Clinical Dialogue Transcription Error Correction using Seq2Seq Models

TL;DR

This paper introduces a seq2seq model approach for correcting transcription errors in clinical dialogues, utilizing a new dataset and domain-specific fine-tuning to improve accuracy of speech-to-text systems in healthcare settings.

Contribution

The study presents a novel seq2seq error correction method, a new clinical dialogue dataset, and demonstrates improved transcription accuracy through domain-specific fine-tuning.

Findings

BART-based model reduces word error rates in clinical ASR outputs

Introduces a new gastrointestinal clinical dialogue dataset

Shows effectiveness of domain-specific fine-tuning for error correction

Abstract

Good communication is critical to good healthcare. Clinical dialogue is a conversation between health practitioners and their patients, with the explicit goal of obtaining and sharing medical information. This information contributes to medical decision-making regarding the patient and plays a crucial role in their healthcare journey. The reliance on note taking and manual scribing processes are extremely inefficient and leads to manual transcription errors when digitizing notes. Automatic Speech Recognition (ASR) plays a significant role in speech-to-text applications, and can be directly used as a text generator in conversational applications. However, recording clinical dialogue presents a number of general and domain-specific challenges. In this paper, we present a seq2seq learning approach for ASR transcription error correction of clinical dialogues. We introduce a new Gastrointestinal Clinical Dialogue (GCD) Dataset which was gathered by healthcare professionals from a NHS Inflammatory Bowel Disease clinic and use this in a comparative study with four commercial ASR systems. Using self-supervision strategies, we fine-tune a seq2seq model on a mask-filling task using a domain-specific PubMed dataset which we have shared publicly for future research. The BART model fine-tuned for mask-filling was able to correct transcription errors and achieve lower word error rates for three out of four commercial ASR outputs.