1. Cyclostationary Modeling of Surface Electromyography Signal During Gait Cycles and Its Application for Cerebral Palsy Diagnosis
- Author
-
Mengjie Chen, Li Yan, Liang Yu, and Liangchao Dong
- Subjects
030222 orthopedics ,Signal processing ,Multidisciplinary ,medicine.diagnostic_test ,Computer science ,business.industry ,Cyclostationary process ,0206 medical engineering ,Spectral correlation density ,Pattern recognition ,02 engineering and technology ,Electromyography ,medicine.disease ,020601 biomedical engineering ,Signal ,Cerebral palsy ,03 medical and health sciences ,0302 clinical medicine ,Gait (human) ,Gait analysis ,medicine ,Artificial intelligence ,business - Abstract
Cerebral palsy (CP) is a group of permanent movement disorders that appear in early childhood. The electromyography (EMG) signal analysis and the gait analysis are two most commonly used methods in the clinic. In this paper, a cyclostationary model of the EMG signal is proposed. The model can combine the aforementioned two methods. The EMG signal acquired during the gait cycles is assumed to be cyclostationary due to the physiological characteristics of the EMG signal production. Then, the spectral correlation density is used to analyze the cyclic frequency (corresponding to the gait cycles) and spectral frequency (the frequency of EMG signal) in a waterfall representation of the two kinds of frequencies. The experiments show that the asymptomatic (normal) subjects and symptomatic subjects (with CP) can be distinguished from the spectral correlation density in a range of cyclic frequencies.
- Published
- 2018