相关实验视频

Updated: Sep 17, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

使用大型语言模型进行FDA批准的人工智能部署后监测:肺栓塞检测使用案例

Vera Sorin¹, Panagiotis Korfiatis¹, Alex K Bratt²

¹Department of Radiology, Mayo Clinic College of Medicine and Science, Mayo Clinic, Rochester, Minnesota.

Journal of the American College of Radiology : JACR

|July 2, 2025

概括

本研究引入了使用大型语言模型 (LLM) 和人类监督来跟踪AI在临床环境中的表现的部署后监测 (PDM) 框架. 该框架有效地检测人工智能工具的性能偏移,确保放射学人工智能的持续质量保证.

科学领域:

临床信息学临床信息学
人工智能在医学中的应用
放射学工作流的优化工作流的优化

背景情况:

人工智能 (AI) 整合到临床工作流程中需要强大的部署后监控 (PDM) 来解决性能差异.
确保临床生产中部署的AI工具的持续质量保证仍然是一个重大挑战.

研究的目的:

开发和评估PDM框架,利用大型语言模型 (LLM) 进行放射学报告的自由文本分类,并纳入人类监督.
证明该框架在监测商业肺栓塞 (PE) 检测AI (CVPED) 中的应用.

主要方法:

对11999个CT肺血管造影研究的回顾性分析.
通过将放射学报告的LLM分类与CVPED输出和人类对差异的审查相结合来确定地面真相.
模拟每日监测框架以追踪CVPED-LLM差异,根据超过95%置信区间7天的情况定义漂移.

主要成果:

在11,999项研究中,1,285 (10.7%) 患有PE,其中373 (3.1%) 显示CVPED和LLM之间的分类不一致.
模型特异性下降2%至3%会使差异增加2至3倍;灵敏度下降10%也会产生类似的效果.
该框架确定了增量PE病例的4个真实警报,漂移检测的最佳后期评估期为2个月.

结论:

集成基于LLM的自由文本分类和一个循环中的人警报系统的PDM框架可以持续监控AI性能.
该框架有效地提醒性能漂移,并在检测肺栓塞等疾病时提供增量临床价值.

关键词:

在这里,我们可以看到AIAIAI.法学士 (LLM) 是一个专业.部署后的监控监控是部署后的监控.制作后的监控和监控.一个ppm的时间.质量保证质量保证质量保证

更多相关视频

Asthma Detection Research Based on Voice Signal Processing and Machine Learning

Asthma Detection Research Based on Voice Signal Processing and Machine Learning

Published on: July 22, 2025

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems

Published on: June 13, 2025

相关概念视频

Pulmonary Embolism II: Diagnostic Studies and Interprofessional Care

Pulmonary Embolism II: Diagnostic Studies and Interprofessional Care

Diagnosing Pulmonary EmbolismDiagnosing pulmonary embolism (PE) involves clinical assessment and advanced imaging tests. The preferred diagnostic tool is the spiral (helical) CT scan or CT angiography (CTA), which uses intravenous contrast media to visualize the pulmonary vasculature and identify emboli.A ventilation-perfusion (V/Q) scan is an alternative for patients unable to receive contrast media. This scan includes both perfusion and ventilation scanning. Perfusion scanning involves...

Pulmonary Embolism I: Introduction

Pulmonary Embolism I: Introduction

Pulmonary embolism (PE) occurs when a thrombus, fat or air embolus, amniotic fluid, or tumor tissue blocks one or more pulmonary arteries. These blockages originate in the venous system or the right side of the heart.EtiologyPE primarily arises from deep vein thrombosis (DVT) and other hypercoagulable states, such as inherited thrombophilias. Additional etiological factors include venous stasis, commonly seen in obesity, and endothelial injury from surgery and trauma. Less common causes include...

Pulmonary Embolism III: Nursing Management

Pulmonary Embolism III: Nursing Management

A pulmonary embolism occurs when a thrombus, amniotic fluid, tumor tissue, fat, or air embolus blocks one or more pulmonary arteries. Effective nursing management and patient education are crucial for improving outcomes and preventing recurrence.Nursing management starts with obtaining a comprehensive patient history, particularly noting any history of deep vein thrombosis (DVT). Assess for clinical manifestations, including dyspnea, chest pain, crackles, heart murmurs, and signs of right-sided...