Comparative analysis and optimization of multi-model artificial intelligence for microscopic detection of fungal structures in superficial fungal infections
An AI system identifies fungal infections accurately in microscope images and deployed successfully to clinical labs, demonstrating real-world diagnostic support.
This study benchmarks multiple YOLO-series object detection architectures for automated microscopic identification of fungal structures in clinical specimens from 1,730 patients, finding that YOLOv11 with optimized IoU threshold achieves near-perfect F1 scores for hyphae detection. The model was deployed as an online diagnostic platform, demonstrating a feasible path from model development to clinical decision support in dermatology and mycology.
What the study was
- Study design
- Comparative study (prospective patient specimens)
- Category
- T4: AI/ML in clinical diagnostics and imaging
- Maturity
- Validated
- Journal
- Frontiers in Cellular and Infection Microbiology
Why it surfaced
Demonstrates AI-driven diagnostic automation at a clinically meaningful accuracy threshold with real patient specimens and direct deployment—a model example of the watchlist goal for AI in clinical diagnostics achieving validated performance.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.