LLM-RadJudge: Achieving Radiologist-Level Evaluation for X-Ray Report Generation

arXiv:2404.00998v1 Announce Type: new
Abstract: Evaluating generated radiology reports is crucial for the development of radiology AI, but existing metrics fail to reflect the task's clinical requirements. This study proposes a novel evaluation framework using large language models (LLMs) to compare radiology reports for assessment. We compare the performance of various LLMs and demonstrate that, when using GPT-4, our proposed metric achieves evaluation consistency close to that of radiologists. Furthermore, to reduce costs and improve accessibility, making this method practical, …

abstract arxiv assessment clinical cs.ai cs.cl development evaluation framework generated language language models large language large language models llm llms metrics novel performance radiologist radiology ray report reports requirements study type x-ray