Prediction and Classification of Alzheimer’s Disease using Machine Learning Techniques in 3D MR Images

Abstract

Memory and thought-related brain cells are damaged permanently by Alzheimer's disease. It has a fatal outcome since it causes death. As a result, early detection of Alzheimer's disease is of utmost importance. Identifying this condition accurately in its initial stages is critical for both patient care and clinical research. As one of the most expensive diseases to treat, Alzheimer's Disease (AD) has drawn the attention of many researchers, who are working towards creating an automated algorithm with high accuracy. It might be challenging to recognize and forecast Alzheimer's disease in its earliest stages. This issue can be resolved by an ML system that can forecast the disease. Machine Learning (ML) has recently gained enormous popularity for its ability to solve issues in variety of domains, including the interpretation of medical imaging. To forecast and categorize Alzheimer's disease, present-day studies utilize 3D Magnetic Resonance Imaging (MRI) scans and machine learning algorithms. This study combines white and grey matter present in MRI images using 3D MRI technology, and subsequently obtains 2D slices in the coronal, sagittal, and axial orientations. After selecting the most relevant slices, feature extraction is performed using Multi-Layer Perceptron (MLP) and SVM methods for predicting and classifying Alzheimer's disease. The researchers assess the system's efficacy using metrics such as Precision, Recall, Accuracy, and F1-Score.