Title: Neural network based prediction of Parkinsonian hand tremor using surface electromyography

Authors: Sourav Chandra; Koushik Bakshi; Amit Konar; D.N. Tibarewala

Addresses: School of Bio Science & Engineering, Jadavpur University, Kolkata 700032, West Bengal, India ' School of Bio Science & Engineering, Jadavpur University, Kolkata 700032, West Bengal, India ' Department of Electronics & Telecommunication Engineering, Jadavpur University, Kolkata 700032, West Bengal, India ' School of Bio Science & Engineering, Jadavpur University, Kolkata 700032, West Bengal, India

Abstract: Tremor is unwanted movement of body part. Tremor model estimations are employed in intelligent tremor prediction and compensation scheme. One way of modelling Parkinsonian tremor is the analysis-synthesis approach, in which theoretically guessed properties of the individual parts are combined mathematically. However, this approach has limited success in predicting certain tremor characteristics. The second method uses different measures of the tremor (frequency, amplitude, effect of loading, etc.) in order to infer the underlying phenomena. Electromyogram (EMG) being a related physiological event, is included in the parameter domain, which helps to estimate certain tremor characteristics. An artificial neural network based study of tremor in Parkinson's disease and its compensation technique is discussed here.

Keywords: Parkinson's disease; hand tremor; sEMG; accelerometer; ANNs; artificial neural networks; RBF; radial basis function; surface EMG; electromyograms; tremor modelling; estimation; tremor prediction; compensation schemes; Parkinsonian tremor.

DOI: 10.1504/IJSISE.2012.050316

International Journal of Signal and Imaging Systems Engineering, 2012 Vol.5 No.4, pp.246 - 252

Received: 19 Feb 2011
Accepted: 23 Mar 2011
Published online: 31 Dec 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article