双子 2.5 Pro,ChatGPT 4.1,DeepSeek V3 和Claude Opus 4 在生成标准化CMR协议方面的可靠性
Răzvan-Andrei Licu1,2,3,4, Giuseppe Muscogiuri5,6, Davide Casartelli1
1School of Medicine and Surgery, University of Milano-Bicocca, Milan, Italy.
European radiology experimental
|January 26, 2026
概括
大型语言模型 (LLM) 显示出产生心血管磁共振 (CMR) 协议的潜力. 双子座2.5 Pro显示了最高的准确性,这表明AI可以帮助标准化心脏MRI协议.
科学领域:
- 放射学和医学成像学 医学成像学
- 医疗保健中的人工智能
- 心血管诊断心血管诊断服务
背景情况:
- 人工智能 (AI) 和大型语言模型 (LLM) 是医学成像中的新兴工具.
- 心血管磁共振 (CMR) 是用于心脏评估的重要先进成像技术.
- 标准化CMR协议对于一致和准确的诊断至关重要.
研究的目的:
- 评估四个领先的LLM (Gemini 2.5 Pro,ChatGPT 4.1,DeepSeek V3,Claude Opus 4) 在生成CMR协议方面的能力.
- 将人工智能生成的协议与基于心血管磁共振学会 (SCMR) 指南的专家定义的标准进行比较.
- 评估LLMs在为各种心脏病例自动化和优化CMR协议生成方面的潜力.
主要方法:
- 一项概念验证研究,涉及140个假设的心脏病例.
- 对于每个案例,LLM生成了CMR协议.
- 人工智能生成的协议与两个经验丰富的心血管放射科医生之间的共识创建的参考标准进行了基准测试.
- 使用描述性统计数据,科恩的卡帕和弗莱斯的卡帕量化了与SCMR指南的一致性.
主要成果:
- 双子 2.5 Pro 实现了与SCMR指南 (71.5%) 的最高整体一致性.
- 在LLM中,他们与SCMR协议达成了适度的协议 (Cohen的 κ = 0.55对于Gemini 2.5 Pro).
- 对强制性CMR序列观察到实质性的协议 (Fleiss' κ = 0.690.74),而对可选序列的协议是公平的.
结论:
- 评估的LLM在生成高效,适应病理的CMR协议方面表现有前途.
- 在专家监督下,人工智能辅助的协议生成可以简化成像工作流程.
- 通过LLM的协议自动化可以提高对先进心脏诊断的访问,特别是在初级医疗保健机构.
更多相关视频
相关概念视频
01:29Reliability and Validity
13.8K
Reliability and validity are two important considerations that must be made with any type of data collection. Reliability refers to the ability to consistently produce a given result. In the context of psychological research, this would mean that any instruments or tools used to collect data do so in consistent, reproducible ways.
13.8K
01:25Distribution Reliability and Automation
503
Distribution reliability in electrical power systems is critical for ensuring an uninterrupted power supply to consumers at minimal cost. According to IEEE Standard Terms, reliability is the probability that a device will function without failure over a specified time period or amount of usage. For electric power distribution, this translates to maintaining continuous power supply and addressing customer concerns over power outages. Several indices, as defined by IEEE Standard 1366-2012, are...
503
03:00Standard Entropy Change for a Reaction
24.2K
Entropy is a state function, so the standard entropy change for a chemical reaction (ΔS°rxn) can be calculated from the difference in standard entropy between the products and the reactants.
24.2K
02:37Standard Enthalpy of Formation
49.1K
Enthalpy changes are typically tabulated for reactions in which both the reactants and products are at the same conditions. A standard state is a commonly accepted set of conditions used as a reference point for the determination of properties under other different conditions. For chemists, the IUPAC standard state refers to materials under a pressure of 1 bar and solutions at 1 M and does not specify a temperature. Many thermochemical tables list values with a standard state of 1 atm. Because...
49.1K
03:02Standard Electrode Potentials
50.1K
On comparing the reactivity of silver and lead, it is observed that the two ionic species, Ag+ (aq) and Pb2+ (aq), show a difference in their redox reactivity towards copper: the silver ion undergoes spontaneous reduction, while the lead ion does not. This relative redox activity can be easily quantified in electrochemical cells by a property called cell potential. This property is commonly known as cell voltage in electrochemistry, and it is a measure of the energy which accompanies the charge...
50.1K
03:38Measurement: Standard Units
79.2K
Every measurement provides three kinds of information: the size or magnitude of the measurement (a number), a standard of comparison for the measurement (a unit), and an indication of the uncertainty of the measurement. While the number and unit are explicitly represented when a quantity is written, the uncertainty is an aspect of the errors in the measurement results.
79.2K