Alzheimer's disease classification using competitive swarm multi‐verse optimizer‐based deep neuro‐fuzzy network

Abstract

SummaryClassification of Alzheimer's disease (AD) from neuroimaging, like magnetic resonance imaging (MRI) through deep learning classifier has been increasing in research in recent decades. However, it is required for enhancing the accuracy of the AD classification for effective treatment. In this work, an efficient model termed competitive swarm multi‐verse optimizer + deep neuro‐fuzzy network (CSMVO + DNFN) is designed to accurately classify stages of AD. Preprocessing is done with a median filter. Then, the resulting image is segmented to find the interested regions by channel‐wise feature pyramid network module (CFPNet‐M). Some features obtained from the segmented image are haralick, convolutional neural network, and texture features. The devised method is more efficient in classifying different stages of AD with MRI modality. Furthermore, the developed model attained higher performance using metrics like the accuracy of 89.9%, sensitivity of 89.6%, and specificity of 87.0% based on the k‐fold value.