Readmission prediction via deep contextual embedding of clinical concepts
| Autor: | Tengfei Ma, Fei Wang, Cao Xiao, David M. Blei, Adji B. Dieng |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
Computer science
Cardiovascular Procedures Social Sciences lcsh:Medicine 02 engineering and technology computer.software_genre 0302 clinical medicine Risk Factors 0202 electrical engineering electronic engineering information engineering Medicine and Health Sciences Electronic Health Records Psychology 030212 general & internal medicine lcsh:Science Interpretability Language Multidisciplinary Cardiac Transplantation Patient Discharge Hospitals 3. Good health Separation Processes 020201 artificial intelligence & image processing Readmission risk Research Article Computer and Information Sciences Neural Networks Political Science MEDLINE Cardiology Context (language use) Surgical and Invasive Medical Procedures Public Policy Machine learning Medicare Research and Analysis Methods Patient Readmission Hospital records 03 medical and health sciences Humans Distillation Heart Failure Hospital readmission Transplantation Models Statistical business.industry Deep learning lcsh:R Cognitive Psychology Biology and Life Sciences Organ Transplantation Health Care Health Care Facilities Cognitive Science lcsh:Q Artificial intelligence business Feature learning computer Neuroscience |
| Zdroj: | PLoS ONE, Vol 13, Iss 4, p e0195024 (2018) PLoS ONE |
| ISSN: | 1932-6203 |
| Popis: | Objective Hospital readmission costs a lot of money every year. Many hospital readmissions are avoidable, and excessive hospital readmissions could also be harmful to the patients. Accurate prediction of hospital readmission can effectively help reduce the readmission risk. However, the complex relationship between readmission and potential risk factors makes readmission prediction a difficult task. The main goal of this paper is to explore deep learning models to distill such complex relationships and make accurate predictions. Materials and methods We propose CONTENT, a deep model that predicts hospital readmissions via learning interpretable patient representations by capturing both local and global contexts from patient Electronic Health Records (EHR) through a hybrid Topic Recurrent Neural Network (TopicRNN) model. The experiment was conducted using the EHR of a real world Congestive Heart Failure (CHF) cohort of 5,393 patients. Results The proposed model outperforms state-of-the-art methods in readmission prediction (e.g. 0.6103 ± 0.0130 vs. second best 0.5998 ± 0.0124 in terms of ROC-AUC). The derived patient representations were further utilized for patient phenotyping. The learned phenotypes provide more precise understanding of readmission risks. Discussion Embedding both local and global context in patient representation not only improves prediction performance, but also brings interpretable insights of understanding readmission risks for heterogeneous chronic clinical conditions. Conclusion This is the first of its kind model that integrates the power of both conventional deep neural network and the probabilistic generative models for highly interpretable deep patient representation learning. Experimental results and case studies demonstrate the improved performance and interpretability of the model. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
| Nepřihlášeným uživatelům se plný text nezobrazuje | K zobrazení výsledku je třeba se přihlásit. |