High-Performance Method for Brain Tumor Feature Extraction in MRI Using Complex Network.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Oncology

Background:

• Accurate differentiation between benign and malignant brain tumors is crucial for effective patient management.
• Magnetic Resonance Imaging (MRI) is a primary tool for brain tumor diagnosis.
• Current classification methods can be limited in accuracy and efficiency.

Purpose of the Study:

• To develop and validate a high-performance method for brain tumor feature extraction and classification using MRI.
• To accurately distinguish between benign and malignant brain tumors.
• To assess the efficacy of a combined complex network and U-Net architecture approach.

Main Methods:

• A novel method integrating complex network and U-Net architecture for brain tumor feature extraction was proposed.
• Machine learning algorithms were employed for tumor classification.
• The method was tested on a dataset of 230 brain tumor patient MRIs (77 high-grade glioma, 153 low-grade glioma).

Main Results:

• The proposed method achieved a classification accuracy of 99.84% in distinguishing between benign and malignant brain tumors.
• The combination of complex network and U-Net architecture demonstrated superior performance in feature extraction.
• Experimental results validated the high accuracy of the classification model.

Conclusions:

• The developed method significantly enhances the accuracy of brain tumor classification.
• The integration of complex networks and U-Net architecture shows great potential for improving diagnostic accuracy.
• This approach could serve as a valuable tool for clinicians in aiding brain tumor diagnosis and treatment planning.