UC San Diego researchers developed an AI tool that accurately predicts colorectal cancer risk in ulcerative colitis patients using clinical and imaging data.
Key Details
- 1AI workflow reviewed records of 55,000 patients in the VA health care system, the largest US dataset of its kind.
- 2Combined large language models and statistical risk models to predict progression from low-grade dysplasia to cancer.
- 3Classified patients into five risk groups based on lesion size, number, visibility, resection completeness, and inflammation severity.
- 4Nearly half of patients were categorized as lowest risk, with the model correctly predicting 99% would not develop cancer within two years.
- 5The tool matched real-world outcomes for over ten years of follow-up data.
- 6The model may allow some low-risk patients to have less frequent surveillance colonoscopies.
Why It Matters
This AI system offers more precise, data-driven cancer risk assessments for clinicians, potentially reducing unnecessary procedures and focusing attention on high-risk patients, improving both care quality and resource allocation in gastrointestinal oncology.
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.