FeaCL: Carotid plaque classification from ultrasound images using feature-level and instance-level contrast learning

Open Webpage

Cheng Li, Kai Wang, Haitao Gan, Ran Zhou, Xinyao Cheng, Zhi Yang

Published: 2025, Last Modified: 08 Nov 2025Comput. Medical Imaging Graph. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Feature-level and instance-level contrast learning is used in pretext task.•Feature level contrast is designed by a cross-correlation matrix fusion strategy.•A triplet network is used to learn effective feature representation.•The performance of downstream task verifies effectiveness in plaque classification.