swh:1:rev:ee2b7ebcb46bab0f302f73f8d6fc913f63fccda5</ext-link></element-citation></ref><ref id="bib37"><element-citation publication-type="software"><person-group person-group-type="author"><name><surname>Rutherford</surname><given-names>S</given-names></name></person-group><year iso-8601-date="2022">2022b</year><data-title>brainviz-app</data-title><version designator="swh:1:rev:021fff9a48b26f2d07bbb4b3fb92cd5202418905">swh:1:rev:021fff9a48b26f2d07bbb4b3fb92cd5202418905</version><source>Software Heritage</source><ext-link ext-link-type="uri" xlink:href="https://archive.softwareheritage.org/swh:1:dir:5172634bcf0ed341052462df7780d000c5bd4f9e;origin=https://github.com/saigerutherford/brainviz-app;visit=swh:1:snp:70e2db995b8f63e9bdcb233217089d22000d2147;anchor=swh:1:rev:021fff9a48b26f2d07bbb4b3fb92cd5202418905">https://archive.softwareheritage.org/swh:1:dir:5172634bcf0ed341052462df7780d000c5bd4f9e;origin=https://github.com/saigerutherford/brainviz-app;visit=swh:1:snp:70e2db995b8f63e9bdcb233217089d22000d2147;anchor=swh:1:rev:021fff9a48b26f2d07bbb4b3fb92cd5202418905</ext-link></element-citation></ref><ref id="bib38"><element-citation publication-type="software"><person-group person-group-type="author"><name><surname>Rutherford</surname><given-names>S</given-names></name></person-group><year iso-8601-date="2022">2022c</year><data-title>lifespanqcscripts</data-title><version designator="swh:1:rev:70894691c74febe2a4d40ab0c84c50094b9e99ce">swh:1:rev:70894691c74febe2a4d40ab0c84c50094b9e99ce</version><source>Software Heritage</source><ext-link ext-link-type="uri" xlink:href="https://archive.softwareheritage.org/swh:1:dir:9c98ca93b3fb3b463607286eec7dfc9c4c3e97db;origin=https://github.com/saigerutherford/lifespan_qc_scripts;visit=swh:1:snp:84918033541e80549e91c96e85a29d191321d0a3;anchor=swh:1:rev:70894691c74febe2a4d40ab0c84c50094b9e99ce">https://archive.softwareheritage.org/swh:1:dir:9c98ca93b3fb3b463607286eec7dfc9c4c3e97db;origin=https://github.com/saigerutherford/lifespan_qc_scripts;visit=swh:1:snp:84918033541e80549e91c96e85a29d191321d0a3;anchor=swh:1:rev:70894691c74febe2a4d40ab0c84c50094b9e99ce</ext-link></element-citation></ref><ref id="bib39"><element-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Snelson</surname><given-names>E</given-names></name><name><surname>Rasmussen</surname><given-names>CE</given-names></name><name><surname>Ghahramani</surname><given-names>Z</given-names></name></person-group><year iso-8601-date="2003">2003</year><conf-name>Warped Gaussian processes</conf-name><article-title>Proceedings of the 16th International Conference on Neural Information Processing Systems</article-title><fpage>337</fpage><lpage>344</lpage></element-citation></ref><ref id="bib40"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tamnes</surname><given-names>CK</given-names></name><name><surname>Ostby</surname><given-names>Y</given-names></name><name><surname>Fjell</surname><given-names>AM</given-names></name><name><surname>Westlye</surname><given-names>LT</given-names></name><name><surname>Due-Tønnessen</surname><given-names>P</given-names></name><name><surname>Walhovd</surname><given-names>KB</given-names></name></person-group><year iso-8601-date="2010">2010</year><article-title>Brain maturation in adolescence and young adulthood: regional age-related changes in cortical thickness and white matter volume and microstructure</article-title><source>Cerebral Cortex (New York, N.Y</source><volume>20</volume><fpage>534</fpage><lpage>548</lpage><pub-id pub-id-type="doi">10.1093/cercor/bhp118</pub-id><pub-id pub-id-type="pmid">19520764</pub-id></element-citation></ref><ref id="bib41"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tomlinson</surname><given-names>RC</given-names></name><name><surname>Burt</surname><given-names>SA</given-names></name><name><surname>Waller</surname><given-names>R</given-names></name><name><surname>Jonides</surname><given-names>J</given-names></name><name><surname>Miller</surname><given-names>AL</given-names></name><name><surname>Gearhardt</surname><given-names>AN</given-names></name><name><surname>Peltier</surname><given-names>SJ</given-names></name><name><surname>Klump</surname><given-names>KL</given-names></name><name><surname>Lumeng</surname><given-names>JC</given-names></name><name><surname>Hyde</surname><given-names>LW</given-names></name></person-group><year iso-8601-date="2020">2020</year><article-title>Neighborhood poverty predicts altered neural and behavioral response inhibition</article-title><source>NeuroImage</source><volume>209</volume><elocation-id>116536</elocation-id><pub-id pub-id-type="doi">10.1016/j.neuroimage.2020.116536</pub-id><pub-id pub-id-type="pmid">31935521</pub-id></element-citation></ref><ref id="bib42"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tso</surname><given-names>IF</given-names></name><name><surname>Angstadt</surname><given-names>M</given-names></name><name><surname>Rutherford</surname><given-names>S</given-names></name><name><surname>Peltier</surname><given-names>S</given-names></name><name><surname>Diwadkar</surname><given-names>VA</given-names></name><name><surname>Taylor</surname><given-names>SF</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Dynamic causal modeling of eye gaze processing in schizophrenia</article-title><source>Schizophrenia Research</source><volume>229</volume><fpage>112</fpage><lpage>121</lpage><pub-id pub-id-type="doi">10.1016/j.schres.2020.11.012</pub-id><pub-id pub-id-type="pmid">33229223</pub-id></element-citation></ref><ref id="bib43"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Weigard</surname><given-names>AS</given-names></name><name><surname>Brislin</surname><given-names>SJ</given-names></name><name><surname>Cope</surname><given-names>LM</given-names></name><name><surname>Hardee</surname><given-names>JE</given-names></name><name><surname>Martz</surname><given-names>ME</given-names></name><name><surname>Ly</surname><given-names>A</given-names></name><name><surname>Zucker</surname><given-names>RA</given-names></name><name><surname>Sripada</surname><given-names>C</given-names></name><name><surname>Heitzeg</surname><given-names>MM</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Evidence accumulation and associated error-related brain activity as computationally-informed prospective predictors of substance use in emerging adulthood</article-title><source>Psychopharmacology</source><volume>238</volume><fpage>2629</fpage><lpage>2644</lpage><pub-id pub-id-type="doi">10.1007/s00213-021-05885-w</pub-id><pub-id pub-id-type="pmid">34173032</pub-id></element-citation></ref><ref id="bib44"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Doan</surname><given-names>NT</given-names></name><name><surname>Kaufmann</surname><given-names>T</given-names></name><name><surname>Alnæs</surname><given-names>D</given-names></name><name><surname>Moberget</surname><given-names>T</given-names></name><name><surname>Agartz</surname><given-names>I</given-names></name><name><surname>Buitelaar</surname><given-names>JK</given-names></name><name><surname>Ueland</surname><given-names>T</given-names></name><name><surname>Melle</surname><given-names>I</given-names></name><name><surname>Franke</surname><given-names>B</given-names></name><name><surname>Andreassen</surname><given-names>OA</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Westlye</surname><given-names>LT</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name></person-group><year iso-8601-date="2018">2018</year><article-title>Mapping the Heterogeneous Phenotype of Schizophrenia and Bipolar Disorder Using Normative Models</article-title><source>JAMA Psychiatry</source><volume>75</volume><fpage>1146</fpage><lpage>1155</lpage><pub-id pub-id-type="doi">10.1001/jamapsychiatry.2018.2467</pub-id><pub-id pub-id-type="pmid">30304337</pub-id></element-citation></ref><ref id="bib45"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Hoogman</surname><given-names>M</given-names></name><name><surname>Buitelaar</surname><given-names>JK</given-names></name><name><surname>Franke</surname><given-names>B</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name></person-group><year iso-8601-date="2020">2020</year><article-title>Individual differences v. the average patient: mapping the heterogeneity in ADHD using normative models</article-title><source>Psychological Medicine</source><volume>50</volume><fpage>314</fpage><lpage>323</lpage><pub-id pub-id-type="doi">10.1017/S0033291719000084</pub-id><pub-id pub-id-type="pmid">30782224</pub-id></element-citation></ref><ref id="bib46"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Rokicki</surname><given-names>J</given-names></name><name><surname>Alnaes</surname><given-names>D</given-names></name><name><surname>Berthet</surname><given-names>P</given-names></name><name><surname>Agartz</surname><given-names>I</given-names></name><name><surname>Kia</surname><given-names>SM</given-names></name><name><surname>Kaufmann</surname><given-names>T</given-names></name><name><surname>Zabihi</surname><given-names>M</given-names></name><name><surname>Moberget</surname><given-names>T</given-names></name><name><surname>Melle</surname><given-names>I</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Andreassen</surname><given-names>OA</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name><name><surname>Westlye</surname><given-names>LT</given-names></name></person-group><year iso-8601-date="2021">2021</year><article-title>Replicating extensive brain structural heterogeneity in individuals with schizophrenia and bipolar disorder</article-title><source>Human Brain Mapping</source><volume>42</volume><fpage>2546</fpage><lpage>2555</lpage><pub-id pub-id-type="doi">10.1002/hbm.25386</pub-id><pub-id pub-id-type="pmid">33638594</pub-id></element-citation></ref><ref id="bib47"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Zabihi</surname><given-names>M.</given-names></name><name><surname>Oldehinkel</surname><given-names>M</given-names></name><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Frouin</surname><given-names>V</given-names></name><name><surname>Goyard</surname><given-names>D</given-names></name><name><surname>Loth</surname><given-names>E</given-names></name><name><surname>Charman</surname><given-names>T</given-names></name><name><surname>Tillmann</surname><given-names>J</given-names></name><name><surname>Banaschewski</surname><given-names>T</given-names></name><name><surname>Dumas</surname><given-names>G</given-names></name><name><surname>Holt</surname><given-names>R</given-names></name><name><surname>Baron-Cohen</surname><given-names>S</given-names></name><name><surname>Durston</surname><given-names>S</given-names></name><name><surname>Bölte</surname><given-names>S</given-names></name><name><surname>Murphy</surname><given-names>D</given-names></name><name><surname>Ecker</surname><given-names>C</given-names></name><name><surname>Buitelaar</surname><given-names>JK</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Marquand</surname><given-names>AF</given-names></name></person-group><year iso-8601-date="2019">2019</year><article-title>Dissecting the Heterogeneous Cortical Anatomy of Autism Spectrum Disorder Using Normative Models</article-title><source>Biological Psychiatry. Cognitive Neuroscience and Neuroimaging</source><volume>4</volume><fpage>567</fpage><lpage>578</lpage><pub-id pub-id-type="doi">10.1016/j.bpsc.2018.11.013</pub-id><pub-id pub-id-type="pmid">30799285</pub-id></element-citation></ref><ref id="bib48"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Zabihi</surname><given-names>M</given-names></name><name><surname>Floris</surname><given-names>DL</given-names></name><name><surname>Kia</surname><given-names>SM</given-names></name><name><surname>Wolfers</surname><given-names>T</given-names></name><name><surname>Tillmann</surname><given-names>J</given-names></name><name><surname>Arenas</surname><given-names>AL</given-names></name><name><surname>Moessnang</surname><given-names>C</given-names></name><name><surname>Banaschewski</surname><given-names>T</given-names></name><name><surname>Holt</surname><given-names>R</given-names></name><name><surname>Baron-Cohen</surname><given-names>S</given-names></name><name><surname>Loth</surname><given-names>E</given-names></name><name><surname>Charman</surname><given-names>T</given-names></name><name><surname>Bourgeron</surname><given-names>T</given-names></name><name><surname>Murphy</surname><given-names>D</given-names></name><name><surname>Ecker</surname><given-names>C</given-names></name><name><surname>Buitelaar</surname><given-names>JK</given-names></name><name><surname>Beckmann</surname><given-names>CF</given-names></name><name><surname>Marquand</surname><given-names>A</given-names></name><collab>EU-AIMS LEAP Group</collab></person-group><year iso-8601-date="2020">2020</year><article-title>Fractionating autism based on neuroanatomical normative modeling</article-title><source>Translational Psychiatry</source><volume>10</volume><fpage>1</fpage><lpage>10</lpage><pub-id pub-id-type="doi">10.1038/s41398-020-01057-0</pub-id><pub-id pub-id-type="pmid">33159037</pub-id></element-citation></ref><ref id="bib49"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Zucker</surname><given-names>RA</given-names></name><name><surname>Ellis</surname><given-names>DA</given-names></name><name><surname>Fitzgerald</surname><given-names>HE</given-names></name><name><surname>Bingham</surname><given-names>CR</given-names></name><name><surname>Sanford</surname><given-names>K</given-names></name></person-group><year iso-8601-date="2009">2009</year><article-title>Other evidence for at least two alcoholisms II: Life course variation in antisociality and heterogeneity of alcoholic outcome</article-title><source>Development and Psychopathology</source><volume>8</volume><fpage>831</fpage><lpage>848</lpage><pub-id pub-id-type="doi">10.1017/S0954579400007458</pub-id></element-citation></ref></ref-list></back><sub-article article-type="editor-report" id="sa0"><front-stub><article-id pub-id-type="doi">10.7554/eLife.72904.sa0</article-id><title-group><article-title>Editor's evaluation</article-title></title-group><contrib-group><contrib contrib-type="author"><name><surname>Baker</surname><given-names>Chris I</given-names></name><role specific-use="editor">Reviewing Editor</role><aff><institution>National Institute of Mental Health, National Institutes of Health</institution><country>United States</country></aff></contrib></contrib-group><related-object id="sa0ro1" link-type="continued-by" object-id="10.1101/2021.08.08.455487" object-id-type="id" xlink:href="https://sciety.org/articles/activity/10.1101/2021.08.08.455487"/></front-stub><body><p>This manuscript is of broad interest to the neuroimaging community. It establishes a detailed reference model of human brain development and lifespan trajectories based on a very large data set, across many cortical and subcortical brain regions. The model not only explains substantial variability on test data, it also successfully uncovers individual differences on a database of psychiatric patients that, in addition to group-level analyses, may be critical for diagnosis, thereby demonstrating high clinical potential. It presents a clear overview of the data resource, including detailed evaluation metrics, and makes code, models and documentation directly available to the community.</p></body></sub-article><sub-article article-type="decision-letter" id="sa1"><front-stub><article-id pub-id-type="doi">10.7554/eLife.72904.sa1</article-id><title-group><article-title>Decision letter</article-title></title-group><contrib-group content-type="section"><contrib contrib-type="editor"><name><surname>Baker</surname><given-names>Chris I</given-names></name><role>Reviewing Editor</role><aff><institution>National Institute of Mental Health, National Institutes of Health</institution><country>United States</country></aff></contrib></contrib-group><contrib-group><contrib contrib-type="reviewer"><name><surname>Taschler</surname><given-names>Bernd</given-names></name><role>Reviewer</role><aff><institution>University of Oxford</institution><country>United Kingdom</country></aff></contrib><contrib contrib-type="reviewer"><name><surname>Esteban</surname><given-names>Oscar</given-names></name><role>Reviewer</role><aff><institution>Stanford University</institution><country>United States</country></aff></contrib><contrib contrib-type="reviewer"><name><surname>Constable</surname><given-names>Todd</given-names></name><role>Reviewer</role><aff><institution>Yale University</institution><country>United States</country></aff></contrib></contrib-group></front-stub><body><boxed-text id="box1"><p>Our editorial process produces two outputs: i) <ext-link ext-link-type="uri" xlink:href="https://sciety.org/articles/activity/10.1101/2021.08.08.455487">public reviews</ext-link> designed to be posted alongside <ext-link ext-link-type="uri" xlink:href="https://www.biorxiv.org/content/10.1101/2021.08.08.455487v2">the preprint</ext-link> for the benefit of readers; ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.</p></boxed-text><p><bold>Decision letter after peer review:</bold></p><p>Thank you for submitting your article "Charting Brain Growth and Aging at High Spatial Precision" for consideration by <italic>eLife</italic>. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by Chris Baker as the Senior and Reviewing Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Bernd Taschler (Reviewer #1); Oscar Esteban (Reviewer #2); Todd Constable (Reviewer #3).</p><p>The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.</p><p>As you will see, all three reviewers are very enthusiastic about this work and have some excellent suggestions for strengthening the manuscript that will require some additional analyses.</p><p>Essential revisions:</p><p>The comments from the three reviewers are highly consistent and identify three main areas where the manuscript can be improved to increase the strength of the results and the utility of the resource.</p><p>1) All reviewers noted concerns about the current evidence for generalization of the findings. The authors should include additional cross-validation tests across sites.</p><p>2) Related to point (1), the bulk of the data come from the UK Biobank. The implications and potential limitations caused by this should be more fully discussed, although the additional cross-validation analyses will help with this.</p><p>3) The manual QC is very impressive, but the whole process could be described in more detail to enable others to reproduce such a QC.</p><p><italic>Reviewer #1 (Recommendations for the authors):</italic></p><p>This is a highly valuable resource that will hopefully grow further in the future. The manuscript is well written and data and results are presented in a clear and detailed way. I especially commend the authors on making their code easy to run, understandable and truly accessible.</p><p>One aspect that, in my opinion, would strengthen the paper is the inclusion of a more comprehensive evaluation on unseen data across sites. With clinical applications in mind where a small, in-house data set is compared to the reference models, it would be useful to understand how much variation is to be expected from scanner/site differences alone. A comparison of the existing evaluation metrics with a scenario in which the models are trained on one set of sites (or even just UKB alone) and tested on a separate set of data that does not include any of the training sites would increase the interpretability of the current results.</p><p>Several recent studies have found recruitment and selection bias in the UKB with respect to the general population and even within the imaging cohort compared to the full 500k. Although briefly mentioned in the limitations, this could be expanded further by discussing recent findings.</p><p><italic>Reviewer #3 (Recommendations for the authors):</italic></p><p>While the numbers are probably sufficient that it doesn't matter – it seems that the train and test sets were only split once – and then the results presented. Proper form might be to randomly split the train/test set multiple times to obtain distributions. It would be much stronger statistically if this was repeated. If this was already repeated then it should be made clearer. The wording refers to train and test set(s) with sets being plural, but I could not find anything explicitly stating how many times this was repeated.</p><p>The data shown in Figure 1 might be better served by splitting this into multiple figures. In A it is impossible to read the y-axis. In C and D the caption states that the lines are centiles of variation but it doesn't say what centiles (for example do they match the centiles of pediatric growth charts 0.4th, 2, 9th, 25, 50th etc?) – this should be stated.</p><p>Figure 1C shows whole cortex results, while D shows subcortical. It would be nice to show data for some cortical brain regions – or even summarized for lobes instead of just whole brain.</p><p>For regions, it would be reassuring to see that the development curves for PFC for example, agree with the previous literature. Or even show that different regions have different temporal growth charts. Similarly, the work could be put in context with the work of Toga et al., Trends Neurosci, 2006 – mapping brain maturation. Or the work of Pigoni et al., Eur Neuropsychopharm, 2021 where they show (in a large sample) that cortical thickness changes in the temporal lobes can be used in classification of first episode psychosis. While the authors state that a thorough analysis of these curves is beyond the scope (and I agree) it would be helpful to have some text that confirms these curves (for healthy or diseased brains) agree with past literature.</p><p>Overall I am enthusiastic to see this work published.</p></body></sub-article><sub-article article-type="reply" id="sa2"><front-stub><article-id pub-id-type="doi">10.7554/eLife.72904.sa2</article-id><title-group><article-title>Author response</article-title></title-group></front-stub><body><disp-quote content-type="editor-comment"><p>Essential revisions:</p><p>The comments from the three reviewers are highly consistent and identify three main areas where the manuscript can be improved to increase the strength of the results and the utility of the resource.</p><p>1) All reviewers noted concerns about the current evidence for generalization of the findings. The authors should include additional cross-validation tests across sites.</p></disp-quote><p>We thank the reviewers for this suggestion and have addressed the concern regarding generalizability in several ways. First, we ran an additional 10 randomized train/test splits of the data in the full sample. These new analyses show the stability of our models, as there is very little variation in the evaluation metrics across all 10 splits. These results are visualized in <xref ref-type="fig" rid="fig3s2">Figure 3—figure supplement 2</xref>. However, the static <xref ref-type="fig" rid="fig3s2">Figure 3—figure supplement 2</xref> is challenging to read, simply because there are many brain regions fit into a single plot. Therefore, we also created an online interactive visualization tool (<ext-link ext-link-type="uri" xlink:href="https://brainviz-app.herokuapp.com/">https://brainviz-app.herokuapp.com/</ext-link>) that shows the image of the brain region and the explained variance when you hover over a point (see <xref ref-type="fig" rid="sa2fig1">Author response image 1</xref> and <xref ref-type="video" rid="video1">Video 1</xref>). This interactive visualization was created for all supplemental tables for easier exploration and interpretations and we now recommend this tool as the primary method to interrogate our findings interactively.</p><fig id="sa2fig1" position="float"><label>Author response image 1.</label><caption><title>Example of the online interactive visualizations created to help interpret the evaluation metrics.</title><p>This interactive figure was created for each evaluation metric (EV, MSLL, skew, and kurtosis) and all test sets (full controls, mQC controls, clinical, and transfer).</p></caption><graphic mime-subtype="jpeg" mimetype="image" xlink:href="elife-72904.xml.media/sa2-fig1.jpg"/></fig><p>Second, we updated and expanded the transfer data set to include 6 open datasets from OpenNeuro.org (N=546) and we provide this example dataset on our GitHub with the transfer code (<ext-link ext-link-type="uri" xlink:href="https://colab.research.google.com/github/predictive-clinical-neuroscience/braincharts/blob/master/scripts/apply_normative_models.ipynb">https://colab.research.google.com/github/predictive-clinical-neuroscience/braincharts/blob/master/scripts/apply_normative_models.ipynb</ext-link>). This simultaneously provides a more comprehensive evaluation of the performance of our model on unseen data and more comprehensive walk-through for new users applying our models to new data (sites unseen in training).</p><p>Finally, we added per-site evaluation metrics (<xref ref-type="fig" rid="fig3s3">Figure 3—figure supplement 3</xref>) to demonstrate that performance is relatively stable across sites and not driven by a single large site (i.e., UKB). As site is strongly correlated with age, these visualizations can also be used to approximate model performance at different age ranges (i.e., 9–10-year-old performance can be assessed by looking at ABCD sites evaluation metrics, and 50–80-year-old performance can be assessed by looking at UKB evaluation metrics). Moreover, we would also like to emphasize that we should not expect that all sites achieve the same performance because the sampling of the different sites is highly heterogeneous in that some sites cover a broad age range (e.g., OASIS, UKB) whereas other sites have an extremely narrow age range (e.g., ABCD).</p><disp-quote content-type="editor-comment"><p>2) Related to point (1), the bulk of the data come from the UK Biobank. The implications and potential limitations caused by this should be more fully discussed, although the additional cross-validation analyses will help with this.</p></disp-quote><p>The responses to revision 1 should help to address this, in that we show per site evaluation metrics, cross validation, and additional transfer examples. These additional analyses show that the model performance is not driven solely by the UKB sample. However, we agree with this comment and have also updated the limitation section regarding the overrepresentation of UKB and included a statement regarding the known sampling bias of UKB.</p><p>“We also identify limitations of this work. We view the word “normative” as problematic. […] By sampling both healthy population samples and case-control studies, we achieve a reasonable estimate of variation across individuals, however, downstream analyses should consider the nature of the reference cohort and whether it is appropriate for the target sample.”</p><disp-quote content-type="editor-comment"><p>3) The manual QC is very impressive, but the whole process could be described in more detail to enable others to reproduce such a QC.</p></disp-quote><p>We have added further details regarding the quality checking procedure to the methods section (following reviewer 2 comment 3a-c suggestions) and improved the clarity of directions for implementing the scripts, including an interactive link to view an example of the manual QC environment, on the QC GitHub page to enable others to reproduce our manual QC pipeline.</p><p>“Quality control (QC) is an important consideration for large samples and is an active research area (Alfaro-Almagro et al., 2018; Klapwijk et al., 2019; Rosen et al., 2018). […] We separated the evaluation metrics into full test set (relying on automated QC) and mQC test set in order to compare model performance between the two QC approaches and were pleased to notice that the evaluation metrics were nearly identical across the two methods.”</p><disp-quote content-type="editor-comment"><p>Reviewer #1 (Recommendations for the authors):</p><p>This is a highly valuable resource that will hopefully grow further in the future. The manuscript is well written and data and results are presented in a clear and detailed way. I especially commend the authors on making their code easy to run, understandable and truly accessible.</p><p>One aspect that, in my opinion, would strengthen the paper is the inclusion of a more comprehensive evaluation on unseen data across sites. With clinical applications in mind where a small, in-house data set is compared to the reference models, it would be useful to understand how much variation is to be expected from scanner/site differences alone. A comparison of the existing evaluation metrics with a scenario in which the models are trained on one set of sites (or even just UKB alone) and tested on a separate set of data that does not include any of the training sites would increase the interpretability of the current results.</p></disp-quote><p>As noted, we have addressed this concern in response to Essential revisions 1) and 2) above.</p><disp-quote content-type="editor-comment"><p>Several recent studies have found recruitment and selection bias in the UKB with respect to the general population and even within the imaging cohort compared to the full 500k. Although briefly mentioned in the limitations, this could be expanded further by discussing recent findings.</p></disp-quote><p>We have addressed this concern in response to Essential revision 2) above.</p><disp-quote content-type="editor-comment"><p>Reviewer #3 (Recommendations for the authors):</p><p>While the numbers are probably sufficient that it doesn't matter – it seems that the train and test sets were only split once – and then the results presented. Proper form might be to randomly split the train/test set multiple times to obtain distributions. It would be much stronger statistically if this was repeated. If this was already repeated then it should be made clearer. The wording refers to train and test set(s) with sets being plural, but I could not find anything explicitly stating how many times this was repeated.</p></disp-quote><p>We have addressed this comment under Essential revision 1) above. Regarding nomenclature, in our previous version the plural use of test sets previously referred to the full test set, mQC test set, transfer test set, and clinical test set (<xref ref-type="table" rid="table1">Table 1</xref>). However, we thank the reviewer for pointing out that this is open to misinterpretation and have included specific test set names when referring to them instead of just using “test sets”. We also have included a resampling of the full controls test set to address the generalizability concern. Additional details on this analysis are in response to Essential revision 1 above.</p><disp-quote content-type="editor-comment"><p>The data shown in Figure 1 might be better served by splitting this into multiple figures. In A it is impossible to read the y-axis. In C and D the caption states that the lines are centiles of variation but it doesn't say what centiles (for example do they match the centiles of pediatric growth charts 0.4th, 2, 9th, 25, 50th etc?) – this should be stated.</p></disp-quote><p>We have increased the resolution and font size in <xref ref-type="fig" rid="fig1">Figure 1</xref> and added labels to the centiles.</p><disp-quote content-type="editor-comment"><p>Figure 1C shows whole cortex results, while D shows subcortical. It would be nice to show data for some cortical brain regions – or even summarized for lobes instead of just whole brain.</p><p>For regions, it would be reassuring to see that the development curves for PFC for example, agree with the previous literature. Or even show that different regions have different temporal growth charts. Similarly, the work could be put in context with the work of Toga et al., Trends Neurosci, 2006 – mapping brain maturation. Or the work of Pigoni et al., Eur Neuropsychopharm, 2021 where they show (in a large sample) that cortical thickness changes in the temporal lobes can be used in classification of first episode psychosis. While the authors state that a thorough analysis of these curves is beyond the scope (and I agree) it would be helpful to have some text that confirms these curves (for healthy or diseased brains) agree with past literature.</p></disp-quote><p>We have included a frontal cortex ROI trajectory in <xref ref-type="fig" rid="fig1">Figure 1</xref> D and there is another cortical ROI shown (calcarine). Also on our studies GitHub(<ext-link ext-link-type="uri" xlink:href="https://github.com/predictive-clinical-neuroscience/braincharts/tree/master/metrics">https://github.com/predictive-clinical-neuroscience/braincharts/tree/master/metrics</ext-link>), we now include images of the trajectory for every ROI so that users can further explore these results if they wish to see a certain ROI that is not represented in <xref ref-type="fig" rid="fig1">Figure 1</xref> as there is not enough space to show all brain regions. We have included additional citations to prior work mapping trajectories in the Results section.</p></body></sub-article></article>