Radiology-GPT: A Large Language Model for Radiology

Zhengliang Liu,Aoxiao Zhong,Yiwei Li,Longtao Yang,Chao Ju,Zihao Wu,Chong Ma,Peng Shu,Cheng Chen,Sekeun Kim,Haixing Dai,Lin Zhao,Lichao Sun,Dajiang Zhu,Jun Liu,Wei Liu,Dinggang Shen,Xiang Li,Quanzheng Li,Tianming Liu
2024-03-20
Abstract:We introduce Radiology-GPT, a large language model for radiology. Using an instruction tuning approach on an extensive dataset of radiology domain knowledge, Radiology-GPT demonstrates superior performance compared to general language models such as StableLM, Dolly and LLaMA. It exhibits significant versatility in radiological diagnosis, research, and communication. This work serves as a catalyst for future developments in clinical NLP. The successful implementation of Radiology-GPT is indicative of the potential of localizing generative large language models, specifically tailored for distinctive medical specialties, while ensuring adherence to privacy standards such as HIPAA. The prospect of developing individualized, large-scale language models that cater to specific needs of various hospitals presents a promising direction. The fusion of conversational competence and domain-specific knowledge in these models is set to foster future development in healthcare AI. A demo of Radiology-GPT is available at <a class="link-external link-https" href="https://huggingface.co/spaces/allen-eric/radiology-gpt" rel="external noopener nofollow">this https URL</a>.
Computation and Language,Artificial Intelligence
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the privacy and regulatory challenges in applying large - language models (LLMs) in the field of radiology. Although existing large commercial language models such as GPT - 4 and PaLM - 2 perform excellently in language processing capabilities, due to HIPAA regulations, privacy issues, and the need for IRB approval, the direct application of these models in the clinical environment is restricted. Specifically, these problems include: 1. **Privacy Protection**: The application of existing large - language models usually requires uploading patient data to external platforms for processing, which violates the requirements of privacy regulations such as HIPAA. 2. **Regulatory Compliance**: The strict regulatory requirements for data processing and use in the clinical environment make it difficult for general large - language models to meet these standards. 3. **Adaptability to the Professional Field**: Although existing large - language models are powerful, they are insufficient in handling professional terms and tasks in specific medical fields and cannot fully meet the needs of radiologists. To solve these problems, the paper proposes Radiology - GPT, which is a large - language model specifically designed for the field of radiology. The main goals of Radiology - GPT are: - **Local Deployment**: Through local training and deployment, ensure that patient data does not leave the hospital's servers, thereby complying with privacy regulations. - **Optimization in the Professional Field**: By performing instruction - tuning on a large amount of radiology data, enable the model to better understand and generate professional terms and content in radiology reports. - **Performance Improvement**: Compared with general - purpose language models, Radiology - GPT shows higher comprehension ability, coherence, relevance, conciseness, and clinical practicality in radiology - related tasks. Through these improvements, Radiology - GPT aims to become a powerful auxiliary tool for radiologists, helping them generate and interpret radiology reports more efficiently while ensuring data security and compliance.