An experimental protocol to support cognitive impairment diagnosis by using handwriting analysis

Nowadays diseases involving cognitive impairments affect millions of people worldwide, with Alzheimer's and Parkinson's diseases being the most common ones. Because of the worldwide average lifes10.1016/j.procs.2018.10.141pan increment, it is expected that their incidence will increase in the next few decades. Among the daily activities, handwriting is one of the first affected by cognitive impairments. For this reasons, researchers have also been investigating the analysis of handwriting alterations as diagnostic signs for this kind of diseases. In this paper we present an experimental protocol that we developed for the analysis of the handwriting dynamics of patients affected by cognitive impairments. The aim of this protocol is to build a large database that would allow to effectively train different classifier systems. We also detail the most common and effective features previously used in the literature to represent handwriting dynamics of the subjects affected by cognitive impairments.