Wavelet Filter for Alzheimer's Classification from MRI Images using Adaboos

Abstract

Alzheimer’s Disease (AD) is a disorder of the brain which is progressive, destroying memory and
the ability to think. The patients of AD suffer from problems such as lack of initiative, change in
personality and behavior which is seen in their daily functions either at work or home and also in
taking care of oneself which eventually results in death. One widely used technique for diagnosing
of the AD is Magnetic Resonance Imaging (MRI) owing to its non-invasive nature and is widely
adopted in several hospitals to examine certain cognitive abnormalities. The MRI images are
processed using image processing techniques to identify the AD. Image processing techniques
extract features from the images and classify them as either normal or abnormal (AD). The primary
reason for the robustness of the Wavelet Transform is its flexibility in choosing bases and also its
low level of complexity of computation. The basic idea behind the AdaBoost algorithm was that it
was a combination of the weak classifiers to build a robust classifier. In this work, an improved
wavelet filter is proposed to classify Alzheimer's using MRI images with a modified AdaBoost.