Research
 

Surface-EMG


The aim of the project is to improve methods for analyzing surface-EMG and to extend the methodology so that it will work as a supplement to the needle-EMG. This project consists of, among others, "Spatial-temporal analysis of multichannel surface EMG".

Several studies suggested that the number of people who suffer from neuro-muscular disease have been increased in the last decades. The treatment of these people has required a substantial amount of financial supports. The cost for disturbances in the muscular skeletal system was calculated to be in the order of 17 billions SEK per annum (Swedish National Board for Health and Welfare). Clinically there is a long tradition of EMG-recording using surface electrodes in work in physiology, ergonomy and rehabilitation medicine. Recording of surface-EMG has several conclusive advantages, e.g.: the method is non-invasive, it is possible to record in most situation, the total activity of the muscle is reflected, etc.

The aim of the project is to improve methods for analyzing surface EMG signals. Existing methods use Fourier-transformation, which are based on assumptions that usually are not satisfied in this kind of application. This implies that the the existing methods for frequency analysis of EMG signals can only study lapse with a certain length at acceptable frequency resolution. In other words, there are substantial difficulties in analyzing muscle fatigue during short contraction cycle. There is a great need to be able to study these relatively quick but natural movements (such as walking) both in working life and in the clinical work with patients. The methods we are going to apply in the present project will meet such needs, and the information in the registered surface EMG signals will be utilized in a much better way, especially for the new type of multichannel electrode which contains even spatial information. Our research has partly obtained techniques (e.g. wavelets). These methods will be further developed to be more clinical applicable.