A VGG attention vision transformer network for benign and malignant classification of breast ultrasound images

Abstract

In this study, we proposed the VGGA-ViT for the BUS classification, which was good at learning both local and global features. The proposed network achieved higher accuracy than the compared previous methods.