{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/mime-multilevel-medical-embedding-of","title":"MiME: Multilevel Medical Embedding of Electronic Health Records for Predictive Healthcare","arxiv_id":"1810.09593","date":"2018-10-22","proceeding":"NeurIPS 2018 12","authors":["Edward Choi","Cao Xiao","Walter F. Stewart","Jimeng Sun"],"abstract":"Deep learning models exhibit state-of-the-art performance for many predictive\nhealthcare tasks using electronic health records (EHR) data, but these models\ntypically require training data volume that exceeds the capacity of most\nhealthcare systems. External resources such as medical ontologies are used to\nbridge the data volume constraint, but this approach is often not directly\napplicable or useful because of inconsistencies with terminology. To solve the\ndata insufficiency challenge, we leverage the inherent multilevel structure of\nEHR data and, in particular, the encoded relationships among medical codes. We\npropose Multilevel Medical Embedding (MiME) which learns the multilevel\nembedding of EHR data while jointly performing auxiliary prediction tasks that\nrely on this inherent EHR structure without the need for external labels. We\nconducted two prediction tasks, heart failure prediction and sequential disease\nprediction, where MiME outperformed baseline methods in diverse evaluation\nsettings. In particular, MiME consistently outperformed all baselines when\npredicting heart failure on datasets of different volumes, especially\ndemonstrating the greatest performance improvement (15% relative gain in PR-AUC\nover the best baseline) on the smallest dataset, demonstrating its ability to\neffectively model the multilevel structure of EHR data.","url_abs":"http://arxiv.org/abs/1810.09593v1","url_pdf":"http://arxiv.org/pdf/1810.09593v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"mime-multilevel-medical-embedding-of","repo_url":"https://github.com/mp2893/mime","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"disease-prediction","task_name":"Disease Prediction"},{"task_slug":"prediction","task_name":"Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1810.09593","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}