Studies from Iran University of Science and Technology Yield New Information about Cancer (Mass Segmentation In Automated Breast Ultrasound Using an Enhanced Attentive Unet).

A study conducted by researchers at Iran University of Science and Technology in Tehran focuses on the development of a fully-automatic segmentation system for breast cancer detection using 3-D Automated Breast Ultrasound (ABUS). The system utilizes a bidirectional convolutional long short-term memory (Bi-ConvLSTM) network for tumor detection and a 2-D attentive UNet for precise segmentation of breast masses. The proposed modifications to the system achieved a Dice similarity index of 85.82% on a dataset of 60 volumes. This research has been peer-reviewed and published in Expert Systems with Applications. [Extracted from the article]