一种用于自动医疗记录生成的持续预训练大语言模型方法
A Continued Pretrained LLM Approach for Automatic Medical Note Generation
AI总结：
本文提出基于13B LLaMA2持续训练的HEAL，面向医疗对话与自动病历生成，以更低成本在PubMedQA、医疗概念识别及记录正确性和完整性上达到或超过GPT-4等模型。
摘要：
大语言模型正在革新NLP任务。然而，对于大多数专业领域而言，使用GPT-4等最先进的大语言模型往往成本高得令人却步。我们推出HEAL，这是首个基于13B LLaMA2持续训练的大语言模型，专为医疗对话构建，并在自动病历书写任务上接受评测。结果表明，HEAL在PubMedQA上以78.4%的准确率优于GPT-4和PMC-LLaMA；在生成医疗记录方面也与GPT-4表现相当。值得注意的是，HEAL在识别更多正确医疗概念上超过GPT-4和Med-PaLM 2，并在正确性与完整性方面优于人类病历员及其他可比模型。
英文摘要：
LLMs are revolutionizing NLP tasks. However, the use of the most advanced LLMs, such as GPT-4, is often prohibitively expensive for most specialized fields. We introduce HEAL, the first continuously trained 13B LLaMA2-based LLM that is purpose-built for medical conversations and measured on automated scribing. Our results demonstrate that HEAL outperforms GPT-4 and PMC-LLaMA in PubMedQA, with an accuracy of 78.4\%. It also achieves parity with GPT-4 in generating medical notes. Remarkably, HEAL surpasses GPT-4 and Med-PaLM 2 in identifying more correct medical concepts and exceeds the performance of human scribes and other comparable models in correctness and completeness.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
