Pathological brain detection based on wavelet entropy and Hu moment invariants

Area of Science:

• Medical Imaging Analysis
• Computer-Aided Diagnosis
• Machine Learning in Healthcare

Background:

• Magnetic Resonance Imaging (MRI) generates large datasets, posing challenges for accurate pathological brain detection.
• Existing methods struggle with the complexity and volume of MRI data, necessitating advanced diagnostic tools.

Purpose of the Study:

• To develop an accurate automatic computer-aided diagnosis (CAD) system for distinguishing pathological brains from normal brains using MRI.
• To enhance the classification accuracy of pathological brain detection systems.

Main Methods:

• A novel two-step approach combining feature extraction and classification was proposed.
• Feature extraction utilized wavelet entropy (WE) and Hu moment invariants (HMI).
• Classification was performed using a generalized eigenvalue proximal support vector machine (GEPSVM) with a radial basis function (RBF) kernel.

Main Results:

• The proposed "WE + HMI + GEPSVM + RBF" method demonstrated superior classification accuracy compared to existing approaches.
• Average classification accuracies of 100%, 100%, and 99.45% were achieved on Dataset-66, Dataset-160, and Dataset-255, respectively.
• The system's effectiveness was validated through 10 runs of k-fold stratified cross-validation.

Conclusions:

• The developed CAD system is effective for pathological brain detection from MRI scans.
• The proposed method shows potential for realistic clinical application due to its high accuracy and efficiency.