Does the magic of BERT apply to medical code assignment? A quantitative study. (December 2021)

• Record Type: Journal Article
• Authors: Ji, Shaoxiong; Hölttä, Matti; Marttinen, Pekka
• Is Part Of: Computers in biology and medicine. Volume 139(2021)
• Publication Date: 2021-12
• Subjects: Medical code assignment -- Pretrained language models -- BERT -- Quantitative study; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2021.104998
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: Unsupervised pretraining is an integral part of many natural language processing systems, and transfer learning with language models has achieved remarkable results in downstream tasks. In the clinical application of medical code assignment, diagnosis and procedure codes are inferred from lengthy clinical notes such as hospital discharge summaries. However, it is not clear if pretrained models are useful for medical code prediction without further architecture engineering. This paper conducts a comprehensive quantitative analysis of various contextualized language models' performances, pretrained in different domains, for medical code assignment from clinical notes. We propose a hierarchical fine-tuning architecture to capture interactions between distant words and adopt label-wise attention to exploit label information. Contrary to current trends, we demonstrate that a carefully trained classical CNN outperforms attention-based models on a MIMIC-III subset with frequent codes. Our empirical findings suggest directions for building robust medical code assignment models. Highlights: A study on knowledge transfer via mixed-domain and task-adaptive language model. A thorough comparative study to answer the research questions. A hierarchical BERT architecture with a label attention mechanism. Vanilla CNN model with appropriate training can improve the predictive performance.