Semi-global Parameterization of Online Handwriting Features for Characterizing Early-Stage Alzheimer and Mild Cognitive Impairment

•A novel semi-global feature parameterization is proposed for handwriting.•Clustering-based feature selection with Normalized Mutual Information is performed.•Handwriting semi-global features outperform standard global features.•Our method discriminates Early-Stage Alzheimer patients from Healthy Controls.•Our method distributes Mild Cognitive Impairment patients over two clusters.