000292112 001__ 292112
000292112 005__ 20250328112420.0
000292112 0247_ $$2doi$$a10.1186/s13244-024-01781-x
000292112 0247_ $$2pmid$$apmid:39112910
000292112 037__ $$aDKFZ-2024-01611
000292112 041__ $$aEnglish
000292112 082__ $$a610
000292112 1001_ $$00000-0002-4133-1194$$aD Almeida, Silvia$$b0$$eFirst author
000292112 245__ $$aHow do deep-learning models generalize across populations? Cross-ethnicity generalization of COPD detection.
000292112 260__ $$aHeidelberg$$bSpringer$$c2024
000292112 3367_ $$2DRIVER$$aarticle
000292112 3367_ $$2DataCite$$aOutput Types/Journal article
000292112 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1723123062_32133
000292112 3367_ $$2BibTeX$$aARTICLE
000292112 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000292112 3367_ $$00$$2EndNote$$aJournal Article
000292112 500__ $$a#EA:E230#LA:E230#
000292112 520__ $$aTo evaluate the performance and potential biases of deep-learning models in detecting chronic obstructive pulmonary disease (COPD) on chest CT scans across different ethnic groups, specifically non-Hispanic White (NHW) and African American (AA) populations.Inspiratory chest CT and clinical data from 7549 Genetic epidemiology of COPD individuals (mean age 62 years old, 56-69 interquartile range), including 5240 NHW and 2309 AA individuals, were retrospectively analyzed. Several factors influencing COPD binary classification performance on different ethnic populations were examined: (1) effects of training population: NHW-only, AA-only, balanced set (half NHW, half AA) and the entire set (NHW + AA all); (2) learning strategy: three supervised learning (SL) vs. three self-supervised learning (SSL) methods. Distribution shifts across ethnicity were further assessed for the top-performing methods.The learning strategy significantly influenced model performance, with SSL methods achieving higher performances compared to SL methods (p < 0.001), across all training configurations. Training on balanced datasets containing NHW and AA individuals resulted in improved model performance compared to population-specific datasets. Distribution shifts were found between ethnicities for the same health status, particularly when models were trained on nearest-neighbor contrastive SSL. Training on a balanced dataset resulted in fewer distribution shifts across ethnicity and health status, highlighting its efficacy in reducing biases.Our findings demonstrate that utilizing SSL methods and training on large and balanced datasets can enhance COPD detection model performance and reduce biases across diverse ethnic populations. These findings emphasize the importance of equitable AI-driven healthcare solutions for COPD diagnosis.Self-supervised learning coupled with balanced datasets significantly improves COPD detection model performance, addressing biases across diverse ethnic populations and emphasizing the crucial role of equitable AI-driven healthcare solutions.Self-supervised learning methods outperform supervised learning methods, showing higher AUC values (p < 0.001). Balanced datasets with non-Hispanic White and African American individuals improve model performance. Training on diverse datasets enhances COPD detection accuracy. Ethnically diverse datasets reduce bias in COPD detection models. SimCLR models mitigate biases in COPD detection across ethnicities.
000292112 536__ $$0G:(DE-HGF)POF4-315$$a315 - Bildgebung und Radioonkologie (POF4-315)$$cPOF4-315$$fPOF IV$$x0
000292112 588__ $$aDataset connected to CrossRef, PubMed, , Journals: inrepo02.dkfz.de
000292112 650_7 $$2Other$$aArtificial intelligence
000292112 650_7 $$2Other$$aChronic obstructive pulmonary disease
000292112 650_7 $$2Other$$aComputed tomography
000292112 650_7 $$2Other$$aDeep learning
000292112 650_7 $$2Other$$aEthnicity
000292112 7001_ $$0P:(DE-He78)a70f21a2bf78bbc1306c3d432ae08dc7$$aNorajitra, Tobias$$b1$$udkfz
000292112 7001_ $$0P:(DE-He78)6a78e3a44a8038881d941fb467eb4e19$$aLüth, Carsten T$$b2$$udkfz
000292112 7001_ $$0P:(DE-He78)4412d586f86ca57943732a2b9318c44f$$aWald, Tassilo$$b3$$udkfz
000292112 7001_ $$0P:(DE-He78)7dc85735e114a4ace658ba1450a2cca6$$aWeru, Vivienn$$b4$$udkfz
000292112 7001_ $$0P:(DE-He78)a657bf15b4cbdf70baed30e14c19d9d3$$aNolden, Marco$$b5$$udkfz
000292112 7001_ $$0P:(DE-He78)04a0b5a49db132d8f00cee326cb743ca$$aJäger, Paul F$$b6$$udkfz
000292112 7001_ $$avon Stackelberg, Oyunbileg$$b7
000292112 7001_ $$aHeußel, Claus Peter$$b8
000292112 7001_ $$aWeinheimer, Oliver$$b9
000292112 7001_ $$aBiederer, Jürgen$$b10
000292112 7001_ $$aKauczor, Hans-Ulrich$$b11
000292112 7001_ $$0P:(DE-He78)33c74005e1ce56f7025c4f6be15321b3$$aMaier-Hein, Klaus$$b12$$eLast author$$udkfz
000292112 773__ $$0PERI:(DE-600)2543323-4$$a10.1186/s13244-024-01781-x$$gVol. 15, no. 1, p. 198$$n1$$p198$$tInsights into imaging$$v15$$x1869-4101$$y2024
000292112 8564_ $$uhttps://inrepo02.dkfz.de/record/292112/files/s13244-024-01781-x.pdf
000292112 8564_ $$uhttps://inrepo02.dkfz.de/record/292112/files/s13244-024-01781-x.pdf?subformat=pdfa$$xpdfa
000292112 8767_ $$8SN-2024-01444-b$$92024-12-05$$d2025-03-27$$eAPC$$jZahlung erfolgt
000292112 909CO $$ooai:inrepo02.dkfz.de:292112$$pVDB$$pOpenAPC$$popenCost
000292112 9101_ $$0I:(DE-588b)2036810-0$$60000-0002-4133-1194$$aDeutsches Krebsforschungszentrum$$b0$$kDKFZ
000292112 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)a70f21a2bf78bbc1306c3d432ae08dc7$$aDeutsches Krebsforschungszentrum$$b1$$kDKFZ
000292112 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)6a78e3a44a8038881d941fb467eb4e19$$aDeutsches Krebsforschungszentrum$$b2$$kDKFZ
000292112 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)4412d586f86ca57943732a2b9318c44f$$aDeutsches Krebsforschungszentrum$$b3$$kDKFZ
000292112 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)7dc85735e114a4ace658ba1450a2cca6$$aDeutsches Krebsforschungszentrum$$b4$$kDKFZ
000292112 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)a657bf15b4cbdf70baed30e14c19d9d3$$aDeutsches Krebsforschungszentrum$$b5$$kDKFZ
000292112 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)04a0b5a49db132d8f00cee326cb743ca$$aDeutsches Krebsforschungszentrum$$b6$$kDKFZ
000292112 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)33c74005e1ce56f7025c4f6be15321b3$$aDeutsches Krebsforschungszentrum$$b12$$kDKFZ
000292112 9131_ $$0G:(DE-HGF)POF4-315$$1G:(DE-HGF)POF4-310$$2G:(DE-HGF)POF4-300$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$aDE-HGF$$bGesundheit$$lKrebsforschung$$vBildgebung und Radioonkologie$$x0
000292112 9141_ $$y2024
000292112 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bINSIGHTS IMAGING : 2022$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0320$$2StatID$$aDBCoverage$$bPubMed Central$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0501$$2StatID$$aDBCoverage$$bDOAJ Seal$$d2023-05-02T09:07:41Z
000292112 915__ $$0StatID:(DE-HGF)0500$$2StatID$$aDBCoverage$$bDOAJ$$d2023-05-02T09:07:41Z
000292112 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bDOAJ : Double anonymous peer review$$d2023-05-02T09:07:41Z
000292112 915__ $$0LIC:(DE-HGF)CCBYNV$$2V:(DE-HGF)$$aCreative Commons Attribution CC BY (No Version)$$bDOAJ$$d2023-05-02T09:07:41Z
000292112 915__ $$0StatID:(DE-HGF)0600$$2StatID$$aDBCoverage$$bEbsco Academic Search$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bASC$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0113$$2StatID$$aWoS$$bScience Citation Index Expanded$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0160$$2StatID$$aDBCoverage$$bEssential Science Indicators$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)1110$$2StatID$$aDBCoverage$$bCurrent Contents - Clinical Medicine$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)9900$$2StatID$$aIF < 5$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0561$$2StatID$$aArticle Processing Charges$$d2023-08-22
000292112 915__ $$0StatID:(DE-HGF)0700$$2StatID$$aFees$$d2023-08-22
000292112 915pc $$0PC:(DE-HGF)0000$$2APC$$aAPC keys set
000292112 915pc $$0PC:(DE-HGF)0001$$2APC$$aLocal Funding
000292112 915pc $$0PC:(DE-HGF)0002$$2APC$$aDFG OA Publikationskosten
000292112 915pc $$0PC:(DE-HGF)0003$$2APC$$aDOAJ Journal
000292112 9202_ $$0I:(DE-He78)E230-20160331$$kE230$$lE230 Medizinische Bildverarbeitung$$x0
000292112 9201_ $$0I:(DE-He78)E230-20160331$$kE230$$lE230 Medizinische Bildverarbeitung$$x0
000292112 9201_ $$0I:(DE-He78)E290-20160331$$kE290$$lNWG Interaktives maschinelles Lernen$$x1
000292112 9201_ $$0I:(DE-He78)C060-20160331$$kC060$$lC060 Biostatistik$$x2
000292112 9200_ $$0I:(DE-He78)E230-20160331$$kE230$$lE230 Medizinische Bildverarbeitung$$x0
000292112 980__ $$ajournal
000292112 980__ $$aVDB
000292112 980__ $$aI:(DE-He78)E230-20160331
000292112 980__ $$aI:(DE-He78)E290-20160331
000292112 980__ $$aI:(DE-He78)C060-20160331
000292112 980__ $$aUNRESTRICTED
000292112 980__ $$aAPC