Advanced Analysis of Online Handwriting in a Multilingual Cohort of Patients with Parkinson's Disease

Advanced Analysis of Online Handwriting in a Multilingual Cohort of Patients with Parkinson's Disease

Paper in proceedings (conference paper)

The majority of Parkinson’s disease (PD) patients suffer from handwriting abnormalities commonly called as Parkinsonic dysgraphia. Several approaches of PD dysgraphia analysis exist, e.g. based on online handwriting processing. However, a small and unilingual cohort of PD patients is often an issue in quantitative PD dysgraphia analysis studies. Therefore, in this work, we aim to perform a discrimination analysis in a multilingual cohort of 73 PD patients and 48 healthy controls (Spanish and Czech). For this purpose, we extracted advanced handwriting features based on fractional order derivatives (FD). Discrimination power of the advanced FD-based features was evaluated by Mann-Whitney U test and random forests classifier. We reached 82 % classification accuracy (86 % sensitivity, 77 % specificity) in the multilingual cohort. In addition, we observed high discrimination power of the FD-based parameters and proofed the high impact of online handwriting processing in cross-cultural PD dysgraphia analysis studies.

The majority of Parkinson’s disease (PD) patients suffer from handwriting abnormalities commonly called as Parkinsonic dysgraphia. Several approaches of PD dysgraphia analysis exist, e.g. based on online handwriting processing. However, a small and unilingual cohort of PD patients is often an issue in quantitative PD dysgraphia analysis studies. Therefore, in this work, we aim to perform a discrimination analysis in a multilingual cohort of 73 PD patients and 48 healthy controls (Spanish and Czech). For this purpose, we extracted advanced handwriting features based on fractional order derivatives (FD). Discrimination power of the advanced FD-based features was evaluated by Mann-Whitney U test and random forests classifier. We reached 82 % classification accuracy (86 % sensitivity, 77 % specificity) in the multilingual cohort. In addition, we observed high discrimination power of the FD-based parameters and proofed the high impact of online handwriting processing in cross-cultural PD dysgraphia analysis studies.

Parkinsonic dysgraphia, Micrographia, Online handwriting, Fractional order derivative, Fractional calculus, Multilingual cohort.

Parkinsonic dysgraphia, Micrographia, Online handwriting, Fractional order derivative, Fractional calculus, Multilingual cohort.

MUCHA, J.; MEKYSKA, J.; FAÚNDEZ ZANUY, M.; SANZ-CARTAGENA, M.; GALÁŽ, Z.; ZVONČÁK, V.; KISKA, T.; SMÉKAL, Z.; LOPEZ-DE-IPINA, K.; REKTOROVÁ, I.

2019

20.03.2019

International Frequency Sensor Association (IFSA) Publishing, S. L.

Barcelona, Spain

978-84-09-10127-6

Advances in Signal Processing and Artificial Intelligence: Proceedings of the 1st International Conference on Advances in Signal Processing and Artificial Intelligence

1

144

147

4