Systemic sclerosis & Raynaud's
CAPI-Detect: machine learning in capillaroscopy reveals new variables influencing diagnosis
Authors' conclusion
CAPI-Detect, an ML-based model, provides an unbiased, quantitative analysis of capillary structure, shape, size and density, significantly improving capillaroscopic pattern identification.
Study context
- Population
- Nailfold videocapillaroscopies randomly and blindly analysed by three to four trained observers1,490 with consensus
- Participants
- 1,780 participants
- Comparison
- CAPI-Detect machine learning (ML) model vs CAPI-Score for classifying capillaroscopy scleroderma patterns
- Outcome
- Accuracy of CSP classification (SSc vs non-SSc, among SSc patterns, normal vs non-specific)
Methodological detailsmachine learning (CatBoost), blinded observer consensus, validation sets
- Machine learning (CatBoost)
- Blinded observer consensus
- Validation sets
Performance
As reported in the abstract
Accuracy of the ML model (partial and full consensus datasets)
Accuracy, SSc vs non-SSc
0.912
Accuracy, among SSc patterns
0.812
Accuracy, normal vs non-specific patterns
0.746
Further findings
Accuracy of the ML model, full consensus only
- Accuracy, SSc vs non-SSc
- 0.910
- Accuracy, among SSc patterns
- 0.925
- Accuracy, normal vs non-specific patterns
- 0.933
CAPI-Detect vs CAPI-Score
- CAPI-Detect outperformed CAPI-Score
Original publication
CAPI-Detect: machine learning in capillaroscopy reveals new variables influencing diagnosis. Rheumatology (Oxford). Published 7 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →