Support vector machine classification of Parkinson's disease and essential tremor subjects based on temporal fluctuation

Tremor is a common symptom shared in both Parkinson's disease (PD) and Essential tremor (ET) subjects. The differential diagnosis of PD and ET tremor is important since the realization of treatment depends on specific medication. A novel feature is developed based on a hypothesis that tremor of PD subject has a larger fluctuation during resting than action task. Tremor signal is collected using a triaxial gyroscope sensor attached to subject's finger during kinetic and resting task. The angular velocity signal is analyzed by transforming a one-dimensional to two-dimensional signal using a relation of signal and its delay versions. Tremor fluctuation is defined as the area of 95% confidence ellipse covering the two-dimensional signal. The tremor fluctuation during kinetic and resting task is used as classification features. The support vector machine is used as a classifier and tested with 10-fold cross-validation. This novel feature provides a perfect PD/ET classification with 100% accuracy, sensitivity and specificity.

[1]  J. M. Spyers-Ashby,et al.  Classification of normal and pathological tremors using a multidimensional electromagnetic system. , 1999, Medical engineering & physics.

[2]  A. Hossen,et al.  Discrimination of Parkinsonian tremor from essential tremor by implementation of a wavelet-based soft-decision technique on EMG and accelerometer signals , 2010, Biomed. Signal Process. Control..

[3]  Mika P. Tarvainen,et al.  Surface EMG and acceleration signals in Parkinson’s disease: feature extraction and cluster analysis , 2008, Medical & Biological Engineering & Computing.

[4]  Irene Litvan,et al.  Examining the Motor Phenotype of Patients with Both Essential Tremor and Parkinson's Disease , 2012, Tremor and other hyperkinetic movements.

[5]  Heather Wharrad,et al.  Distinguishing between physiological and essential tremor using discriminant and cluster analyses of parameters derived from the frequency spectrum , 2000 .

[6]  Jue Wang,et al.  Multi-features fusion diagnosis of tremor based on artificial neural network and D-S evidence theory , 2008, Signal Process..

[7]  Ronachai Pongthornseri,et al.  Low-cost, 3-dimension, office-based inertial sensors for automated tremor assessment: technical development and experimental verification. , 2014, Journal of Parkinson's disease.

[8]  P. Brown,et al.  Distinguishing the Central Drive to Tremor in Parkinson's Disease and Essential Tremor , 2015, The Journal of Neuroscience.

[9]  Mehmet Engin,et al.  The classification of human tremor signals using artificial neural network , 2007, Expert Syst. Appl..

[10]  Zelma H. T. Kiss,et al.  Smart watch accelerometry for analysis and diagnosis of tremor , 2014, Journal of Neuroscience Methods.

[11]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[12]  J. Jankovic,et al.  Exploring the relationship between essential tremor and Parkinson's disease. , 2007, Parkinsonism & related disorders.

[13]  Ronachai Pongthornseri,et al.  Temporal fluctuations of tremor signals from inertial sensor: a preliminary study in differentiating Parkinson’s disease from essential tremor , 2015, Biomedical engineering online.

[14]  D. Vaillancourt,et al.  Combined measures of movement and force variability distinguish Parkinson’s disease from essential tremor , 2011, Clinical Neurophysiology.

[15]  S M Rissanen,et al.  EMG signal morphology and kinematic parameters in essential tremor and Parkinson's disease patients. , 2014, Journal of electromyography and kinesiology : official journal of the International Society of Electrophysiological Kinesiology.

[16]  Li Li,et al.  Mini Review: linkages between essential tremor and Parkinson’s disease? , 2013, Front. Cell. Neurosci..

[17]  Martin K. Purvis,et al.  Parkinson's disease and essential tremor classification on mobile device , 2014, Pervasive Mob. Comput..