Determination of affected brain regions at various stages of Alzheimer's disease

The objective of study was to explore those brain areas that were affected at each stage during the progression of Alzheimer's disease (AD). Six affected brain areas were explored at mild cognitive impairment, four at first stage and six at each of second and third stage of Alzheimer's disease. The common brain regions among these stages were cuneus, precuneus, calcarine cortex, middle frontal gyrus, superior frontal gyrus, and frontal superior medial gyrus. The fMRI data at the resting state of 18 AD patients who were converted from MCI to stage 3 of Alzheimer's were taken from ADNI public source database. Among these patients, there were ten males and eight females. Independent component analysis was used to explore affected brain regions and an algorithm based on deep learning convolutional neural network was proposed for binary classification among the stages of Alz-heimer's disease. The proposed CNN model delivered 94.6 % accuracy for separating stage 1 of Alzheimer's disease from mild cognitive impairment. 96.7 % accuracy was acquired to distinguish stage 2 of Alzheimer's disease from mild cognitive impairment, and stage 3 of Alzheimer's disease was separated from mild cognitive impairment with an accuracy of 97.8 %.

Automated summary

The study aimed to identify brain areas affected during each stage of Alzheimer's disease (AD). Six areas were affected during mild cognitive impairment, four during the first stage, and six during each of the second and third stages. The common brain regions in all stages were identified. The resting state fMRI data of 18 AD patients from the ADNI database were analyzed using independent component analysis and a deep learning convolutional neural network algorithm. The proposed CNN model achieved high accuracy in classifying the stages of AD.