Using AR model to analyze injured nerve with needle EMG signal.
- Author:
Chuan QIN
1
;
Zhizhong WANG
;
Gang WANG
;
Bo MA
Author Information
1. Department of Biomedical Engineering, Shanghai Jiaotong University, Shanghai 200030, China.
- Publication Type:Journal Article
- MeSH:
Action Potentials;
physiology;
Electromyography;
instrumentation;
methods;
Electrophysiology;
Humans;
Models, Neurological;
Models, Theoretical;
Muscle Contraction;
physiology;
Muscle, Skeletal;
innervation;
physiology;
Needles;
Neural Networks (Computer);
Pattern Recognition, Automated;
Peripheral Nerve Injuries;
Signal Processing, Computer-Assisted;
instrumentation
- From:
Journal of Biomedical Engineering
2004;21(4):636-639
- CountryChina
- Language:Chinese
-
Abstract:
The two main factors to affect the style of the recruitment are temporal recruitment and spatial recruitment. This study sought a new way to analyze the recruitment with the modern spectrum method. The abnormal spatial recruitment and temporal recruitment of varied injury degrees of intramuscular neuron were compared through the AR model. At last, AR coefficients were extracted and passed through BP artificial neuron network to classify different NEMG signals and good result was gained.