Academic Journals Database
Disseminating quality controlled scientific knowledge

A Comparison of the SOFM with LVQ, SOFM without LVQ and Statistical Technique

Author(s): Anjana Bhardwaj | Manish | A.K. Arora

Journal: International Journal of Engineering and Advanced Technology
ISSN 2249-8958

Volume: 1;
Issue: 6;
Start page: 40;
Date: 2012;
VIEW PDF   PDF DOWNLOAD PDF   Download PDF Original page

Keywords: Artificial Neural Network | Electromyography | learning vector quantization | Motor unit Action Potentials | Selforganizing feature maps.

Abstract – The shapes and firing rates of MUAP’s (motor unitaction potentials) in an EMG (electromyographic) signal provide animportant source of information for the diagnosis of neuromusculardisorders. In order to extract this information from EMG signalsrecorded at low to moderate force levels, it is required: i) to identifythe MUAP’s composing the EMG signal, ii) to classify MUAP’swith similar shape. For the classification of MUAP’s two differentpattern recognition techniques are presented: i) An artificial neuralnetwork (ANN) technique based on unsupervised learning, using amodified version of the self-organizing feature maps (SOFM)algorithm and learning vector quantization (LVQ), and ii) Astatistical pattern recognition technique based on Euclideandistance. A total of 521 MUAP’s obtained from 2 normal subjects,4 subjects suffering from myopathy, and 5 subjects suffering frommotor neuron disease were analyzed. The success rate for the ANNtechnique was 97.6%, the success rate for SOFM technique was94.8%, and for statistical technique it was 95.3%. So SOFMtechnique along with LVQ is batter technology than the SOFMwithout LVQ technique and Statistical technique.
Save time & money - Smart Internet Solutions      Why do you need a reservation system?