Computational Analysis and Deep Learning for Medical Care

14Early Diagnosis Tool for Alzheimer’s Disease Using 3D Slicer

V. Krishna Kumar^1*, M.S. Geetha Devasena¹ and G. Gopu²

¹Computer Science and Engineering, Sri Ramakrishna Engineering College, Coimbatore, India

²Electronics and Communication Engineering, Sri Ramakrishna Engineering College, Coimbatore, India

Abstract

Alzheimer’s disease (AD) is a progressive and chronic neurodegenerative disease caused by loss of neurons and neural connections. Though physical and neurological assessments can be helpful in clinical analysis, there is a need for developing new techniques for early prediction of this disease. The proposed model strives to improve the healthcare offered to the patients by predicting the onset of AD at a much earlier date. The data is acquired from ADNI (Alzheimer’s Disease Neuro Imaging) dataset. The variables used in this model are scores from MMSE, FAQ, CDR, and logical memory delayed recall. Various machine learning algorithms such as KNN, random forest, SVM with linear kernel, generalized linear model (GLM), and bagged CART are then implemented using the above variables to determine the accuracy of each algorithm. However, further improvement to the model is done by using 3D slicer. The hippocampus is the first region to be affected by atrophy when Alzheimer’s sets in. The proposed model uses the 3D slicer to process the MRI data of the hippocampus of the subject to derive the Dice coefficient which is then added as a prediction variable to the model. The data ...