Question Answering for Complex Electronic Health Records Database using Unified Encoder-Decoder Architecture

TL;DR

This paper introduces UniQA, a unified encoder-decoder model with input masking for translating natural language questions into SQL or SPARQL queries on complex electronic health record data, significantly improving accuracy.

Contribution

The paper proposes UniQA, a novel unified architecture with input masking and auxiliary training for EHR question answering, addressing medical terminology complexity and typos.

Findings

Achieved 14.2% performance gain on MIMICSQL dataset.

Approximately 28.8% improvement on typo-ridden datasets.

Consistent results on MIMICSPARQL dataset.

Abstract

An intelligent machine that can answer human questions based on electronic health records (EHR-QA) has a great practical value, such as supporting clinical decisions, managing hospital administration, and medical chatbots. Previous table-based QA studies focusing on translating natural questions into table queries (NLQ2SQL), however, suffer from the unique nature of EHR data due to complex and specialized medical terminology, hence increased decoding difficulty. In this paper, we design UniQA, a unified encoder-decoder architecture for EHR-QA where natural language questions are converted to queries such as SQL or SPARQL. We also propose input masking (IM), a simple and effective method to cope with complex medical terms and various typos and better learn the SQL/SPARQL syntax. Combining the unified architecture with an effective auxiliary training objective, UniQA demonstrated a significant performance improvement against the previous state-of-the-art model for MIMICSQL* (14.2% gain), the most complex NLQ2SQL dataset in the EHR domain, and its typo-ridden versions (approximately 28.8% gain). In addition, we confirmed consistent results for the graph-based EHR-QA dataset, MIMICSPARQL*.