{"id":"https://openalex.org/W2975035192","doi":"https://doi.org/10.1145/3368555.3384468","title":"Hidden stratification causes clinically meaningful failures in machine learning for medical imaging","display_name":"Hidden stratification causes clinically meaningful failures in machine learning for medical imaging","publication_year":2020,"publication_date":"2020-03-20","ids":{"openalex":"https://openalex.org/W2975035192","doi":"https://doi.org/10.1145/3368555.3384468","mag":"2975035192","pmid":"https://pubmed.ncbi.nlm.nih.gov/33196064"},"language":"en","primary_location":{"id":"doi:10.1145/3368555.3384468","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3368555.3384468","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM Conference on Health, Inference, and Learning","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref","datacite","pubmed"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/1909.12475","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5112563918","display_name":"Luke Oakden\u2010Rayner","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127558","display_name":"Australian Centre for Robotic Vision","ror":"https://ror.org/02zv9xv82","country_code":"AU","type":"facility","lineage":["https://openalex.org/I4210127558"]},{"id":"https://openalex.org/I5681781","display_name":"University of Adelaide","ror":"https://ror.org/00892tw58","country_code":"AU","type":"education","lineage":["https://openalex.org/I5681781"]}],"countries":["AU"],"is_corresponding":true,"raw_author_name":"Luke Oakden-Rayner","raw_affiliation_strings":["Australian Institute for Machine Learning, University of Adelaide","Australian Institute for Machine Learning University of Adelaide"],"affiliations":[{"raw_affiliation_string":"Australian Institute for Machine Learning, University of Adelaide","institution_ids":["https://openalex.org/I4210127558","https://openalex.org/I5681781"]},{"raw_affiliation_string":"Australian Institute for Machine Learning University of Adelaide","institution_ids":["https://openalex.org/I4210127558","https://openalex.org/I5681781"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049093095","display_name":"Jared Dunnmon","orcid":"https://orcid.org/0000-0002-4002-0562"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jared Dunnmon","raw_affiliation_strings":["Department of Computer Science, Stanford University"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029215323","display_name":"Gustavo Carneiro","orcid":"https://orcid.org/0000-0002-5571-6220"},"institutions":[{"id":"https://openalex.org/I4210127558","display_name":"Australian Centre for Robotic Vision","ror":"https://ror.org/02zv9xv82","country_code":"AU","type":"facility","lineage":["https://openalex.org/I4210127558"]},{"id":"https://openalex.org/I5681781","display_name":"University of Adelaide","ror":"https://ror.org/00892tw58","country_code":"AU","type":"education","lineage":["https://openalex.org/I5681781"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Gustavo Carneiro","raw_affiliation_strings":["Australian Institute for Machine Learning, University of Adelaide","Australian Institute for Machine Learning University of Adelaide"],"affiliations":[{"raw_affiliation_string":"Australian Institute for Machine Learning, University of Adelaide","institution_ids":["https://openalex.org/I4210127558","https://openalex.org/I5681781"]},{"raw_affiliation_string":"Australian Institute for Machine Learning University of Adelaide","institution_ids":["https://openalex.org/I4210127558","https://openalex.org/I5681781"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5109519525","display_name":"Christopher R\u00e9","orcid":null},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Christopher Re","raw_affiliation_strings":["Department of Computer Science, Stanford University"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Stanford University","institution_ids":["https://openalex.org/I97018004"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5112563918"],"corresponding_institution_ids":["https://openalex.org/I4210127558","https://openalex.org/I5681781"],"apc_list":null,"apc_paid":null,"fwci":2.6765,"has_fulltext":false,"cited_by_count":30,"citation_normalized_percentile":{"value":0.91996793,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":90,"max":99},"biblio":{"volume":"2020","issue":null,"first_page":"151","last_page":"159"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12422","display_name":"Radiomics and Machine Learning in Medical Imaging","score":0.9987000226974487,"subfield":{"id":"https://openalex.org/subfields/2741","display_name":"Radiology, Nuclear Medicine and Imaging"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},"topics":[{"id":"https://openalex.org/T12422","display_name":"Radiomics and Machine Learning in Medical Imaging","score":0.9987000226974487,"subfield":{"id":"https://openalex.org/subfields/2741","display_name":"Radiology, Nuclear Medicine and Imaging"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},{"id":"https://openalex.org/T10862","display_name":"AI in cancer detection","score":0.9975000023841858,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.9907000064849854,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/spurious-relationship","display_name":"Spurious relationship","score":0.823910117149353},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.7435137033462524},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.6592470407485962},{"id":"https://openalex.org/keywords/stratification","display_name":"Stratification (seeds)","score":0.6387721300125122},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6098323464393616},{"id":"https://openalex.org/keywords/population-stratification","display_name":"Population stratification","score":0.5836942195892334},{"id":"https://openalex.org/keywords/risk-stratification","display_name":"Risk stratification","score":0.5628317594528198},{"id":"https://openalex.org/keywords/population","display_name":"Population","score":0.4149583578109741},{"id":"https://openalex.org/keywords/medicine","display_name":"Medicine","score":0.22704797983169556},{"id":"https://openalex.org/keywords/biology","display_name":"Biology","score":0.08239573240280151}],"concepts":[{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.823910117149353},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.7435137033462524},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6592470407485962},{"id":"https://openalex.org/C192943249","wikidata":"https://www.wikidata.org/wiki/Q1893382","display_name":"Stratification (seeds)","level":5,"score":0.6387721300125122},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6098323464393616},{"id":"https://openalex.org/C166976648","wikidata":"https://www.wikidata.org/wiki/Q7229824","display_name":"Population stratification","level":5,"score":0.5836942195892334},{"id":"https://openalex.org/C3020404979","wikidata":"https://www.wikidata.org/wiki/Q1058438","display_name":"Risk stratification","level":2,"score":0.5628317594528198},{"id":"https://openalex.org/C2908647359","wikidata":"https://www.wikidata.org/wiki/Q2625603","display_name":"Population","level":2,"score":0.4149583578109741},{"id":"https://openalex.org/C71924100","wikidata":"https://www.wikidata.org/wiki/Q11190","display_name":"Medicine","level":0,"score":0.22704797983169556},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.08239573240280151},{"id":"https://openalex.org/C99454951","wikidata":"https://www.wikidata.org/wiki/Q932068","display_name":"Environmental health","level":1,"score":0.0},{"id":"https://openalex.org/C88548481","wikidata":"https://www.wikidata.org/wiki/Q2397491","display_name":"Seed dormancy","level":4,"score":0.0},{"id":"https://openalex.org/C135763542","wikidata":"https://www.wikidata.org/wiki/Q106016","display_name":"Genotype","level":3,"score":0.0},{"id":"https://openalex.org/C3527866","wikidata":"https://www.wikidata.org/wiki/Q162267","display_name":"Dormancy","level":3,"score":0.0},{"id":"https://openalex.org/C164705383","wikidata":"https://www.wikidata.org/wiki/Q10379","display_name":"Cardiology","level":1,"score":0.0},{"id":"https://openalex.org/C153209595","wikidata":"https://www.wikidata.org/wiki/Q501128","display_name":"Single-nucleotide polymorphism","level":4,"score":0.0},{"id":"https://openalex.org/C55493867","wikidata":"https://www.wikidata.org/wiki/Q7094","display_name":"Biochemistry","level":1,"score":0.0},{"id":"https://openalex.org/C100701293","wikidata":"https://www.wikidata.org/wiki/Q193838","display_name":"Germination","level":2,"score":0.0},{"id":"https://openalex.org/C104317684","wikidata":"https://www.wikidata.org/wiki/Q7187","display_name":"Gene","level":2,"score":0.0},{"id":"https://openalex.org/C59822182","wikidata":"https://www.wikidata.org/wiki/Q441","display_name":"Botany","level":1,"score":0.0}],"mesh":[],"locations_count":6,"locations":[{"id":"doi:10.1145/3368555.3384468","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3368555.3384468","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM Conference on Health, Inference, and Learning","raw_type":"proceedings-article"},{"id":"pmid:33196064","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/33196064","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM Conference on Health, Inference, and Learning","raw_type":null},{"id":"pmh:oai:arXiv.org:1909.12475","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1909.12475","pdf_url":"https://arxiv.org/pdf/1909.12475","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:pubmedcentral.nih.gov:7665161","is_oa":true,"landing_page_url":"https://www.ncbi.nlm.nih.gov/pmc/articles/7665161","pdf_url":null,"source":{"id":"https://openalex.org/S2764455111","display_name":"PubMed Central","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Proc ACM Conf Health Inference Learn (2020)","raw_type":"Text"},{"id":"doi:10.48550/arxiv.1909.12475","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.1909.12475","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"},{"id":"mag:2975035192","is_oa":false,"landing_page_url":null,"pdf_url":null,"source":null,"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":null}],"best_oa_location":{"id":"pmh:oai:arXiv.org:1909.12475","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1909.12475","pdf_url":"https://arxiv.org/pdf/1909.12475","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G4410736033","display_name":null,"funder_award_id":"N000141712266","funder_id":"https://openalex.org/F4320338298","funder_display_name":"Office of Naval Research Global"},{"id":"https://openalex.org/G4976659380","display_name":null,"funder_award_id":"U54EB020405","funder_id":"https://openalex.org/F4320337373","funder_display_name":"Center for Information Technology"},{"id":"https://openalex.org/G579084456","display_name":null,"funder_award_id":"FA87501720095, FA86501827865, FA86501827882","funder_id":"https://openalex.org/F4320332180","funder_display_name":"Defense Advanced Research Projects Agency"},{"id":"https://openalex.org/G899720461","display_name":null,"funder_award_id":"CCF1763315, CCF1563078, CC1937301","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320306202","display_name":"Gordon and Betty Moore Foundation","ror":"https://ror.org/006wxqw41"},{"id":"https://openalex.org/F4320307102","display_name":"Intel Corporation","ror":"https://ror.org/01ek73717"},{"id":"https://openalex.org/F4320307764","display_name":"Microsoft","ror":"https://ror.org/00d0nc645"},{"id":"https://openalex.org/F4320308737","display_name":"Facebook","ror":"https://ror.org/01zbnvs85"},{"id":"https://openalex.org/F4320309065","display_name":"Analog Devices","ror":"https://ror.org/01545pm61"},{"id":"https://openalex.org/F4320309327","display_name":"Google","ror":"https://ror.org/00njsd438"},{"id":"https://openalex.org/F4320311914","display_name":"Kurata Memorial Hitachi Science and Technology Foundation","ror":"https://ror.org/02exqgm79"},{"id":"https://openalex.org/F4320314786","display_name":"Xilinx","ror":"https://ror.org/01rb7bk56"},{"id":"https://openalex.org/F4320322589","display_name":"Taiwan Semiconductor Manufacturing Company","ror":"https://ror.org/02wx79d08"},{"id":"https://openalex.org/F4320322617","display_name":"Okawa Foundation for Information and Telecommunications","ror":"https://ror.org/01enbtr31"},{"id":"https://openalex.org/F4320327982","display_name":"NEC Corporation","ror":null},{"id":"https://openalex.org/F4320332180","display_name":"Defense Advanced Research Projects Agency","ror":"https://ror.org/02caytj08"},{"id":"https://openalex.org/F4320333053","display_name":"Intelligence Community Postdoctoral Research Fellowship Program","ror":null},{"id":"https://openalex.org/F4320337373","display_name":"Center for Information Technology","ror":"https://ror.org/03jh5a977"},{"id":"https://openalex.org/F4320338298","display_name":"Office of Naval Research Global","ror":"https://ror.org/00rk2pe57"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":36,"referenced_works":["https://openalex.org/W1970149092","https://openalex.org/W1987971958","https://openalex.org/W1994213117","https://openalex.org/W2085487226","https://openalex.org/W2106479238","https://openalex.org/W2282821441","https://openalex.org/W2557738935","https://openalex.org/W2581082771","https://openalex.org/W2611650229","https://openalex.org/W2767106145","https://openalex.org/W2769484769","https://openalex.org/W2770241596","https://openalex.org/W2776565386","https://openalex.org/W2802561155","https://openalex.org/W2803124023","https://openalex.org/W2806021783","https://openalex.org/W2806853752","https://openalex.org/W2811374795","https://openalex.org/W2896817483","https://openalex.org/W2901782819","https://openalex.org/W2901954625","https://openalex.org/W2902874468","https://openalex.org/W2905810301","https://openalex.org/W2953328958","https://openalex.org/W2956228567","https://openalex.org/W2957824372","https://openalex.org/W2958731824","https://openalex.org/W2963521553","https://openalex.org/W2963543962","https://openalex.org/W2966462242","https://openalex.org/W2969096242","https://openalex.org/W2971350322","https://openalex.org/W2986571455","https://openalex.org/W3022291425","https://openalex.org/W3101156210","https://openalex.org/W3121683006"],"related_works":["https://openalex.org/W3011721937","https://openalex.org/W2194775991","https://openalex.org/W1834627138","https://openalex.org/W3172255281","https://openalex.org/W1977086905","https://openalex.org/W2913033917","https://openalex.org/W3027100496","https://openalex.org/W3184845774","https://openalex.org/W2801765286","https://openalex.org/W2940943291","https://openalex.org/W2923727989","https://openalex.org/W2891019206","https://openalex.org/W3187231518","https://openalex.org/W3205789029","https://openalex.org/W3008151949","https://openalex.org/W2982698495","https://openalex.org/W2780963912","https://openalex.org/W3194445373","https://openalex.org/W2971269870","https://openalex.org/W3035053267"],"abstract_inverted_index":{"Machine":[0],"learning":[1,84,196],"models":[2],"for":[3,103],"medical":[4,125,199],"image":[5],"analysis":[6],"often":[7],"suffer":[8],"from":[9,64],"poor":[10],"performance":[11,29,163],"on":[12,116,122,168],"important":[13,170],"subsets":[14,143],"of":[15,30,82,99,165,178,185,193],"a":[16,31,45,71,190],"population":[17],"that":[18,61,135,157,183],"are":[19],"not":[20],"identified":[21],"during":[22],"training":[23],"or":[24,153],"testing.":[25],"For":[26],"example,":[27],"overall":[28],"cancer":[32,49],"detection":[33],"model":[34,40],"may":[35,41],"be":[36,189],"high,":[37],"but":[38,47],"the":[39,67,79,97,117,175],"still":[42],"consistently":[43],"miss":[44],"rare":[46],"aggressive":[48],"subtype.":[50],"We":[51],"refer":[52],"to":[53,90],"this":[54,93],"problem":[55],"as":[56],"<i>hidden":[57],"stratification</i>,":[58],"and":[59,108,121,156,181],"observe":[60],"it":[62,158],"results":[63],"incompletely":[65],"describing":[66],"meaningful":[68],"variation":[69],"in":[70,140,161,198],"dataset.":[72],"While":[73],"hidden":[74,105,136,186],"stratification":[75,106,137,187],"can":[76,138,159],"substantially":[77],"reduce":[78],"clinical":[80,176],"efficacy":[81],"machine":[83,195],"models,":[85],"its":[86],"effects":[87,111],"remain":[88],"difficult":[89],"measure.":[91],"In":[92],"work,":[94],"we":[95,132,173],"assess":[96],"utility":[98],"several":[100],"possible":[101],"techniques":[102],"measuring":[104],"effects,":[107],"characterize":[109],"these":[110,129],"both":[112],"via":[113],"synthetic":[114],"experiments":[115],"CIFAR-100":[118],"benchmark":[119],"dataset":[120],"multiple":[123],"real-world":[124],"imaging":[126,142],"datasets.":[127],"Using":[128],"measurement":[130],"techniques,":[131],"find":[133],"evidence":[134],"occur":[139],"unidentified":[141],"with":[144],"low":[145,147],"prevalence,":[146],"label":[148],"quality,":[149],"subtle":[150],"distinguishing":[151],"features,":[152],"spurious":[154],"correlates,":[155],"result":[160],"relative":[162],"differences":[164],"over":[166],"20%":[167],"clinically":[169],"subsets.":[171],"Finally,":[172],"discuss":[174],"implications":[177],"our":[179],"findings,":[180],"suggest":[182],"evaluation":[184],"should":[188],"critical":[191],"component":[192],"any":[194],"deployment":[197],"imaging.":[200]},"counts_by_year":[{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":3},{"year":2021,"cited_by_count":11},{"year":2020,"cited_by_count":10},{"year":2019,"cited_by_count":2}],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2025-10-10T00:00:00"}