Vascular Dementia Characterization by Dynamic Time Warping in the Context of Cognitive Disorders

Distinguishing neurodegenerative diseases in an effective and efficient way is an increasingly important challenge. This study investigates the use of Dynamic Time Warping (DTW) as a functional connectivity (FC) measure, to distinguish patients with Subjective Cognitive Impairments (SCI), Mild Cognitive Impairments (MCI), Alzheimer's Disease (AD) and Vascular Dementia (VaD) with EEG. The FC analysis is performed per cortical region, defining three topological ranges: short-, mid- and long-ranges. We analyze the FC at different threshold values, corresponding to the percentage of the highest values to be kept. The study highlights different FC topologies in the four conditions. Results show that DTW allows a significant distinction of VaD from SCI, MCI and AD, particularly for the short-range FC. Also, SCI, AD and VaD are significantly distinct when considering the mid-range FC for high threshold values. MCI patients show an intermediate behaviour between SCI and AD groups. Interestingly, the study also reveals that VaD group exhibits the lowest short-range connectivity and the highest long-range connectivity.