SU-Next: An Image Segmentation Model of Breast Tumors Based on U-Next and Attention Mechanisms

Abstract Breast cancer is one of the most common malignant tumors endangering women’s health today, and timely screening targeted treatment such can effectively reduce incidence breast cancer. The commonly used method diagnosis to detect pathological image tumor find spread cells difference in tissue morphology over time. There are a large number complex problems as interlaminar variation characteristics structural morphological diversity sections tumors, must be accurately segmented before detection, accuracy segmentation directly affects results subsequent detection. Aiming at problem that U-Next network model not clearly defined marginal region we propose SU-Next model. adds an attention module U-Next, which focuses on feature values related current task discards irrelevant features. By testing BUSI dataset, effectiveness this was verified. Taking IoU Dice evaluation indicators, they reached 64.26% 77.40%, respectively, compared with improved by 1% 1.19% respectively.