Classification of Subjective Cognitive Decline in Alzheimer’s Disease Through Resting-State Hemodynamic Response Function

Abstract

As a super-early stage of Alzheimer’s disease (AD), there is no unified structural change in imaging nor significant difference in assessments in the Subjective Cognitive Decline (SCD), lack of effective clinical diagnosis. The hemodynamic response function (HRF), as the basis of functional magnetic resonance imaging(fMRI), represents the tendency of specific cortex voxel over time after specific stimulation, reflecting the sensitivity and activation of neurons. We equalize the spontaneous dynamic change of the cerebral cortex in resting state to equivalent stimulation for activation, extend the application of HRF. The possible lesion is selected through priori knowledge both academically and clinically. The parameters of HRF: peak, time to peak and FWHM were extracted as the basis for classification via support vector machine (SVM). The SCD group was found deactivation in specific frontal, hippocampus, inferior temporal and occipital regions especially. The classification accuracy between the healthy controls (HC) can reach 74%, which could be great referable in clinical pre-diagnosis. It highlights the possible ROI for SCD stage, provides new methods and research materials for pre-diagnosis of early-stage Alzheimer’s disease as a typical brain-related disease.