Experimental Investigations of EMG-Torque Modeling for the Human Upper Limb
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The electrical activity of skeletal muscle—the electromyogram (EMG)—is of value to many different application areas, including ergonomics, clinical biomechanics and prosthesis control. For many applications, the EMG is related to muscular tension, joint torque and/or applied forces. In these cases, a goal is for an EMG-torque model to emulate the natural relationship between the central nervous system (as evidenced in the surface EMG) and peripheral joints and muscles. This thesis work concentrated on experimental investigations of EMG-torque modeling. My contributions include: 1) continuing to evaluate the advantage of advanced EMG amplitude estimators, 2) studying system identification techniques (regularizing the least squares fit and increasing training data duration) to improve EMG-torque model performance, and 3) investigating the influence of joint angle on EMG-torque modeling. Results show that the advanced EMG amplitude estimator reduced the model error by 21%—71% compared to conventional estimators. Use of the regularized least squares fit with 52 seconds of training data reduced the model error by 20% compared to the least squares fit without regulation when using 26 seconds of training data. It is also demonstrated that the influence of joint angle can be modeled as a multiplicative factor in slowly force-varying and force-varying contractions at various, fixed angles. The performance of the models that account for the joint angle are not statistically different from a model that was trained at each angle separately and thus does not interpolate across angles. The EMG-torque models that account for joint angle and utilize advanced EMG amplitude estimation and system identification techniques achieved an error of 4.06±1.2% MVCF90 (i.e., error referenced to maximum voluntary contraction at 90° flexion), while models without using these advanced techniques and only accounting for a joint angle of 90° generated an error of 19.15±11.2% MVCF90. This thesis also summarizes other collaborative research contributions performed as part of this thesis. (1) EMG-force modeling at the finger tips was studied with the purpose of assessing the ability to determine two or more independent, continuous degrees of freedom of control from the muscles of the forearm [with WPI and Sherbrooke University]. (2) Investigation of EMG bandwidth requirements for whitening for real-time applications of EMG whitening techniques [with WPI colleagues]. (3) Investigation of the ability of surface EMG to estimate joint torque at future times [with WPI colleagues]. (4) Decomposition of needle EMG data was performed as part of a study to characterize motor unit behavior in patients with amyotrophic lateral sclerosis (ALS) [with Spaulding Rehabilitation Hospital, Boston, MA].
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it