Multi-modal MRI (Magnetic Resonance Imaging) brain tumor segmentation method based on convolution attention

By introducing a convolutional attention mechanism into the Transformer model, the problems of insufficient local structure perception and low computational efficiency in medical image segmentation are solved, achieving more accurate multimodal information fusion and segmentation, which is suitable for efficient deployment of high-resolution 3D medical images.

CN121837641APending Publication Date: 2026-04-10UNIV OF SHANGHAI FOR SCI & TECH
0 Cites 0 Cited by

Patent Information

Application Number
CN202610062169.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-01-17
Publication Date
2026-04-10

AI Technical Summary

Technical Problem

Existing Transformer models lack the ability to extract features sensitive to local structures in medical image segmentation, have inefficient global attention mechanisms during multimodal fusion, and suffer from insufficient generalization performance due to data scarcity.

Method used

A multimodal MRI brain tumor segmentation method based on convolutional attention is constructed by replacing traditional linear transformations with convolutional operations to generate query, key, and value vectors. Enhanced features are obtained through parallel encoding, feature fusion, and convolutional attention processing to achieve the final segmentation result.

Benefits of technology

It improves the model's ability to perceive local structures, enhances the accuracy and robustness of multimodal information fusion, reduces computational complexity and overfitting risk, and is suitable for efficient deployment of high-resolution 3D medical images.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to a multi-modal MRI brain tumor segmentation method based on convolution attention, and the method comprises the steps: carrying out the parallel coding of multi-modal medical image data, and extracting the specific low-layer space features of each modal; performing feature fusion and high-level coding on the low-level spatial features to obtain shared features; carrying out convolution attention processing on the shared features to obtain enhanced features; and carrying out decoding processing on the enhanced features to obtain a segmentation result. According to the method, more accurate tumor boundary segmentation and tissue differentiation can be realized in medical image analysis such as brain MRI (Magnetic Resonance Imaging), meanwhile, the parameter efficiency, the calculation speed and the generalization ability under limited data of the model are remarkably improved, and the clinical practicability is enhanced.
Need to check novelty before this filing date? Find Prior Art