Medical Code Assignment with Gated Convolution and Note-Code Interaction

Medical Code Assignment with Gated Convolution and Note-Code Interaction Shaoxiong Ji †, Shirui Pan ‡, Pekka Marttinen † † Department of Computer Science, Aalto University ‡ Faculty of Information Technology, Monash University {shaoxiong.ji; pekka.marttinen}@aalto.fi shirui.pan@monash.edu Abstract Medical code assignment from clinical text is a fundamental task in clinical information system management. As medical notes are typically lengthy and the medical coding sys- tem’s code space is large, this task is a long- standing challenge. Recent work applies deep neural network models to encode the medical notes and assign medical codes to clinical doc- uments. However, these methods are still in- effective as they do not fully encode and cap- ture the lengthy and rich semantic information of medical notes nor explicitly exploit the in- teractions between the notes and codes. We propose a novel method, gated convolutional neural networks, and a note-code interaction (GatedCNN-NCI), for automatic medical code assignment to overcome these challenges. Our methods capture the rich semantic information of the lengthy clinical text for better repre- sentation by utilizing embedding injection and gated information propagation in the medical note encoding module. With a novel note-code interaction design and a graph message pass- ing mechanism, we explicitly capture the un- derlying dependency between notes and codes, enabling effective code prediction. A weight sharing scheme is further designed to decrease the number of trainable parameters. Empiri- cal experiments on real-world clinical datasets show that our proposed model outperforms state-of-the-art models in most cases, and our model size is on par with light-weighted base- lines. 1 Introduction Automatic medical code assignment is a routine healthcare task for medical information manage- ment and clinical decision support. The Interna- tional Classification of Diseases (ICD) coding sys- tem, maintained by the World Health Organization (WHO), is widely used among various coding sys- tems. Thus, the medical code assignment task is also called ICD coding. It uses all types of clini- cal notes to predict medical codes in a supervised manner with human-annotated codes (Perotte et al., 2014), which is formulated as a multi-class multi- label text classification problem in the medical do- main. While there are increasing works in the community in automatic medical code assign- ment (Prakash et al., 2017; Shi et al., 2017; Mullen- bach et al., 2018; Ji et al., 2020), this task remains challenging from the perspectives of note represen- tation and code prediction. First, medical note rep- resentation, a critical step in understanding medical notes, is formidably challenging due to the lengthy and complex semantic information in the discharge documents. There are typically thousands of tokens in a medical note due to the various diagnoses and procedures experienced by a patient. Furthermore, clinical notes also contain a vocabulary with many professional words and phrases, making it hard for a neural network model to encode and understand critical information. Second, the medical coding system has a very high and sparse dimensional la- bel space, which renders the code prediction task incredibly difficult. For example, ICD9 and ICD10 coding systems have …

Content truncated to 500 words