An automated system for detecting and measuring nailfold capillaries

Area of Science:

• Biomedical Engineering
• Medical Imaging Analysis
• Rheumatology

Background:

• Nailfold capillaroscopy is a standard qualitative method for evaluating Raynaud's phenomenon.
• Current methods lack quantitative analysis, limiting diagnostic precision.
• Distinguishing primary Raynaud's phenomenon from systemic sclerosis is clinically crucial.

Purpose of the Study:

• To develop and validate a fully automated machine learning system for quantitative biomarker extraction from nailfold capillaroscopy images.
• To assess the system's ability to differentiate between primary Raynaud's phenomenon and systemic sclerosis.

Main Methods:

• A layered machine learning approach was employed for image analysis.
• The system automatically detects and localizes capillaries within images.
• Quantitative measurements of vessel morphology were extracted.

Main Results:

• The automated system demonstrated expert-level performance in capillary detection and localization.
• Statistically significant differences in vessel morphology were identified between patient groups.
• The system successfully differentiated primary Raynaud's phenomenon from systemic sclerosis.

Conclusions:

• Automated quantitative analysis of nailfold capillaroscopy images is feasible and effective.
• This system offers a promising tool for improving the diagnosis and management of connective tissue diseases.
• Quantitative biomarkers may enhance the early detection of systemic sclerosis.