Optimization of Near-Infrared Spectroscopy Signal Derived from Muscle Tissue
Department
Biomedical Engineering
Faculty Advisor
Wei-Chiang Lin
Start Date
30-9-2020 2:00 PM
End Date
30-9-2020 3:00 PM
Abstract
Current approaches for controlling prosthesis include myoelectric control using electromyography (EMG) and EEG based brain-computer interfaces. In this project, we will explore the feasibility of prosthesis control using muscle hemodynamics. To measure muscle hemodynamics near infrared spectroscopy (NIRS) will be used to monitor local optical changes in muscle tissue in response to muscle activity. This method presents benefits such as its portability, noninvasiveness, and the lack of effect of electromagnetic noise. NIRS may also have the capability to detect partial activation of muscle and hence provide more independent signals for control purposes. A prototype mini NIR spectroscopy system was designed and developed. It contains a high power NIR light emitting diodes (LEDs) and a chip-based spectral sensor (AS7263, AMS, Austria). A special case was designed to control the distance between the NIR LEDs and the detector, which has been shown to influence the resulting signal measurements. An ideal distance between the NIRS LED and detector ensures that the muscle tissue contributes significantly to the measured signal. A Monte Carlo simulation for photon migration was performed to theoretically evaluate the performance of the prototype mini NIR spectroscopy system. The results of the Monte Carlo simulation indicated an inverse relationship between source-detector separation and diffuse reflectance, as well as an inverse relationship between subcutaneous fat layer thickness and diffuse reflectance signal derived from muscle. The simulation results also suggested that an approximate 2cm source-detector separation provides the optimal signal strength and signal derivation from the muscle tissue. These theoretical findings are the basis for recent adaptations to the case design, which has been altered to effectively contain the light source and NIRS detector. The adapted case will be used in an experimental procedure to obtain NIRS signals from predetermined areas such as the lower forearm and thigh. Using the chip-based spectral sensor, light source, and code written in MATLAB, we aim to obtain a NIRS signal from the muscle. We anticipate a measurable distinction between signal derived from the relaxed muscle and signal derived from the activated muscle. Ultimately, this signal will be used investigate NIRS applications for prosthesis control.
File Type
Event
Optimization of Near-Infrared Spectroscopy Signal Derived from Muscle Tissue
Current approaches for controlling prosthesis include myoelectric control using electromyography (EMG) and EEG based brain-computer interfaces. In this project, we will explore the feasibility of prosthesis control using muscle hemodynamics. To measure muscle hemodynamics near infrared spectroscopy (NIRS) will be used to monitor local optical changes in muscle tissue in response to muscle activity. This method presents benefits such as its portability, noninvasiveness, and the lack of effect of electromagnetic noise. NIRS may also have the capability to detect partial activation of muscle and hence provide more independent signals for control purposes. A prototype mini NIR spectroscopy system was designed and developed. It contains a high power NIR light emitting diodes (LEDs) and a chip-based spectral sensor (AS7263, AMS, Austria). A special case was designed to control the distance between the NIR LEDs and the detector, which has been shown to influence the resulting signal measurements. An ideal distance between the NIRS LED and detector ensures that the muscle tissue contributes significantly to the measured signal. A Monte Carlo simulation for photon migration was performed to theoretically evaluate the performance of the prototype mini NIR spectroscopy system. The results of the Monte Carlo simulation indicated an inverse relationship between source-detector separation and diffuse reflectance, as well as an inverse relationship between subcutaneous fat layer thickness and diffuse reflectance signal derived from muscle. The simulation results also suggested that an approximate 2cm source-detector separation provides the optimal signal strength and signal derivation from the muscle tissue. These theoretical findings are the basis for recent adaptations to the case design, which has been altered to effectively contain the light source and NIRS detector. The adapted case will be used in an experimental procedure to obtain NIRS signals from predetermined areas such as the lower forearm and thigh. Using the chip-based spectral sensor, light source, and code written in MATLAB, we aim to obtain a NIRS signal from the muscle. We anticipate a measurable distinction between signal derived from the relaxed muscle and signal derived from the activated muscle. Ultimately, this signal will be used investigate NIRS applications for prosthesis control.