从脑MRI报告中评估用于诊断印象生成的大型语言模型的评估结果:一个多中心基准和读者研究
Ming-Liang Wang1, Rui-Peng Zhang1, Wen-Juan Wu2
1Department of Radiology, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China.
NPJ digital medicine
|January 22, 2026
概括
DeepSeek-R1,一个大型语言模型 (LLM),从脑MRI报告中显著改进了自动辐射诊断. 它提高了放射科医生的准确性和效率,特别是结构化数据和差异诊断提示.
科学领域:
- 放射学 放射学是一门学科.
- 人工智能的人工智能
- 医疗信息学 医疗信息学
背景情况:
- 从脑MRI报告中自动化放射性诊断是复杂的,需要显著的领域专业知识.
- 现有的方法难以应对这些报告中的高度复杂性和细微信息.
研究的目的:
- 评估10个大型语言模型 (LLM) 在从脑MRI发现中生成诊断的性能.
- 评估先进的LLM,DeepSeek-R1对放射学诊断准确性和效率的影响.
主要方法:
- 在4293份脑部MRI报告中评估了10个LLM,涵盖15种疾病类别.
- 比较单一诊断与前三位差异诊断提示策略.
- 对500份报告评估了DeepSeek-R1对放射科医生的性能 (准确性和阅读时间) 的影响.
主要成果:
- 在评估的LLM中,DeepSeek-R1表现最高,特别是在结构化发现和临床信息方面.
- 排名前三的差异诊断提示实现了97.6%的患者级准确性,超过单一诊断提示 (87.1%).
- DeepSeek-R1集成提高了放射科医生的诊断准确性 (AUPRC: 0.774-0.893) 并减少了阅读时间 (61秒到53秒).
结论:
- 像DeepSeek-R1这样的高级LLM对于在脑MRI报告中有效的自动诊断印象生成至关重要.
- 优化的提示和输入策略提高了LLM的性能.
- 这一框架显示了作为一个支持工具来提高放射学工作流程效率的潜力.
相关概念视频
Higher Mental Functions of the Brain: Language
3.5K
Language is a system of communication that allows the expression of thoughts, ideas, and feelings. The brain processes language in both hemispheres.
Language formation and comprehension take place in the dominant hemisphere. The dominant hemisphere is responsible for understanding the meaning of spoken, written, or sign language, as well as the ability to communicate. For most people, the left hemisphere is the dominant one. The right hemisphere, then, gives tone and emotional context to the...
Language formation and comprehension take place in the dominant hemisphere. The dominant hemisphere is responsible for understanding the meaning of spoken, written, or sign language, as well as the ability to communicate. For most people, the left hemisphere is the dominant one. The right hemisphere, then, gives tone and emotional context to the...
3.5K
First Impression
200
First impressions play a crucial role in social perception, shaping how individuals assess others in professional, academic, and interpersonal contexts. Psychological research highlights the significance of cognitive biases, such as the primacy and recency effects, which influence how people interpret and recall information.The Primacy Effect and Cognitive AnchoringThe primacy effect describes the tendency for initial information to impact judgment disproportionately. When individuals encounter...
200
Imaging Studies I: CT and MRI
838
Introduction: MRI and CT scans are crucial advancements in medical imaging techniques, playing a vital role in diagnosing conditions related to the gastrointestinal (GI) system. Each scan serves distinct purposes, targets specific areas, and requires unique nursing duties.
Description of the Procedures
Computed Tomography (CT) scan:
Computed Tomography (CT) scans use X-ray technology to generate detailed images of bones, organs, and tissues. During the scan, the patient lies on a moving table...
Description of the Procedures
Computed Tomography (CT) scan:
Computed Tomography (CT) scans use X-ray technology to generate detailed images of bones, organs, and tissues. During the scan, the patient lies on a moving table...
838
Language
898
Language is a unique communication system that uses words and systematic rules to organize and transmit information. Unlike other forms of communication, which may involve postures, movements, odors, or vocalizations, language relies on symbols and grammar. This makes human communication distinct from that of other species, who also communicate but do not use language in the same way humans do.
Corballis and Suddendorf (2007) and Tomasello and Rakoczy (2003) highlight the role of language in...
Corballis and Suddendorf (2007) and Tomasello and Rakoczy (2003) highlight the role of language in...
898
Managing Impressions
165
Impression management encompasses individuals' deliberate efforts to shape how others perceive them during social interactions. This behavior is often employed to conform to social norms, secure approval, or pursue specific goals. While it involves selective self-presentation, it is not necessarily deceptive; individuals frequently present authentic aspects of themselves that align with situational demands.Common strategies include:Ingratiation: where individuals use flattery or agreeableness...
165
Appendicitis-II: Diagnostic Studies and Management
541
Diagnosing and managing appendicitis requires a structured and comprehensive approach that spans from initial assessment to postoperative care. Here is an overview of the process:
Diagnosing Appendicitis
It requires a multifaceted approach, starting with a detailed physical examination to pinpoint the location and nature of the pain and identify any associated symptoms. Laboratory tests play a crucial role. A complete Blood Count (CBC) typically reveals leukocytosis (an increased number of...
Diagnosing Appendicitis
It requires a multifaceted approach, starting with a detailed physical examination to pinpoint the location and nature of the pain and identify any associated symptoms. Laboratory tests play a crucial role. A complete Blood Count (CBC) typically reveals leukocytosis (an increased number of...
541


