Imaging segmentation mechanism for rectal tumors using improved U-Net

• Published in: BMC medical imaging Vol. 24; no. 1; p. 95
• Main Authors: Zhang, Kenan, Yang, Xiaotang, Cui, Yanfen, Zhao, Jumin, Li, Dengao
• Format: Journal Article
• Language: English
• Published: England BioMed Central Ltd 23.04.2024; BioMed Central; BMC
• Subjects: Colorectal cancer; Deep Learning; Diagnostic imaging; Gastrointestinal diseases; Humans; Image Interpretation, Computer-Assisted - methods; Image processing; Image segmentation; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Medical care; Medical diagnosis; Medical imaging equipment; Modules; Mortality; MR image; Neural networks; Neural Networks, Computer; Organs; Quality management; Radiation therapy; Radiotherapy; Rectal cancer; Rectal Neoplasms - diagnostic imaging; Rectal Neoplasms - pathology; Rectum; Segmentation; Semantic segmentation; Semantics; Tumors; U-Net; Rectal cancer; Semantic segmentation; U-Net; MR image
• ISSN: 1471-2342; 1471-2342
• DOI: 10.1186/s12880-024-01269-6

In radiation therapy, cancerous region segmentation in magnetic resonance images (MRI) is a critical step. For rectal cancer, the automatic segmentation of rectal tumors from an MRI is a great challenge. There are two main shortcomings in existing deep learning-based methods that lead to incorrect segmentation: 1) there are many organs surrounding the rectum, and the shape of some organs is similar to that of rectal tumors; 2) high-level features extracted by conventional neural networks often do not contain enough high-resolution information. Therefore, an improved U-Net segmentation network based on attention mechanisms is proposed to replace the traditional U-Net network. The overall framework of the proposed method is based on traditional U-Net. A ResNeSt module was added to extract the overall features, and a shape module was added after the encoder layer. We then combined the outputs of the shape module and the decoder to obtain the results. Moreover, the model used different types of attention mechanisms, so that the network learned information to improve segmentation accuracy. We validated the effectiveness of the proposed method using 3773 2D MRI datasets from 304 patients. The results showed that the proposed method achieved 0.987, 0.946, 0.897, and 0.899 for Dice, MPA, MioU, and FWIoU, respectively; these values are significantly better than those of other existing methods. Due to time savings, the proposed method can help radiologists segment rectal tumors effectively and enable them to focus on patients whose cancerous regions are difficult for the network to segment. The proposed method can help doctors segment rectal tumors, thereby ensuring good diagnostic quality and accuracy.