Publication detail

Fractional Derivatives of Online Handwriting: a New Approach of Parkinsonic Dysgraphia Analysis

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

Original Title

Fractional Derivatives of Online Handwriting: a New Approach of Parkinsonic Dysgraphia Analysis

English Title

Fractional Derivatives of Online Handwriting: a New Approach of Parkinsonic Dysgraphia Analysis

Type

conference paper

Language

en

Original Abstract

Parkinson’s disease (PD) is the second most frequent neurodegenerative disorder. One typical hallmark of PD is disruption in execution of practised skills such as handwriting. This paper introduces a new methodology of kinematic features calculation based on fractional derivatives applied on PD handwriting. Discrimination power of basic kinematic features (velocity, acceleration, jerk) was evaluated by classification analysis (using support vector machines and random forests). For this purpose, 30 PD patients and 36 healthy controls were enrolled. In comparison with results reported in other works, the newly designed features based on fractional derivatives increased classification accuracy by 8% in univariate analysis and by 10% when employing the multivariate one. This study reveals an impact of fractional derivatives based features in analysis of Parkinsonic dysgraphia.

English abstract

Parkinson’s disease (PD) is the second most frequent neurodegenerative disorder. One typical hallmark of PD is disruption in execution of practised skills such as handwriting. This paper introduces a new methodology of kinematic features calculation based on fractional derivatives applied on PD handwriting. Discrimination power of basic kinematic features (velocity, acceleration, jerk) was evaluated by classification analysis (using support vector machines and random forests). For this purpose, 30 PD patients and 36 healthy controls were enrolled. In comparison with results reported in other works, the newly designed features based on fractional derivatives increased classification accuracy by 8% in univariate analysis and by 10% when employing the multivariate one. This study reveals an impact of fractional derivatives based features in analysis of Parkinsonic dysgraphia.

Keywords

Archimedean spiral; binary classification; fractal calculus; fractional derivative; online handwriting; Parkinson’s disease;

Released

04.06.2018

Location

Atény, Řecko

ISBN

978-1-5386-4695-3

Book

41st International Conference on Telecommunications and Signal Processing (TSP)

Pages from

214

Pages to

217

Pages count

4

URL

BibTex


@inproceedings{BUT148762,
  author="Ján {Mucha} and Vojtěch {Zvončák} and Zoltán {Galáž} and Jiří {Mekyska} and Tomáš {Kiska} and Zdeněk {Smékal} and Marcos {Faúndez Zanuy} and Luboš {Brabenec} and Irena {Rektorová} and Karmele {Lopez-de-Ipina}",
  title="Fractional Derivatives of Online Handwriting: a New Approach of Parkinsonic Dysgraphia Analysis",
  annote="Parkinson’s disease (PD) is the second most frequent neurodegenerative disorder. One typical hallmark of PD is disruption in execution of practised skills such as handwriting. This paper introduces a new methodology of kinematic features calculation based on fractional derivatives applied on PD handwriting. Discrimination power of basic kinematic features (velocity, acceleration, jerk) was evaluated by classification analysis (using support vector machines and random forests). For this purpose, 30 PD patients and 36 healthy controls were enrolled. In  comparison with results reported in other works, the newly designed features based on fractional derivatives increased classification accuracy by 8% in univariate analysis and by 10% when employing the multivariate one. This study reveals an impact of fractional derivatives based features in analysis of Parkinsonic dysgraphia.",
  booktitle="41st International Conference on Telecommunications and Signal Processing (TSP)",
  chapter="148762",
  doi="10.1109/TSP.2018.8441293",
  howpublished="online",
  year="2018",
  month="june",
  pages="214--217",
  type="conference paper"
}