Publications by authors named "Etheve Braz-Ma"
Microvasc Res
January 2025
Article Synopsis
- The study aimed to evaluate machine learning and deep learning techniques for detecting systemic scleroderma using nailfold capillaroscopy images from patients in the French SCLEROCAP study.
- Machine learning classifiers showed varying performance, with the best results (F1 score of 0.79) achieved using the light gradient boosting model, while deep learning with DenseNet-121 greatly outperformed with an accuracy of 0.94.
- This research highlights the potential of advanced machine learning and deep learning methods for improving the accuracy of systemic scleroderma diagnosis through image analysis.
