Developing an accurate and efficient automated system for the early detection of Alzheimer’s disease (AD) is of prime importance for effective treatment. Recently, there has been great interest in Computer Aided Diagnosis (CAD) system for AD. However, differentiating normal control from AD patients is a very tough task due to their patterns and image intensities. Wavelet transform is a mathematical tool used to divide a given signal or image into different scale components and it has been successfully applied to many fields including medical. The main goal of this paper is to develop a CAD system for AD diagnosis using soft computing model. In this paper, a CAD system of AD diagnosis using Discrete Wavelet Transform (DWT) and Multilayer Perceptron (MLP) is formulated, which is capable of differentiating normal control from AD. The proposed system employs DWT for extracting most important features from the image. Principal Component Analysis (PCA) is adopted to reduce the dimension of the feature vector. Finally, reduced feature set is passed to MLP to distinguish normal control and AD from Positron Emission Tomography (PET) image. Proposed method is implemented, analyzed and compared with other existing method in terms of classification accuracy, sensitivity and specificity. Results demonstrated that the proposed CAD system yields better accuracy than other CAD methods reported in the literature. Additionally, the proposed CAD can be used as a diagnostic tool for AD with the capability of defining early stages of the disease.
Volume 11 | 07-Special Issue
Pages: 1223-1230