Analýza písma pre klinické počítačové systémy pre podporu rozhodovania

The goal of this project is to provide deep handwriting analysis and investigate whether handwriting can be used for diagnosis and monitoring of hepatic encephalopathy and whether there exist correlation between handwriting and other biomarkers in hepatic encephalopathy. The number of novel signal processing algorithms, which reveal alteration in handwriting more accurately than existing approaches, are proposed, giving rise to new group of handwriting features. The robust feature selection algorithms select the optimal subset of these features, which are fed into non-parametric regression and classification algorithms, mapping the signal processing algorithm outputs to stage of hepatic encephalopathy.