Multiclass classification of initial stages of Alzheimer´s disease using structural MRI phase images

Alzheimer´s disease (AD) is the most common type of dementia that is affecting the elderly population worldwide. We present here a novel method based on the progressive two class proximal support vector machine based decision (pTCDC- PSVM) classifier to distinguish between the elderly patients with AD, mild cognitive impairment (MCI) and normal controls (NC). Structural phase images are formed to extract useful features using independent component analysis (ICA) technique which are subsequently used for the classification purposes. The results obtained show the efficacy of our approach and the significant advantages associated with the use of structural magnetic resonance imaging (MRI) phase images in discriminating the early categories of Alzheimer´s disease.