Managing Map Cardinality in Automatic Disease Classification Mapping: Balancing Precision, Recall and Coverage

2026-06-28

AI Summary

This study addresses the limitations of existing automated disease classification mapping methods, which are largely confined to one-to-one scenarios and struggle to simultaneously achieve high precision, recall, and coverage in complex one-to-many mappings. To overcome this challenge, the work introduces, for the first time, a blocking-and-matching framework from entity resolution into this domain: it first generates candidate match blocks using embedding-based techniques (blocking), then employs large language models to perform fine-grained matching within each block to identify all valid mappings (matching). Evaluated across multiple classification version pairs—including ICD-9-CM↔ICD-10-CM and ICD-10-AM↔ICD-11—the proposed approach significantly improves both precision and coverage while maintaining high recall, thereby effectively supporting real-world, complex mapping scenarios.

Abstract

Automatic mapping between disease classification systems, such as the International Classification of Diseases (ICD), is a challenging yet essential task for integrating health data and conducting longitudinal data analysis. Existing embedding-based methods primarily focus on \emph{one-to-one} mappings, overlooking more complex \emph{one-to-many} scenarios. The threshold-based and top-K methods offer natural extensions; however, they involve inherent trade-offs between \emph{precision}, \emph{recall} and \emph{mapping coverage} -- the proportion of source codes with at least one mapping to a target code. To address this challenge, we introduce a novel method, which is inspired by the \emph{blocking-and-matching} pipeline commonly used in \emph{entity resolution}. In particular, we first generate a block of candidate matches (\emph{blocking}) and then employ a large language model (LLM) to identify all valid mappings within each block (\emph{matching}). Empirically, we show that the proposed method achieves higher precision with comparable recall and broader coverage across multiple ICD version pairs (ICD-9-CM$\leftrightarrow$ICD-10-CM and ICD-10-AM$\leftrightarrow$ICD-11). Our source code and dataset is available at: https://tinyurl.com/46kyn7wp.

Problem

• disease classification mapping
• one-to-many mapping
• precision-recall trade-off
• mapping coverage
• ICD

Innovation

• blocking-and-matching
• large language model
• disease classification mapping
• one-to-many mapping
• mapping coverage