DIMOO: Deep Learning Inspired Multiobjective Optimization Approach to Detect Premature Progression of Alzheimer’s Disease

Authors

  • Sai Sindhuri Nasina , A. Rama Mohan Reddy

Abstract

— Alzheimer’s disease is considered as one of the popular brain pathological disorder that results
in the permenant damage of memory cells. Early detection of this disease may help the medical experts in
treating the affected subjects and prevent the advancement of the disease. Regardless of other neuroimaging
based diagonostic mechanisms MRI(magnetic resonance imaging) is considered as a proven efficient
modality that evaluates the significant anatomic alterations of the brain from the pathological point of view.
Existing research studies indicate that MRI mode of diagnosing has depicted relatively adequate outcomes in
terms of classification of the MRI image from normal to Alzheimer’s disease.Despite in many cases of the
clinical analysis, it is evidenced that MRI subsequently produces the insensitive outcomes during the
analysis timeframe between the normal controls and the MCD (Moderate Cognitive Decline/ Impairement)
phase. Addressing this problem, the studies in the article presents a deep learning inspired optimization
model to evaluate optimal threshold value that aids in accurate image segmentation to classify MRIs of
normal and MCD. Experimental results depict that the proposed model out performs several existing
mechanisms in the context of image segmentation.

Published

2020-01-31

Issue

Section

Articles