{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,8,12]],"date-time":"2025-08-12T21:42:13Z","timestamp":1755034933293,"version":"3.37.3"},"reference-count":40,"publisher":"Oxford University Press (OUP)","issue":"12","license":[{"start":{"date-parts":[[2023,9,5]],"date-time":"2023-09-05T00:00:00Z","timestamp":1693872000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100020022","name":"Parkinson Association of Alberta","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100020022","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100009003","name":"Hotchkiss Brain Institute","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100009003","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100015569","name":"Canadian Consortium on Neurodegeneration in Aging","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100015569","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100017658","name":"Canadian Open Neuroscience Platform","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100017658","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000038","name":"Natural Sciences and Engineering Research Council of Canada","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100000038","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001804","name":"Canada Research Chairs","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001804","id-type":"DOI","asserted-by":"publisher"}]},{"name":"River Fund at Calgary Foundation"},{"DOI":"10.13039\/501100000024","name":"Canadian Institutes for Health Research","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100000024","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Tourmaline Chair in Parkinson disease"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2023,11,17]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:sec>\n                  <jats:title>Objective<\/jats:title>\n                  <jats:p>This work investigates if deep learning (DL) models can classify originating site locations directly from magnetic resonance imaging (MRI) scans with and without correction for intensity differences.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Material and Methods<\/jats:title>\n                  <jats:p>A large database of 1880 T1-weighted MRI scans collected across 41 sites originally for Parkinson\u2019s disease (PD) classification was used to classify sites in this study. Forty-six percent of the datasets are from PD patients, while 54% are from healthy participants. After preprocessing the T1-weighted scans, 2 additional data types were generated: intensity-harmonized T1-weighted scans and log-Jacobian deformation maps resulting from nonlinear atlas registration. Corresponding DL models were trained to classify sites for each data type. Additionally, logistic regression models were used to investigate the contribution of biological (age, sex, disease status) and non-biological (scanner type) variables to the models\u2019 decision.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>A comparison of the 3 different types of data revealed that DL models trained using T1-weighted and intensity-harmonized T1-weighted scans can classify sites with an accuracy of 85%, while the model using log-Jacobian deformation maps achieved a site classification accuracy of 54%. Disease status and scanner type were found to be significant confounders.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Discussion<\/jats:title>\n                  <jats:p>Our results demonstrate that MRI scans encode relevant site-specific information that models could use as shortcuts that cannot be removed using simple intensity harmonization methods.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Conclusion<\/jats:title>\n                  <jats:p>The ability of DL models to exploit site-specific biases as shortcuts raises concerns about their reliability, generalization, and deployability in clinical settings.<\/jats:p>\n               <\/jats:sec>","DOI":"10.1093\/jamia\/ocad171","type":"journal-article","created":{"date-parts":[[2023,9,5]],"date-time":"2023-09-05T16:29:38Z","timestamp":1693931378000},"page":"1925-1933","source":"Crossref","is-referenced-by-count":12,"title":["Image-encoded biological and non-biological variables may be used as shortcuts in deep learning models trained on multisite neuroimaging data"],"prefix":"10.1093","volume":"30","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7455-3383","authenticated-orcid":false,"given":"Raissa","family":"Souza","sequence":"first","affiliation":[{"name":"Department of Radiology, Cumming School of Medicine, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Hotchkiss Brain Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Biomedical Engineering Graduate Program, University of Calgary , Calgary, AB T2N 4N1, Canada"}]},{"given":"Matthias","family":"Wilms","sequence":"additional","affiliation":[{"name":"Hotchkiss Brain Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Alberta Children\u2019s Hospital Research Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Department of Pediatrics, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Department of Community Health Sciences, University of Calgary , Calgary, AB T2N 4N1, Canada"}]},{"given":"Milton","family":"Camacho","sequence":"additional","affiliation":[{"name":"Department of Radiology, Cumming School of Medicine, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Hotchkiss Brain Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Biomedical Engineering Graduate Program, University of Calgary , Calgary, AB T2N 4N1, Canada"}]},{"given":"G Bruce","family":"Pike","sequence":"additional","affiliation":[{"name":"Department of Radiology, Cumming School of Medicine, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Hotchkiss Brain Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"}]},{"given":"Richard","family":"Camicioli","sequence":"additional","affiliation":[{"name":"Department of Medicine (Neurology), Neuroscience and Mental Health Institute, University of Alberta , Edmonton, AB T6G 2E1, Canada"}]},{"given":"Oury","family":"Monchi","sequence":"additional","affiliation":[{"name":"Hotchkiss Brain Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Department of Radiology, Radio-Oncology and Nuclear Medicine, Universit\u00e9 de Montr\u00e9al , Montr\u00e9al, QC H3C 3J7, Canada"},{"name":"Centre de Recherche, Institut Universitaire de G\u00e9riatrie de Montr\u00e9al , Montr\u00e9al, QC H3W 1W4, Canada"},{"name":"Department of Clinical Neurosciences, Cumming School of Medicine, University of Calgary , Calgary, AB T2N 4N1, Canada"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2556-3224","authenticated-orcid":false,"given":"Nils D","family":"Forkert","sequence":"additional","affiliation":[{"name":"Department of Radiology, Cumming School of Medicine, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Hotchkiss Brain Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Alberta Children\u2019s Hospital Research Institute, University of Calgary , Calgary, AB T2N 4N1, Canada"},{"name":"Department of Clinical Neurosciences, Cumming School of Medicine, University of Calgary , Calgary, AB T2N 4N1, Canada"}]}],"member":"286","published-online":{"date-parts":[[2023,9,5]]},"reference":[{"issue":"6","key":"2023111709545748500_ocad171-B1","doi-asserted-by":"crossref","first-page":"062001","DOI":"10.1088\/1741-2552\/abbff2","article-title":"Supervised machine learning tools: a tutorial for clinicians","volume":"17","author":"Lo Vercio","year":"2020","journal-title":"J Neural Eng"},{"issue":"7","key":"2023111709545748500_ocad171-B2","doi-asserted-by":"crossref","first-page":"795","DOI":"10.1016\/j.amjmed.2019.01.017","article-title":"Artificial intelligence transforms the future of health care","volume":"132","author":"Noorbakhsh-Sabet","year":"2019","journal-title":"Am J Med"},{"issue":"4","key":"2023111709545748500_ocad171-B3","doi-asserted-by":"crossref","first-page":"416","DOI":"10.1139\/gen-2020-0131","article-title":"Machine learning for precision medicine","volume":"64","author":"Maceachern","year":"2021","journal-title":"Genome"},{"issue":"1","key":"2023111709545748500_ocad171-B4","doi-asserted-by":"crossref","first-page":"35","DOI":"10.1186\/s41747-018-0061-6","article-title":"Artificial intelligence in medical imaging: threat or opportunity? Radiologists again at the forefront of innovation in medicine","volume":"2","author":"Pesapane","year":"2018","journal-title":"Eur Radiol Exp"},{"issue":"1","key":"2023111709545748500_ocad171-B5","doi-asserted-by":"crossref","first-page":"12236","DOI":"10.1038\/s41598-021-91669-3","article-title":"Influence of cardiovascular risk-factors on morphological changes of cerebral arteries in healthy adults across the life span","volume":"11","author":"Mouches","year":"2021","journal-title":"Sci Rep"},{"issue":"8","key":"2023111709545748500_ocad171-B6","doi-asserted-by":"crossref","first-page":"2554","DOI":"10.1002\/hbm.25805","article-title":"Multimodal biological brain age prediction using magnetic resonance imaging and angiography with the identification of predictive regions","volume":"43","author":"Mouches","year":"2022","journal-title":"Hum Brain Mapp"},{"key":"2023111709545748500_ocad171-B7","doi-asserted-by":"crossref","first-page":"648548","DOI":"10.3389\/fneur.2021.648548","article-title":"Utility of multi-modal MRI for differentiating of Parkinson\u2019s disease and progressive supranuclear palsy using machine learning","volume":"12","author":"Talai","year":"2021","journal-title":"Front Neurol"},{"issue":"6","key":"2023111709545748500_ocad171-B8","first-page":"061102","article-title":"Fairness-related performance and explainability effects in deep learning models for brain image analysis","volume":"9","author":"Stanley","year":"2022","journal-title":"J Med Imaging (Bellingham)"},{"issue":"3","key":"2023111709545748500_ocad171-B9","doi-asserted-by":"crossref","first-page":"827","DOI":"10.1016\/j.neuroimage.2008.09.053","article-title":"Sensitivity of voxel-based morphometry analysis to choice of imaging protocol at 3 T","volume":"44","author":"Tardif","year":"2009","journal-title":"Neuroimage"},{"issue":"7","key":"2023111709545748500_ocad171-B10","doi-asserted-by":"crossref","first-page":"943","DOI":"10.1002\/hbm.20908","article-title":"Regional impact of field strength on voxel-based morphometry results","volume":"31","author":"Tardif","year":"2010","journal-title":"Hum Brain Mapp"},{"key":"2023111709545748500_ocad171-B11","doi-asserted-by":"crossref","first-page":"369","DOI":"10.1007\/978-3-030-59713-9_36","volume-title":"Medical Image Computing and Computer Assisted Intervention \u2013 MICCAI 2020","author":"Dinsdale","year":"2020"},{"key":"2023111709545748500_ocad171-B12","doi-asserted-by":"crossref","first-page":"193","DOI":"10.1007\/978-3-030-05831-9_16","article-title":"Inter-scanner harmonization of high angular resolution DW-MRI using null space deep learning","volume":"2019","author":"Nath","year":"2019","journal-title":"Comput Diffus MRI"},{"year":"2019","author":"Glocker","key":"2023111709545748500_ocad171-B13"},{"issue":"11","key":"2023111709545748500_ocad171-B14","doi-asserted-by":"crossref","first-page":"665","DOI":"10.1038\/s42256-020-00257-z","article-title":"Shortcut learning in deep neural networks","volume":"2","author":"Geirhos","year":"2020","journal-title":"Nat Mach Intell"},{"issue":"3","key":"2023111709545748500_ocad171-B15","doi-asserted-by":"crossref","first-page":"e1001779","DOI":"10.1371\/journal.pmed.1001779","article-title":"UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age","volume":"12","author":"Sudlow","year":"2015","journal-title":"PLoS Med"},{"key":"2023111709545748500_ocad171-B16"},{"issue":"2","key":"2023111709545748500_ocad171-B17","doi-asserted-by":"crossref","first-page":"456","DOI":"10.1002\/jmri.26197","article-title":"The Canadian dementia imaging protocol: harmonizing national cohorts","volume":"49","author":"Duchesne","year":"2019","journal-title":"J Magn Reson Imaging"},{"issue":"1","key":"2023111709545748500_ocad171-B18","doi-asserted-by":"crossref","first-page":"56","DOI":"10.1017\/S0317167100005795","article-title":"Axial signs and magnetic resonance imaging correlates in Parkinson\u2019s disease","volume":"34","author":"Acharya","year":"2007","journal-title":"Can J Neurol Sci"},{"issue":"6","key":"2023111709545748500_ocad171-B19","doi-asserted-by":"crossref","first-page":"893","DOI":"10.1002\/mds.27674","article-title":"Network basis of the dysexecutive and posterior cortical cognitive profiles in Parkinson\u2019s disease","volume":"34","author":"Lang","year":"2019","journal-title":"Mov Disord"},{"issue":"Pt 4","key":"2023111709545748500_ocad171-B20","doi-asserted-by":"crossref","first-page":"1120","DOI":"10.1093\/brain\/awu036","article-title":"Mild cognitive impairment is linked with faster rate of cortical thinning in patients with Parkinson\u2019s disease longitudinally","volume":"137","author":"Hanganu","year":"2014","journal-title":"Brain"},{"key":"2023111709545748500_ocad171-B21"},{"issue":"11","key":"2023111709545748500_ocad171-B22","doi-asserted-by":"crossref","first-page":"e0188196","DOI":"10.1371\/journal.pone.0188196","article-title":"Exploring the reproducibility of functional connectivity alterations in Parkinson\u2019s disease","volume":"12","author":"Badea","year":"2017","journal-title":"PLoS One"},{"key":"2023111709545748500_ocad171-B23"},{"issue":"3","key":"2023111709545748500_ocad171-B24","doi-asserted-by":"crossref","first-page":"421","DOI":"10.1002\/mds.24926","article-title":"Brain iron deposition fingerprints in Parkinson\u2019s disease and progressive supranuclear palsy","volume":"27","author":"Boelmans","year":"2012","journal-title":"Mov Disord"},{"year":"2019","author":"LaMontagne","key":"2023111709545748500_ocad171-B25"},{"key":"2023111709545748500_ocad171-B26","doi-asserted-by":"crossref","first-page":"180134","DOI":"10.1038\/sdata.2018.134","article-title":"Structural and functional brain scans from the cross-sectional Southwest University adult lifespan dataset","volume":"5","author":"Wei","year":"2018","journal-title":"Sci Data"},{"issue":"17","key":"2023111709545748500_ocad171-B27","doi-asserted-by":"crossref","first-page":"4952","DOI":"10.1002\/hbm.24750","article-title":"Automated brain extraction of multisequence MRI using artificial neural networks","volume":"40","author":"Isensee","year":"2019","journal-title":"Hum Brain Mapp"},{"issue":"6","key":"2023111709545748500_ocad171-B28","doi-asserted-by":"crossref","first-page":"1310","DOI":"10.1109\/TMI.2010.2046908","article-title":"N4ITK: improved N3 bias correction","volume":"29","author":"Tustison","year":"2010","journal-title":"IEEE Trans Med Imaging"},{"key":"2023111709545748500_ocad171-B29","doi-asserted-by":"crossref","first-page":"370","DOI":"10.1016\/j.dib.2017.04.013","article-title":"A dataset of multi-contrast population-averaged brain MRI atlases of a Parkinson\u2019s disease cohort","volume":"12","author":"Xiao","year":"2017","journal-title":"Data Brief"},{"issue":"6","key":"2023111709545748500_ocad171-B30","doi-asserted-by":"crossref","first-page":"822","DOI":"10.1109\/TMI.2007.892646","article-title":"Statrationistical properties of Jacobian maps and the realization of unbiased large-deformation nonlinear image regist","volume":"26","author":"Leow","year":"2007","journal-title":"IEEE Trans Med Imaging"},{"issue":"6","key":"2023111709545748500_ocad171-B31","doi-asserted-by":"crossref","first-page":"1072","DOI":"10.1002\/(SICI)1522-2594(199912)42:6<1072::AID-MRM11>3.0.CO;2-M","article-title":"On standardizing the MR image intensity scale","volume":"42","author":"Ny\u00fal","year":"1999","journal-title":"Magn Reson Med"},{"issue":"3","key":"2023111709545748500_ocad171-B32","doi-asserted-by":"crossref","first-page":"908","DOI":"10.1002\/jmri.27908","article-title":"Deep generative medical image harmonization for improving cross-site generalization in deep learning predictors","volume":"55","author":"Bashyam","year":"2022","journal-title":"J Magn Reson Imaging"},{"key":"2023111709545748500_ocad171-B33","doi-asserted-by":"crossref","first-page":"101871","DOI":"10.1016\/j.media.2020.101871","article-title":"Accurate brain age prediction with lightweight deep neural networks","volume":"68","author":"Peng","year":"2021","journal-title":"Med Image Anal"},{"year":"2017","author":"Smilkov","key":"2023111709545748500_ocad171-B34"},{"author":"Kubota","key":"2023111709545748500_ocad171-B35"},{"author":"Nielson","key":"2023111709545748500_ocad171-B36"},{"key":"2023111709545748500_ocad171-B37","doi-asserted-by":"crossref","first-page":"160","DOI":"10.1016\/j.mri.2019.05.041","article-title":"DeepHarmony: a deep learning approach to contrast harmonization across scanner changes","volume":"64","author":"Dewey","year":"2019","journal-title":"Magn Reson Imaging"},{"key":"2023111709545748500_ocad171-B38","first-page":"38","article-title":"Lesion-preserving unpaired image-to-image translation between MRI and CT from ischemic stroke patients","author":"Gutierrez","year":"2022"},{"key":"2023111709545748500_ocad171-B39","first-page":"1","article-title":"Neural stain-style transfer learning using GAN for histopathological images","author":"Cho","year":"2017"},{"issue":"4","key":"2023111709545748500_ocad171-B40","doi-asserted-by":"crossref","first-page":"e0284440","DOI":"10.1371\/journal.pone.0284440","article-title":"Effect of geometric distortion correction on thickness and volume measurements of cortical parcellations in 3D T1w gradient echo sequences","volume":"18","author":"Thaler","year":"2023","journal-title":"PLoS One"}],"container-title":["Journal of the American Medical Informatics Association"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/jamia\/article-pdf\/30\/12\/1925\/53477565\/ocad171.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/jamia\/article-pdf\/30\/12\/1925\/53477565\/ocad171.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,11,17]],"date-time":"2023-11-17T13:28:10Z","timestamp":1700227690000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/jamia\/article\/30\/12\/1925\/7260634"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,9,5]]},"references-count":40,"journal-issue":{"issue":"12","published-online":{"date-parts":[[2023,9,5]]},"published-print":{"date-parts":[[2023,11,17]]}},"URL":"https:\/\/doi.org\/10.1093\/jamia\/ocad171","relation":{},"ISSN":["1067-5027","1527-974X"],"issn-type":[{"type":"print","value":"1067-5027"},{"type":"electronic","value":"1527-974X"}],"subject":[],"published-other":{"date-parts":[[2023,12,1]]},"published":{"date-parts":[[2023,9,5]]}}}