针对腹部图像的多器官分割方法及成像方法
By constructing a multi-organ segmentation model for abdominal images based on KNN and graph representation learning, the consistency and accuracy issues in abdominal image segmentation are solved, achieving efficient and reliable multi-organ segmentation results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CENT SOUTH UNIV
- Filing Date
- 2024-03-20
- Publication Date
- 2026-07-17
AI Technical Summary
Existing methods for segmenting multiple organs in abdominal images suffer from problems such as being time-consuming and labor-intensive, relying on doctors' experience, having poor consistency, and lacking reliability and accuracy. Furthermore, existing algorithms perform poorly in segmenting complex abdominal structures.
A multi-organ segmentation model based on the KNN algorithm and graph representation learning algorithm is constructed, including a patch transformation layer, an encoder layer, a decoder layer and an output layer. Feature extraction and aggregation are performed using a graph transformation sub-layer, a k-hop neighborhood feature mixing sub-layer and a feedforward neural network sub-layer. Abdominal image segmentation is performed by combining Chebyshev polynomial recursion and convolution operations.
This method achieves high reliability and accuracy in multi-organ segmentation of abdominal images, improves the consistency and efficiency of segmentation results, and overcomes the shortcomings of traditional methods.
Smart Images
Figure CN118229721B_ABST