3D-MRI brain glioma intelligent segmentation based on improved 3D U-net network

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Neuro-oncology

Background:

• Accurate glioma segmentation from 3D MRI is crucial for clinical decision-making.
• Existing segmentation methods face challenges with complex tumor structures and data variability.

Purpose of the Study:

• To develop and validate an intelligent 3D-MRI glioma segmentation method using deep learning.
• To enhance the accuracy and generalization of glioma segmentation for improved medical diagnosis, grading, and treatment planning.

Main Methods:

• Utilized the BraTS2023 dataset, applying preprocessing techniques like 3D clipping, resampling, artifact elimination, and normalization.
• Incorporated a space pyramid pool module for multi-scale feature perception and a multi-scale fusion attention mechanism to focus on tumor details.
• Employed a combined Dice and Focal loss function to address class imbalance and improve learning of challenging voxels.

Main Results:

• The enhanced 3D U-Net model achieved stable training loss and demonstrated superior segmentation performance.
• Achieved high Dice Similarity Coefficient (DSC), Recall, and Precision scores across Whole Tumor (WT), Core Tumor (TC), and Enhanced Tumor (ET) segments.
• Specific scores included DSC of 0.9168 (WT), 0.8954 (TC), and 0.8674 (ET), indicating robust segmentation capabilities.

Conclusions:

• The proposed deep learning method significantly enhances glioma segmentation accuracy and reliability.
• The high performance indices in WT, TC, and ET segments provide a strong scientific foundation for clinical diagnosis and treatment strategies.
• This advancement offers improved guidance for medical diagnosis, grading, and treatment selection in neuro-oncology.