Title: Data-Mining models for the Diagnosis of EMG-based Neuromuscular Diseases

Authors: Babita Pandey, R.B. Mishra

Addresses: Department of Computer Engineering, IT, Banaras Hindu University, Varanasi, UP 221005, India. ' Department of Computer Engineering, IT, Banaras Hindu University, Varanasi, UP 221005, India

Abstract: Data-Mining (DM) methods such as Decision Tree (DT), Artificial Neural Network (ANN) and Sensitivity Analysis (SA) have been extensively used in the medical and biomedical domains but these methods and their combinations have been rarely used in the diagnosis of Neuromuscular Diseases (NMDs). In this paper, we use various methods of DM such as: DT, ANN and SA and their combination. A comparative study is made on these combined methods to determine the methods with highest degree of accuracy. The results show that in combined SA–ANN models SA–ANN (RBFN) has highest accuracy. In SA–DT model and DT–ANN model, all the three methods, i.e., quick, dynamic and RBFN have equal accuracy.

Keywords: ANNs; artificial neural networks; decision trees; sensitivity analysis; NMDs; neuromuscular diseases; data mining; biomedical engineering; electromyography; EMG; neuromuscular disease diagnosis.

DOI: 10.1504/IJBET.2011.041118

International Journal of Biomedical Engineering and Technology, 2011 Vol.6 No.2, pp.109 - 128

Received: 08 May 2021
Accepted: 12 May 2021

Published online: 09 Jul 2011 *

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