Computer-Aided System for Breast Cancer Lesion Segmentation and Classification Using Ultrasound Images

We proposed an automated ultrasound-based Computer-Aided Diagnosis (CAD) system designed to detect and classify breast cancer lesions. The dataset used in the study comprised 6,319 images from 2,889 patients. The AI algorithm used deep learning methodologies, including the EfficientNetV2-B0 architecture for image classification (benign/malignant) and the Attention U-Net coupled with the Cosh log Dice loss function for breast lesion segmentation. The CAD system demonstrated an impressive sensitivity of 89.0% and specificity of 92.0% for classification, along with a segmentation Dice score of 86.0%. The integration of such CAD systems into breast imaging workflows holds promise for diminishing the influence of human errors, consequently reducing diagnostic costs, and expediting the breast US imaging process.

My presentation for our gradution project paper in EHB 2023 Conference.