Detail publikace

Monitoring Parkinson Disease from speech articulation kinematics

Originální název

Monitoring Parkinson Disease from speech articulation kinematics

Anglický název

Monitoring Parkinson Disease from speech articulation kinematics

Jazyk

en

Originální abstrakt

Parkinson Disease (PD) is a neuromotor illness affecting general movements of different muscles, those implied in speech production being among them. The relevance of speech in monitoring illness progression has been documented in these last two decades. Most of the studies have concentrated in dysarthria and dysphonia induced by the syndrome. The present work is devoted to explore how PD affects the dynamic behavior of the speech neuromotor biomechanics (neuromechanics) involved in deficient articulation (dysarthria), in contrast to classical measurements based on static features as extreme and central vowel triangle positions. A statistical distribution of the kinematic velocity of the lower jaw and tongue is introduced, which presents interesting properties regarding pattern recognition and classification. This function may be used to establish distances between different articulation profiles in terms of information theory. Results show that these distances are correlated with a set of tests currently used by neurologists in PD progress evaluation, and could be used in elaborating new speech testing protocols.

Anglický abstrakt

Parkinson Disease (PD) is a neuromotor illness affecting general movements of different muscles, those implied in speech production being among them. The relevance of speech in monitoring illness progression has been documented in these last two decades. Most of the studies have concentrated in dysarthria and dysphonia induced by the syndrome. The present work is devoted to explore how PD affects the dynamic behavior of the speech neuromotor biomechanics (neuromechanics) involved in deficient articulation (dysarthria), in contrast to classical measurements based on static features as extreme and central vowel triangle positions. A statistical distribution of the kinematic velocity of the lower jaw and tongue is introduced, which presents interesting properties regarding pattern recognition and classification. This function may be used to establish distances between different articulation profiles in terms of information theory. Results show that these distances are correlated with a set of tests currently used by neurologists in PD progress evaluation, and could be used in elaborating new speech testing protocols.

BibTex


@article{BUT142890,
  author="Pedro {Gomez-Vilda} and Jiří {Mekyska} and Andrés {Gómez-Rodellar} and Daniel {Palacios-Alonso} and María Victoria {Rodellar Biarge} and Agustín {Álvarez-Marquina}",
  title="Monitoring Parkinson Disease from speech articulation kinematics",
  annote="Parkinson Disease (PD) is a neuromotor illness affecting general movements of different muscles, those implied in speech production being among them. The relevance of speech in monitoring illness progression has been documented in these last two decades. Most of the studies have concentrated in dysarthria and dysphonia induced by the syndrome. The present work is devoted to explore how PD affects the dynamic behavior of the speech neuromotor biomechanics (neuromechanics) involved in deficient articulation (dysarthria), in contrast to classical measurements based on static features as extreme and central vowel triangle positions. A statistical distribution of the kinematic velocity of the lower jaw and tongue is introduced, which presents interesting properties regarding pattern recognition and classification. This function may be used to establish distances between different articulation profiles in terms of information theory. Results show that these distances are correlated with a set of tests currently used by neurologists in PD progress evaluation, and could be used in elaborating new speech testing protocols.",
  chapter="142890",
  doi="10.3989/loquens.2017.036",
  howpublished="online",
  number="1",
  volume="4",
  year="2017",
  month="december",
  pages="1--12",
  type="journal article"
}