Mayo Clinic researchers created an AI model to analyze patient photos for surgical site infection detection with strong accuracy.
Key Details
- 1AI system developed by Mayo Clinic automates detection of surgical site infections using patient-submitted wound images.
- 2Model was trained on 20,000+ images from more than 6,000 patients across nine hospitals.
- 3Two-stage Vision Transformer model: incision detection (94% accuracy) and infection detection (81% AUC).
- 4Demonstrated performance consistency across diverse patient demographics, addressing bias concerns.
- 5Potential to accelerate infection detection, reducing delays in follow-up care and costs.
- 6Further prospective studies are planned for clinical validation.
Why It Matters
This AI-based system could streamline postoperative care by facilitating early infection identification, especially during virtual follow-ups, improving outcomes and clinician workflows. The tool highlights progress in applying imaging AI to routine, patient-generated health data.
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.