Phe2vec
- Robust phenotyping of patients from electronic health records (EHRs) at scale is a challenge in clinical informatics. Here, we introduce Phe2vec, an automated framework for disease phenotyping from EHRs based on unsupervised learning and assess its effectiveness against standard rule-based algorithms from Phenotype KnowledgeBase (PheKB). Phe2vec is based on pre-computing embeddings of medical concepts and patients' clinical history. Disease phenotypes are then derived from a seed concept and its neighbors in the embedding space. Patients are linked to a disease if their embedded representation is close to the disease phenotype. Comparing Phe2vec and PheKB cohorts head-to-head using chart review, Phe2vec performed on par or better in nine out of ten diseases. Differently from other approaches, it can scale to any condition and was validated against widely adopted expert-based standards. Phe2vec aims to optimize clinical informatics research by augmenting current frameworks to characterize patients by condition and derive reliableRobust phenotyping of patients from electronic health records (EHRs) at scale is a challenge in clinical informatics. Here, we introduce Phe2vec, an automated framework for disease phenotyping from EHRs based on unsupervised learning and assess its effectiveness against standard rule-based algorithms from Phenotype KnowledgeBase (PheKB). Phe2vec is based on pre-computing embeddings of medical concepts and patients' clinical history. Disease phenotypes are then derived from a seed concept and its neighbors in the embedding space. Patients are linked to a disease if their embedded representation is close to the disease phenotype. Comparing Phe2vec and PheKB cohorts head-to-head using chart review, Phe2vec performed on par or better in nine out of ten diseases. Differently from other approaches, it can scale to any condition and was validated against widely adopted expert-based standards. Phe2vec aims to optimize clinical informatics research by augmenting current frameworks to characterize patients by condition and derive reliable disease cohorts.…
Author details: | Jessica K. De Freitas, Kipp W. JohnsonORCiD, Eddye Golden, Girish N. Nadkarni, Joel T. Dudley, Erwin BöttingerORCiDGND, Benjamin S. GlicksbergORCiD, Riccardo MiottoORCiD |
---|---|
DOI: | https://doi.org/10.1016/j.patter.2021.100337 |
ISSN: | 2666-3899 |
Pubmed ID: | https://pubmed.ncbi.nlm.nih.gov/34553174 |
Title of parent work (English): | Patterns |
Subtitle (English): | Automated disease phenotyping based on unsupervised embeddings from electronic health records |
Publisher: | Elsevier |
Place of publishing: | Amsterdam |
Publication type: | Article |
Language: | English |
Date of first publication: | 2021/09/10 |
Publication year: | 2021 |
Release date: | 2024/01/12 |
Volume: | 2 |
Issue: | 9 |
Article number: | 100337 |
Number of pages: | 9 |
Funding institution: | Hasso Plattner Foundation; Alzheimer's Drug Discovery Foundation; National Center for Advancing Translational Sciences, National Institutes of HealthUnited States Department of Health & Human ServicesNational Institutes of Health (NIH) - USANIH National Center for Advancing Translational Sciences (NCATS) [U54 TR001433-05] |
Organizational units: | An-Institute / Hasso-Plattner-Institut für Digital Engineering gGmbH |
DDC classification: | 0 Informatik, Informationswissenschaft, allgemeine Werke / 00 Informatik, Wissen, Systeme / 004 Datenverarbeitung; Informatik |
Peer review: | Referiert |
Publishing method: | Open Access / Gold Open-Access |
DOAJ gelistet | |
License (German): | CC-BY-NC-ND - Namensnennung, nicht kommerziell, keine Bearbeitungen 4.0 International |