A Computed Aided Diagnosis tool for Alzheimer's disease based on 11C-PiB PET imaging technique

Abstract

Pittsburgh compound B Positron Emission Tomography (PiB PET) imaging is a new technique to detect amyloid-beta (Aβ). Aβ is a pathological bio-data which appears distinctly in most neuro-degeneration diseases, such as Alzheimer's disease (AD). Although PiB PET imaging is relative mature, the accurate diagnosis of AD based on PiB PET images still remains a challenge for radiologists. To solve above problem, this paper proposes a Computed Aided Diagnosis (CAD) tool, which combines three machine learning kernels: Principal Component Analysis (PCA), Independent Component analysis (ICA) and Support Vector Machine (SVM). The experimental results with 120 groups of PiB PET images showed that the proposed CAD tool can yield a high accuracy in AD diagnosis.