Contrastive learning improves critical event prediction in COVID-19 patients.

Autor:	Wanyan T; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA.; School of Informatics, Computing, and Engineering, Indiana University, Bloomington, IN, USA., Honarvar H; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Jaladanki SK; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Zang C; Department of Population Health Sciences, Weill Cornell Medicine, New York, NY, USA., Naik N; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Somani S; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA., De Freitas JK; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA.; Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Paranjpe I; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Vaid A; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Zhang J; Renmin University of China, Beijing, China., Miotto R; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Wang Z; Department of Electrical and Computer Engineering, University of Texas at Austin, Austin, TX, USA., Nadkarni GN; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA.; Division of Nephrology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, USA.; The Charles Bronfman Institute for Personalized Medicine, Icahn School of Medicine at Mount Sinai, New York, NY, USA., Zitnik M; Department of Biomedical Informatics, Harvard University, USA., Azad A; School of Informatics, Computing, and Engineering, Indiana University, Bloomington, IN, USA., Wang F; Department of Population Health Sciences, Weill Cornell Medicine, New York, NY, USA., Ding Y; Dell Medical School, University of Texas at Austin, Austin, TX, USA.; School of Informatics, University of Texas at Austin, Austin, TX, USA., Glicksberg BS; Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, USA.; Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New York, NY, USA.
Jazyk:	angličtina
Zdroj:	Patterns (New York, N.Y.) [Patterns (N Y)] 2021 Dec 10; Vol. 2 (12), pp. 100389. Date of Electronic Publication: 2021 Oct 25.
DOI:	10.1016/j.patter.2021.100389
Abstrakt:	Deep learning (DL) models typically require large-scale, balanced training data to be robust, generalizable, and effective in the context of healthcare. This has been a major issue for developing DL models for the coronavirus disease 2019 (COVID-19) pandemic, where data are highly class imbalanced. Conventional approaches in DL use cross-entropy loss (CEL), which often suffers from poor margin classification. We show that contrastive loss (CL) improves the performance of CEL, especially in imbalanced electronic health records (EHR) data for COVID-19 analyses. We use a diverse EHR dataset to predict three outcomes: mortality, intubation, and intensive care unit (ICU) transfer in hospitalized COVID-19 patients over multiple time windows. To compare the performance of CEL and CL, models are tested on the full dataset and a restricted dataset. CL models consistently outperform CEL models, with differences ranging from 0.04 to 0.15 for area under the precision and recall curve (AUPRC) and 0.05 to 0.1 for area under the receiver-operating characteristic curve (AUROC). Competing Interests: The authors declare no competing interests. (© 2021 The Authors.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu