TGMS-UNet: A dual-branch segmentation network for ultrasound endometrium based on sequence guidance and multi-scale feature correction.
Authors
Affiliations (9)
Affiliations (9)
- Key Laboratory of Medical Imaging Precision Theranostics and Radiation Protection, College of Hunan Province, Hengyang Medical School, The Affiliated Changsha Central Hospital, University of South China, Changsha, China.
- Institute of Medical Imaging, Hengyang Medical School, University of South China, Hengyang, China.
- Institute for Future Sciences, University of South China, Changsha, Hunan, China.
- School of Computer, University of South China, Hengyang, China.
- School of Computer Science and Engineering, Central South University, Changsha, China.
- Department of Ultrasound Medicine, Department of Medical Imaging, Hengyang Medical School, The Affiliated Changsha Central Hospital, University of South China, Changsha, Hunan, China.
- Department of Ultrasound Medicine, The Third Affiliated Hospital of Guangzhou Medical University, Guangzhou, China.
- College of Mechanical Engineering, University of South China, Hengyang, Hunan, China.
- Reproductive Medicine Center, Changsha Jiangwan Maternity Hospital, Changsha, Hunan, China.
Abstract
Ultrasound imaging of the endometrium is often affected by speckle noise, low contrast, and physiological cycle variations, which lead to blurred boundaries and ambiguous structural representation. To address these challenges, this study proposes a segmentation method specifically designed for ambiguous ultrasound scenarios, which provides stronger attention guidance to indistinct boundaries, thereby improving the accuracy of structural contour segmentation. The proposed TGMS-UNet is a dual-branch segmentation network that converts geometric contours into sequential representations by calculating the distance and angle variations from boundary points to the centroid. The sequence-image feature alignment module employs global attention and scanning mechanisms to achieve cross-modal fusion and integrate contour prior knowledge, while the feature correction and adaptive fusion module uses a gating mechanism to rectify multi-scale feature misalignment and suppress erroneous information. The network is trained end-to-end, enabling the fusion of multi-scale image features and sequence-guided boundary information within a unified framework. Experimental results show that TGMS-UNet achieves a Dice of 0.9226 and an IoU of 0.8822 on a private endometrial ultrasound dataset of 1,063 images, and a Dice of 0.8355 and an IoU of 0.8012 on the public BUSI breast ultrasound dataset, outperforming existing mainstream segmentation models on key metrics. The proposed TGMS-UNet demonstrates strong ability to handle ambiguous, low-contrast ultrasound images, particularly in accurately delineating blurred boundaries. By leveraging contour-derived prior information and enhancing feature consistency across scales, it improves segmentation precision and reliability.