Evaluation of different algorithms to identify the scleroderma pattern in nailfold videocapillaroscopy

Area of Science:

• Rheumatology
• Dermatology
• Medical Diagnostics

Background:

• Nailfold videocapillaroscopy (NVC) is crucial for investigating Raynaud's phenomenon (RP) and systemic sclerosis (SSc).
• Existing methods for identifying the scleroderma pattern via NVC show significant heterogeneity.
• Novel algorithms like Fast Track and CAPI-score aim to standardize SSc pattern identification.

Purpose of the Study:

• To evaluate the accuracy of various NVC parameters and algorithms in identifying the scleroderma pattern.
• To compare the diagnostic performance of the Fast Track algorithm and CAPI-score.
• To assess NVC accuracy in patients with and without RP and autoimmune rheumatic diseases.

Main Methods:

• Analysis of 258 NVCs to determine the accuracy and area under the curve (AUC) of qualitative and quantitative parameters.
• Discrimination between scleroderma and non-scleroderma patterns using statistical analysis.
• Evaluation of specific parameters including capillary density and the presence of giant capillaries.

Main Results:

• The scleroderma pattern was identified in 39.15% of NVCs.
• A capillary density of ≤8 capillaries/mm showed optimal cut-off (AUC 0.911) with 87.94% accuracy.
• Combining giant capillaries with ≤8 capillaries/mm achieved 91.08% accuracy for the scleroderma (SD) pattern.

Conclusions:

• A combination of giant capillaries and capillary density, inspired by the Fast Track and CAPI-score, is highly accurate for defining the scleroderma pattern.
• Optimal capillary density cut-off is ≤8 capillaries/mm for distinguishing the SD pattern.
• The presence of giant capillaries is a highly specific indicator for the scleroderma pattern.