Preprocessing of MRI Data for the Early Detection of Alzheimer’s Disease Using DWT

A neurological condition known as Alzheimer’s disease (AD) usually occurs in individuals aged 60 and above. AD is progressive in nature. The AD causes destruction of neurons in the brain, initially causing memory loss, which upon progressing causes the loss of taste and eventually leads to the person’s death. Cognitive normal (CN), mild cognitive impairment (MCI), and Alzheimer’s disease are the different phases of cognition. At the MCI stage, patients can either develop or not progress into AD. AD can be detected through the diagnosis of MRI scans of patients. This process can be automated using the concepts of machine learning or neural networks. This paper discusses several techniques to reduce the computational complexity of the MRI images by segmentation of the 3D MRI data to obtain 2D slices of the brain and applying different data preprocessing techniques in the form of discrete wavelet transforms to process the data before doing classification using different machine learning models. This paper aims at finding the best preprocessing technique and machine learning algorithm to obtain high accuracy. This work shows that the Haar wavelet gives the best accuracy with the random forest algorithm. The proposed method resulted in an accuracy of 97.5%.