Date Approved


Graduate Degree Type


Degree Name

Engineering (M.S.E.)

Degree Program

School of Engineering

First Advisor

Dr. Nabeeh Kandalaft

Second Advisor

Dr. Samhita Rhodes

Third Advisor

Dr. Brian Krug

Fourth Advisor

Dr. Karl Brakora

Fifth Advisor

Dr. Gordon Alderink

Academic Year



This is an investigation of the use of surface electromyography (sEMG) as a tool to improve human interfacing devices (HID) information bandwidth through the transduction of the fingertip workspace. It combines the work of Merletti et al and Jarque-Bou et al to design an open-source framework for Fingertip Workspace based Human Interfacing Devices (HID). In this framework, the fingertip workspace is defined as the system of forearm and hand muscle force through a tensor which describes hand anthropometry. The thesis discusses the electrophysiology of muscle tissue along with the anatomy and physiology of the arm in pursuit of optimizing sensor location, muscle force measurements, and viable command gestures. Algorithms for correlating sEMG to hand joint angle are investigated using MATLAB for both static and moving gestures. Seven sEMG spots and Fingertip Joint Angles recorded by Jarque Bou et al are investigated for the application of sEMG to Human Interfacing Devices (HID). Such technology is termed Gesture Computer Interfacing (GCI) and has been shown feasible through devices such as CTRL Labs interface, and models such as those of Sartori, Merletti, and Zhao. Muscles under sEMG spots in this dataset and the actions related to them are discussed, along with what muscles and hand actions are not visible within this dataset. Viable gestures for detection algorithms are discussed based on the muscles discerned to be visible in the dataset through intensity, spectral moment, power spectra, and coherence. Detection and isolation of such viable actions is fundamental to designing an EMG driven musculoskeletal model of the hand needed to facilitate GCI. Enveloping, spectral moment, power spectrum, and coherence analysis are applied to a Sollerman Hand Function Test sEMG dataset of twenty-two subjects performing 26 activities of living to differentiate pinching and grasping tasks. Pinches and grasps were found to cause very different activation patterns in sEMG spot 3 relating to flexion of digits I - V. Spectral moment was found to be less correlated with differentiation and provided information about the degree of object manipulation performed and extent of fatigue during each task. Coherence was shown to increase between flexors and extensors with intensity of task but was found corrupted by crosstalk with increasing intensity of muscular activation. Some spectral results correlated between finger flexor and extensor power spectra showed anticipatory coherence between the muscle groups at the end of object manipulation. An sEMG amplification system capable of capturing HD-sEMG with a bandwidth of 300 and 500 Hz at a sampling frequency of 2 kHz was designed for future work. The system was designed in ordinance with current IEEE research on sensor-electrode characteristics. Furthermore, discussion of solutions to open issues in HD-sEMG is provided. This work did not implement the designed wristband but serves as a literature review and open-source design using commercially available technologies.

Included in

Engineering Commons