A new study evaluates the diagnostic accuracy of three leading generative multimodal AI models in interpreting CT images for lung cancer detection.
Key Details
- 1Three models compared: Gemini-pro-vision (Google), Claude-3-opus (Anthropic), and GPT-4-turbo (OpenAI).
- 2On 184 malignant lung cases, Gemini achieved highest single-image accuracy (>90%), followed by Claude-3-opus, GPT lowest (65.2%).
- 3Gemini's performance dropped to 58.5% with continuous CT slices, indicating challenges with spatial reasoning in imaging.
- 4Simplified text prompts improved diagnostic AUCs: Gemini (0.76), GPT (0.73), and Claude (0.69).
- 5Claude-3-opus showed superior consistency and lower variation in lesion feature analysis.
- 6External validation with TCGA and MIDRC datasets supported findings, especially with simplified prompt strategies.
Why It Matters
This benchmark provides essential insight into the current capabilities and limitations of leading multimodal LLMs for radiological image analysis. Understanding model strengths, weaknesses, and prompt engineering strategies will guide their optimal integration into clinical workflows.
Source
EurekAlert
Related News
•EurekAlert
AI-Powered OCT Enables Rapid 'Optical Biopsy' for Early Endometrial Cancer Detection
A team at Washington University has developed a catheter-based 3D OCT system with AI to quickly and noninvasively detect early endometrial cancers.
•EurekAlert
AI Clinical Reasoning in Diagnostics and Digital Fatigue in Healthcare
Recent JMIR features explore large language models in clinical diagnostics and digital fatigue among healthcare professionals.
•EurekAlert
KAIST, MIT, Microsoft Develop Efficient AI Image Upsampling for Robotics
KAIST, MIT, and Microsoft have created 'Upsample Anything,' a training-free AI method to restore high-resolution visual data from compressed images with up to 16x improved GPU memory efficiency.