This document summarizes a study on segmenting cysts in breast ultrasound images using texture features and an active contour method. The authors apply the Chan-Vese level set method to segment cyst regions based on texture features calculated from the images using different kernel sizes. Segmentation performance is evaluated using measures like area error rate, DICE coefficient, sensitivity and Hausdorff distance. The results show that mean texture features and preprocessing the images with Qui's mask produce more accurate segmentations with lower error rates compared to other texture features and kernels.