A CAD system design using iteratively reweighted fuzzy c-means and deep tree training for Alzheimer's disease diagnosis

Abstract

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by memory loss and other cognitive abnormalities. In this research work, an MRI-based Computer-Aided Diagnosis (CAD) system is presented for clinical practice. While current algorithms used for detecting Alzheimer's disease perform inefficiently. Those algorithms are unable to detect minor changes in the disease due to discrepancies in feature extraction, segmentation, and classification methods. To overcome the issues in algorithm, a new iteratively reweighted adaptive spatial fuzzy c-means approach (IRW-ASFCM) is used to segment the brain tissue efficiently. Initially, the Magnetic resonance images (MRI) will pre-processed by using the Statistical parametric mapping (SPM) tool. After pre-processing, the segmentation is done using IRW-ASFCM to get effective tissue segments. Then, a 2-level selective Wavelet Kernel (2LS-WaveCNN ensemble) based CNN ensemble model will be used to extract the feature map from the Gray matter (GM), White matter (WM) and Cerebrospinal fluid (CSF) regions after segmenting the images. The Wave CNN networks built from hidden layers are trained using Deep Tree Training (DTT) for AD classification. The simulation results demonstrate that the suggested CAD system is superior to the state-of-the-art systems in terms of accuracy (99.54 %), sensitivity (99.26 %), and specificity (99.64 %) for multi-class problems.