Study of Issues in the Development of Electronically Operated Prosthetic Arm

Abstract

The arms and legs are very important parts of the human body. The arm physiology is an extremely complex to study. People’s working capability is badly affected when they suffer an amputated arm. Artificial replacements with prosthetic devices to get a satisfactory level of performance for essential functions with the currently available prosthetic technology are very difficult. The natural solution to amputation is biological regeneration of the missing limb. As this is not currently possible it is important to find viable solutions for providing multifunctional prostheses. Realization of a successful prosthesis is the need with complete understanding of the associated practical, clinical and economic factors. Presently, myoelectric arm prostheses are becoming popular because they are operated by a natural contraction of intact muscles. With the requirement of such multifunctional low cost prosthesis, following objectives were defined for present work- 1. Study of various methodologies and algorithms for SEMG signal analysis to control myoelectric arm 2. A comparative study on the analysis of SEMG signal based on different postures/movements of arm. 3. To develop more powerful, flexible, and efficient SEMG signal interpretation with the locations of electrodes on different acupressure points of human arm i.e. from shoulder to wrist. 4. Development of a control methodology to use myoelectric signals from muscles which act in synergy with hand function as the control signals to operate the prosthesis. 5. On the basis of above studies, development of a prototype model for realization of control methodology for elbow movement. Surface Electromyogram (SEMG) is a complex signal, which is controlled by the nervous system and is dependent on the anatomical and physiological properties of muscles. SEMG detector (electrodes) at the surface of the skin collects signals from different motor units at a time, which may generate interaction of different signals. These SEMG signal gets corrupted by noise while traveling through different tissues. The sources of these noises are environmental (such as 50 Hz power-line) or biological iv (such as motion artifact) interference. Due to these complexities, detection of SEMG signals with powerful and advance techniques is becoming a very important requirement in biomedical engineering. In the present work SEMG acquisition system was designed keeping the economics and application in mind. The amplifier provides the input for the analog interface of the computer using sound card (which is default adapter for softscope). The SEMG acquisition system was developed to measure and record the signals from the subjects. Users have the options to select a variety of voltage scales and time scales to display the output signals with great advantage of portability. After acquiring SEMG signal, important aspect is to extract the features which are representative of SEMG pattern for different arm functions. The study of different parameters in relation to SEMG variations with force level has been carried out and it has been found that both amplitude-related parameters and selected statistical parameters give good view of force level. A function slope has been tried for as SEMG parameter and it produced good representation of the SEMG. It has also been observed that prolonged use of the prosthetic device may result in lower amplitudes of SEMG signals. Thus, a prosthetic controller based on amplitude should have the capability of adjusting the threshold level in order to get proper degree of controllability. Acupressure points are places on the skin that are especially sensitive to bioelectrical impulses in the body and conduct those impulses readily. Traditionally, Asian cultures conceived of the points as junctures of special pathways that carried the human energy that the Chinese call chi and the Japanese call ki. Stimulating these points with pressure, needles, or heat triggers the release of endorphins, which are the neurochemicals that relieve pain and relaxes muscles. In tune with objective 3, as an additional study to find electrode locations to have multifunctional prosthesis, acupressure points were considered. The study explored that those electrode sites may be considered where even very weak correlation between the SEMG and (more than one) limb functions exist. This study verifies the SEMG activities on acupressure points if selected sensibly which encourage advancing the work for specific movement measurement. Further, the pressure points are compared with other locations on arm for SEMG observations. In the next stage of this work, the interpretation of SEMG signals from the locations of human arm i.e. from shoulder to wrist was done to discriminate hand/wrist v movements. The objective of the present study was to develop such a system so as to assess different muscles activities on arm during four basic movements namely opening (op) / closing (cl) / down (d) / up (u) of hand wrist/grip and find out the suitable number of channels for movement classification with best electrode locations. Six locations including two pressure points were considered. Principal component analysis was used to determine the best location with different sets of electrode numbers. After the principal component analysis of data from different locations, it is found that in two channel locations namely “cbet (below elbow thumb side) and cbeb (below elbow little finger side)” and in three channel locations namely “cam (below arm middle palm side)”, “camb (on opposite side of cam)” and “cbet (below elbow thumb side)” were best suited to discriminate the four chosen movements. However, the pressure points showed low performance with respect to other locations. Best location were related to the maximum loading percentage to PC1 and appropriate numbers of channel of the acquisition system to monitor movements again depended upon the loading percentage to PC1 but shall vary within i.e. different movement shall have different contribution percentage for better discrimination. Finally, the SEMG analysis on the elbow movement for the four movements [extension, flexion, supination and pronation] is discussed which results in control methodology for single-channel and two-channel prototype elbow. A prototype elbow based on SEMG analysis for the above elbow had been developed using microcontroller based hardware. The design employs SEMG-RMS to DC conversion using AD536 IC which gave satisfactory results. With proper synchronizing of the prototype arm position and the SEMG contraction the specific sequence of movements are achieved successfully. The prototype has been realized differently for two movement control (extension/flexion) and four movement control (extension/flexion/pronation/supination). This work presents a successful creation of inexpensive SEMG platform. A principal component analysis of data from different locations to discriminate the four chosen movements was usefully used.