在电子健康记录中癌症诊断分类使用大型语言模型和BioBERT:模型性能评估研究
Soheil Hashtarkhani1, Rezaur Rashid1, Christopher L Brett2
1Center for Biomedical Informatics, Department of Pediatrics, College of Medicine, University of Tennessee Health Science Center, 50 N Dunlap Street, Memphis, TN, 38103, United States, 1 9012875836.
JMIR cancer
|October 2, 2025
概括
生物BERT和GPT-4o显示出从电子健康记录中对癌症诊断进行分类的希望. 虽然BioBERT在结构化数据方面表现出色,但GPT-4o在自由文本方面表现更好,这表明AI在医疗管理和研究中的潜力.
科学领域:
- 医疗信息学 医疗信息学
- 医疗保健中的人工智能
- 自然语言处理自然语言处理.
背景情况:
- 电子健康记录 (EHR) 包含各种数据格式,需要用于预测医疗模型的高效预处理.
- 人工智能 (AI) 和自然语言处理 (NLP) 工具为自动化诊断分类提供了潜力,但需要严格评估临床可靠性.
研究的目的:
- 评估包括GPT-3.5,GPT-4o,Llama 3.2,Gemini 1.5和BioBERT在内的大型语言模型 (LLM) 的性能.
- 评估它们在从结构化 (国际疾病分类 (ICD) 代码) 和非结构化 (自由文本) EHR数据中对癌症诊断进行分类时的准确性.
主要方法:
- 从3456个患者记录中分析了762个独特的癌症诊断.
- 测试模型将诊断分类为14个预定义的类别.
- 两位瘤学专家对分类进行了验证.
主要成果:
- 在ICD代码 (84.2) 中,BioBERT获得了最高的权重宏观F1分数,并且在准确度上与GPT-4o相匹配 (90.8).
- 在自由文本诊断 (71.8 vs 61.5) 和准确性 (81.9 vs 81.6) 的权重宏观F1得分方面,GPT-4o的表现优于BioBERT.
- GPT-3.5,Gemini和Llama的整体性能较低;常见的错误包括转移,中枢神经系统瘤和模两可的术语.
结论:
- 目前的人工智能模型的性能足以用于癌症诊断分类的管理和研究目的.
- 临床应用需要标准化的文档和强有力的人类监督,以做出关键的决策.
更多相关视频
相关概念视频
Combination Therapies and Personalized Medicine
5.9K
Combining two or more treatment methods increases the life span of cancer patients while reducing damage to vital organs or tissue from the overuse of a single treatment. Combination therapy also targets different cancer-inducing pathways, thus reducing the chances of developing resistance to treatment.
The combination of the drug acetazolamide and sulforaphane is a good example of combination therapy to treat cancer. The cells in the interior of a large tumor often die due to the hypoxic and...
The combination of the drug acetazolamide and sulforaphane is a good example of combination therapy to treat cancer. The cells in the interior of a large tumor often die due to the hypoxic and...
5.9K
Cancer Survival Analysis
648
Cancer survival analysis focuses on quantifying and interpreting the time from a key starting point, such as diagnosis or the initiation of treatment, to a specific endpoint, such as remission or death. This analysis provides critical insights into treatment effectiveness and factors that influence patient outcomes, helping to shape clinical decisions and guide prognostic evaluations. A cornerstone of oncology research, survival analysis tackles the challenges of skewed, non-normally...
648


