Pathological brain detection in MRI scanning by wavelet packet Tsallis entropy and fuzzy support vector machine

Area of Science:

• Medical Imaging
• Machine Learning
• Biomedical Engineering

Background:

• Computer-aided diagnosis (CAD) systems are crucial for interpreting medical images.
• Accurate pathological brain detection (PBD) aids physicians in diagnosis.
• Magnetic resonance imaging (MRI) is a key modality for brain imaging.

Purpose of the Study:

• To develop a novel automatic PBD system for distinguishing pathological from healthy brains using MRI.
• To simplify PBD into a binary classification task.
• To evaluate the efficacy of the proposed method against existing state-of-the-art techniques.

Main Methods:

• Feature extraction using wavelet packet Tsallis entropy (WPTE).
• Classification using fuzzy support vector machines (FSVM).
• Validation on 3 benchmark datasets using K-fold cross-validation.

Main Results:

• The proposed WPTE + FSVM method achieved superior classification accuracy compared to 17 state-of-the-art methods.
• WPTE demonstrated superiority over discrete wavelet transform.
• Tsallis entropy outperformed Shannon entropy, and FSVM excelled standard SVM.

Conclusions:

• The "WPTE + FSVM" method is highly effective for pathological brain detection.
• The study highlights the advantages of WPTE and Tsallis entropy in image analysis.
• The findings support the clinical utility of advanced machine learning techniques in neuroimaging.