How word choice affects cognitive impairment detection by handwriting analysis: A preliminary study

In this paper we present the results of a preliminary study in which we considered two copy tasks of regular words and non-words, collecting the handwriting data produced by 99 subjects by using a graphic tablet. The rationale of our approach is to analyze kinematic and pressure properties of handwriting by extracting some standard features proposed in the literature for testing the discriminative power of non-words task to distinguish patients from healthy controls. To this aim, we considered two classification methods, namely Random Forest and Decision Tree, and a standard statistical ANOVA analysis. The obtained results are very encouraging and seem to confirm the hypothesis that machine learning-based analysis of handwriting on the difference of Word/Non-Word tasks can be profitably used to support the Cognitive Impairment diagnosis.