Extraction of Sensory part of Ulnar Nerve Signal Using Blind Source Separation Method
Subject Areas : International Journal of Smart Electrical EngineeringAlireza Kashaninia 1 , S Nooreddin Jafari 2
1 - Assistant Professor in Electrical Engineering, Islamic Azad University, Central Tehran Branch (IAUCTB), Advanced Bionic Systems (ABS)
lab, Tehran, Iran.
2 - M. Sc Degree in Electrical Engineering, Department of electrical engineering, Islamic Azad University, Langaroud branch, Guilan, Iran
Keywords: PNS1, ENG2, surface electrode, Ulnar nerve signals, sensory signal, motor signal, BSS3, PCA4, ICA5, Correlation analysis,
Abstract :
A recorded nerve signal via an electrode is composed of many evokes or action potentials, (originated from individual axons) which may be considered as different initial sources. Recovering these primitive sources in its turn may lead us to the anatomic originations of a nerve signal which will give us outstanding foresights in neural rehabilitations. Accordingly, clinical interests may be raised on extraction of sensory and motor components of the nerve signals in neural injuries. One example is to extract sensory fraction in sacral nerve to sense the bladder filling up in paraplegic or quadriplegic people [3]. Blind Source Separation (BSS) methods seem good solutions for extraction of the initial sources which are contributing in recorded mixed sources. Considering the nerve signal as a superposition of many axonal or fascicular signals, we have encouraged to try BSS methods to see whether it can recover the sensory and motor sources of a recorded nerve signal. Accordingly, both PCA and ICA techniques were examined in a case study (human left arm), in which the response of the ADM muscle to the Ulnar nerve stimulation were recorded in two points. The corresponded sensory signal was recorded on the pinkie at the same time (all recordings were done via surface electrodes). It was shown that ICA (supremely better than PCA) was able to separate initial sources (ADM recorded signals) into two signals so that one of them was most similar to the sensory (Pinkie) signal. The level of similarity was quantified via correlation analysis. As the result, it is concluded that ICA is capable of extracting Sensory and Motor signals in PNS.