A Unified Review of Deep Learning for Automated Medical Coding

Abstract

Automated medical coding, an essential task for healthcare operation and delivery, makes unstructured data manageable by predicting medical codes from clinical documents. Recent advances in deep learning and natural language processing have been widely applied to this task. However, deep learning-based medical coding lacks a unified view of the design of neural network architectures. This review proposes a unified framework to provide a general understanding of the building blocks of medical coding models and summarizes recent advanced models under the proposed framework. Our unified framework decomposes medical coding into four main components, i.e., encoder modules for text feature extraction, mechanisms for building deep encoder architectures, decoder modules for transforming hidden representations into medical codes, and the usage of auxiliary information. Finally, we introduce the benchmarks and real-world usage and discuss key research challenges and future directions.

ACM Computing Surveys

References in corpus (24)

• Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
• A Survey on Knowledge Graphs: Representation, Acquisition and Applications
• Longformer: The Long-Document Transformer
• BioGPT: Generative Pre-trained Transformer for Biomedical Text Generation and Mining
• ClinicalBERT: Modeling Clinical Notes and Predicting Hospital Readmission
• Human-in-the-loop Artificial Intelligence
• Natural Language Processing Advancements By Deep Learning: A Survey
• Suicidal Ideation and Mental Disorder Detection with Attentive Relation Networks
• Does the Magic of BERT Apply to Medical Code Assignment? A Quantitative Study
• Clinical Assistant Diagnosis for Electronic Medical Record Based on Convolutional Neural Network
• Hierarchical Label-wise Attention Transformer Model for Explainable ICD Coding
• Can GPT-3.5 Generate and Code Discharge Summaries?
• Experimental Evaluation and Development of a Silver-Standard for the MIMIC-III Clinical Coding Dataset
• A Systematic Literature Review of Automated ICD Coding and Classification Systems using Discharge Summaries
• Multitask Balanced and Recalibrated Network for Medical Code Prediction
• Medical code prediction with multi-view convolution and description-regularized label-dependent attention
• Ontology Enrichment from Texts: A Biomedical Dataset for Concept Discovery and Placement
• Hierarchical BERT for Medical Document Understanding
• Few-Shot Electronic Health Record Coding through Graph Contrastive Learning
• DKEC: Domain Knowledge Enhanced Multi-Label Classification for Diagnosis Prediction
• Towards Semi-Structured Automatic ICD Coding via Tree-based Contrastive Learning
• An Automatic ICD Coding Network Using Partition-Based Label Attention
• Improving Predictions of Tail-end Labels using Concatenated BioMed-Transformers for Long Medical Documents
• Contextual Semantic Embeddings for Ontology Subsumption Prediction