Spatial Attention Mechanism and Cascade Feature Extraction in a U-Net Model for Enhancing Breast Tumor Segmentation

Abstract

In the field of medical imaging, the accurate segmentation of breast tumors is a critical task for the diagnosis and treatment of breast cancer. To address the challenges posed by fuzzy boundaries, vague tumor shapes, variation in tumor size, and illumination variation, we propose a new approach that combines a U-Net model with a spatial attention mechanism. Our method utilizes a cascade feature extraction technique to enhance the subtle features of breast tumors, thereby improving segmentation accuracy. In addition, our model incorporates a spatial attention mechanism to enable the network to focus on important regions of the image while suppressing irrelevant areas. This combination of techniques leads to significant improvements in segmentation accuracy, particularly in challenging cases where tumors have fuzzy boundaries or vague shapes. We evaluate our suggested technique on the Mini-MIAS dataset and demonstrate state-of-the-art performance, surpassing existing methods in terms of accuracy, sensitivity, and specificity. Specifically, our method achieves an overall accuracy of 91%, a sensitivity of 91%, and a specificity of 93%, demonstrating its effectiveness in accurately identifying breast tumors.

Multi‐focal channel attention for medical image segmentation

“…This integration enables a more precise spatial aggregation of contextual information, resulting in improved accuracy. Cross-fusion channel attention (CAS-UNet) (You et al, 2023) introduces a method for segmenting blood vessels in the retina that relies on an attention mechanism. The method makes three modifications to the U-Net architecture: first, it replaces the original convolutional block with an enhanced attention block; second, it incorporates an additional AG to the skip-connection layer for spatial enhancement; and third, it employs the SoftPool pooling method to minimize information loss.…”