This repository contains data resources accompanying the article:
Simisola Odimayo, Chollette C. Olisah, Khadija Mohammed "SNeurodCNN: Structure-focused Neurodegeneration Convolutional Neural Network for Modeling and Classification of Alzheimer’s Disease",
In this article, we designed and developed convolutional neural network for classifying AD and MCI, with the potential for extracting features useful in predicting the conversion of MCI to AD. Two datasets, comprising of midsagittal and parasagittal images, consisting of 180 AD, 105 sMCI and 83 pMCI patients were acquired from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) repository. The images were labelled based on their respective classes, pre-processed using the Gamma correction method to enhance pixel brightness. Subsequently, the dataset was split into an 80:20 ratio, with 80% for training and 20% for testing in the classification task. The proposed classifier was trained on the two datasets and successfully classified the midsagittal data into groups AD and MCI with an accuracy of 98.1% along with 97.2% specificity and 99.0% sensitivity and the parasagittal data with 97.8% accuracy and 97.0% and 98.5% for specificity and sensitivity respectively.
List of all 180 Alzheimer Disease subjects [Alzheimers Disease DEMOGRAPHICS.csv] List of all 105 Stable Mild Cognitive Impairment subjects Stable Mild cognitive impairment patient DEMOGRAPHICS.csv List of all 83 Progressive Mild Cognitive Impairment subjects [Progressive Mild Cognitie Impairment Demogrpahics.csv]