A Comparative Study of the Efficiency of the Methods Based on Artificial Intelligence Techniques to Estimate the Joint Angles of the Arm Using Surface Electromyogram Signal Processing

Research areas:

• Volume 3,Issue 4, Jan. 2015

Year:

2015

Type of Publication:

Article

Keywords:

Joint Arm, Electromyogram, Signal Processing, Artificial Intelligence

Authors:

Mostafa Langarizade; Navid Moshtaghi Yazdani; Arezoo Yazdani Seqerloo

Journal:

IJAIM

Volume:

3

Number:

4

Pages:

231-238

Month:

Jan.-Feb.

ISSN:

2320-5121

Abstract:

Today, several studies are conducted in the field of artificial organs to somewhat retrieve the disabilities caused by the amputation of organs such as arms. It is important to estimate the angle of the elbow joint, in this context. This research aims to compare the effectiveness of various methods of artificial intelligence to estimate the joint angles of the arm. The surface electromyogram data of 10 men with a mean age of 28 years is used in this study and no skeletal abnormalities were examined in all cases, and the examined arm was the right arm in all individuals. The results indicated that neural network along with different structures are well able to estimate the joint angles with high accuracy, however this method depends on the subject and is required to be reset for each subject. In the proposed method of this study, the dependency to the subject was removed and an acceptable accuracy was obtained in estimating the joint angles, respectively. Accordingly, we can conclude that there is a reasonable ability to estimate the joint angles using artificial intelligence techniques, especially neural networks and these methods can be well used by conducting further researches in the field of artificial organs.

Full text: IJAIM-388_final.pdf