Bidirectional Representation Learning From Transformers Using Multimodal Electronic Health Record Data to Predict Depression

• Published in: IEEE journal of biomedical and health informatics Vol. 25; no. 8; pp. 3121 - 3129
• Main Authors: Meng, Yiwen, Speier, William, Ong, Michael K., Arnold, Corey W.
• Format: Journal Article
• Language: English
• Published: United States IEEE 01-08-2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Artificial intelligence; Biological system modeling; Chronic illnesses; Clinical decision support; Data models; Decision making; Decision support systems; Deep learning; Depression; Diseases; electronic health record; Electronic health records; Electronic medical records; Learning algorithms; Machine learning; Medical diagnosis; Medical diagnostic imaging; Medical screening; Mental depression; natural language processing; Prediction models; Predictive models; Representations; temporal representation and reasoning; Transformers; Vocabulary
• ISSN: 2168-2194; 2168-2208
• DOI: 10.1109/JBHI.2021.3063721

Advancements in machine learning algorithms have had a beneficial impact on representation learning, classification, and prediction models built using electronic health record (EHR) data. Effort has been put both on increasing models' overall performance as well as improving their interpretability, particularly regarding the decision-making process. In this study, we present a temporal deep learning model to perform bidirectional representation learning on EHR sequences with a transformer architecture to predict future diagnosis of depression. This model is able to aggregate five heterogenous and high-dimensional data sources from the EHR and process them in a temporal manner for chronic disease prediction at various prediction windows. We applied the current trend of pretraining and fine-tuning on EHR data to outperform the current state-of-the-art in chronic disease prediction, and to demonstrate the underlying relation between EHR codes in the sequence. The model generated the highest increases of precision-recall area under the curve (PRAUC) from 0.70 to 0.76 in depression prediction compared to the best baseline model. Furthermore, the self-attention weights in each sequence quantitatively demonstrated the inner relationship between various codes, which improved the model's interpretability. These results demonstrate the model's ability to utilize heterogeneous EHR data to predict depression while achieving high accuracy and interpretability, which may facilitate constructing clinical decision support systems in the future for chronic disease screening and early detection.