Summary of Medsage: Enhancing Robustness Of Medical Dialogue Summarization to Asr Errors with Llm-generated Synthetic Dialogues, by Kuluhan Binici et al.
MEDSAGE: Enhancing Robustness of Medical Dialogue Summarization to ASR Errors with LLM-generated Synthetic Dialogues
by Kuluhan Binici, Abhinav Ramesh Kashyap, Viktor Schlegel, Andy T. Liu, Vijay Prakash Dwivedi, Thanh-Tung Nguyen, Xiaoxue Gao, Nancy F. Chen, Stefan Winkler
First submitted to arxiv on: 26 Aug 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed MEDSAGE approach aims to improve the performance of medical dialogue summarization systems by generating synthetic samples for data augmentation using Large Language Models (LLMs). The method leverages LLMs’ in-context learning capabilities to generate ASR-like errors based on a few available medical dialogue examples with audio recordings. By incorporating this noisy data into the training process, experimental results show that MEDSAGE can significantly improve the robustness and accuracy of medical dialogue summarization systems. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Medical dialogue summarization is an important task that helps healthcare professionals quickly understand patient information. However, current Automatic Speech Recognition (ASR) systems are not accurate enough to provide reliable transcripts. This makes it difficult to fine-tune models for this specific domain. To address this challenge, researchers developed a new approach called MEDSAGE. It uses Large Language Models to generate synthetic samples that mimic the errors made by ASR systems. These noisy data samples are then used to train medical dialogue summarization models. The results show that MEDSAGE improves the accuracy and robustness of these models. |
Keywords
» Artificial intelligence » Data augmentation » Summarization
