{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"omics_type":["Unknown"],"volume":["7(1)"],"submitter":["Ong Ly C"],"pubmed_abstract":["Healthcare datasets are becoming larger and more complex, necessitating the development of accurate and generalizable AI models for medical applications. Unstructured datasets, including medical imaging, electrocardiograms, and natural language data, are gaining attention with advancements in deep convolutional neural networks and large language models. However, estimating the generalizability of these models to new healthcare settings without extensive validation on external data remains challenging. In experiments across 13 datasets including X-rays, CTs, ECGs, clinical discharge summaries, and lung auscultation data, our results demonstrate that model performance is frequently overestimated by up to 20% on average due to shortcut learning of hidden data acquisition biases (DAB). Shortcu"],"journal":["NPJ digital medicine"],"pagination":["124"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC11094145"],"repository":["biostudies-literature"],"pubmed_title":["Shortcut learning in medical AI hinders generalization: method for estimating AI model generalization without external data."],"pmcid":["PMC11094145"],"pubmed_authors":["Unnikrishnan B","Duhamel J","Hope A","Brudno M","Patel T","Moayedi Y","Tadic T","Ross H","McIntosh C","Ong Ly C","Kandel S"],"additional_accession":[]},"is_claimable":false,"name":"Shortcut learning in medical AI hinders generalization: method for estimating AI model generalization without external data.","description":"Healthcare datasets are becoming larger and more complex, necessitating the development of accurate and generalizable AI models for medical applications. Unstructured datasets, including medical imaging, electrocardiograms, and natural language data, are gaining attention with advancements in deep convolutional neural networks and large language models. However, estimating the generalizability of these models to new healthcare settings without extensive validation on external data remains challenging. In experiments across 13 datasets including X-rays, CTs, ECGs, clinical discharge summaries, and lung auscultation data, our results demonstrate that model performance is frequently overestimated by up to 20% on average due to shortcut learning of hidden data acquisition biases (DAB). Shortcu","dates":{"release":"2024-01-01T00:00:00Z","publication":"2024 May","modification":"2026-04-08T19:53:07.896Z","creation":"2026-04-08T14:33:51.726Z"},"accession":"S-EPMC11094145","cross_references":{"pubmed":["38744921"],"doi":["10.1038/s41746-024-01118-4"]}}