Clinical dialogue transcription error correction with self-supervision.

Nanayakkara, Gayani; Wiratunga, Nirmalie; Corsar, David; Martin, Kyle; Wijekoon, Anjana

Abstract

A clinical dialogue is a conversation between a clinician and a patient to share medical information, which is critical in clinical decision-making. The reliance on manual note-taking is highly inefficient and leads to transcription errors when digitising notes. Speech-to-text applications designed using Automatic Speech Recognition (ASR) can potentially overcome these errors using post-ASR error correction. Pre-trained language models are increasingly used in this area. However, the performance suffers from the lack of domain-specific vocabulary and the mismatch between error correction and pre-training objectives. This research explores these challenges in gastrointestinal specialism by introducing self-supervision strategies to fine-tune pre-trained language models for clinical dialogue error correction. We show that our mask-filling objective specialised for the medical domain (med-mask-filling) outperforms the best performing commercial ASR system by 10.27%.

Citation

NANAYAKKARA, G., WIRATUNGA, N., CORSAR, D., MARTIN, K. and WIJEKOON, A. 2023. Clinical dialogue transcription error correction with self-supervision. In Bramer, M. and Stahl, F. (eds.) Artificial intelligence XL: proceedings of the 43rd SGAI international conference on artificial intelligence (AI-2023), 12-14 December 2023, Cambridge, UK. Lecture notes in computer science, 14381. Cham: Springer [online], pages 33-46. Available from: https://doi.org/10.1007/978-3-031-47994-6_3

• Presentation Conference Type: Conference Paper (published)
• Conference Name: 43rd SGAI (Specialist Group on Artificial Intelligence) Artificial intelligence international conference 2023 (AI 2023)
• Start Date: Dec 12, 2023
• End Date: Dec 14, 2023
• Acceptance Date: Aug 29, 2023
• Online Publication Date: Nov 8, 2023
• Publication Date: Dec 31, 2023
• Deposit Date: Mar 19, 2024
• Publicly Available Date: Nov 9, 2024
• Publisher: Springer
• Peer Reviewed: Peer Reviewed
• Pages: 33-46
• Series Title: Lecture notes in computer science
• Series Number: 14381
• Series ISSN: 0302-9743; 1611-3349
• Book Title: Artificial intelligence XL: proceedings of the 43rd SGAI (Specialist Group on Artificial Intelligence) Artificial intelligence international conference 2023 (AI 2023)
• ISBN: 9783031479939
• DOI: https://doi.org/10.1007/978-3-031-47994-6_3
• Keywords: Automatic speech recognition; Error correction; Language models
• Public URL: https://rgu-repository.worktribe.com/output/2278561