Related Experiment Video
Updated: Oct 5, 2026
Using Facial Electromyography to Assess Facial Muscle Reactions to Experienced and Observed Affective Touch in Humans
Published on: March 15, 2019
Sex, age, and race representation in electromyography research: A cross-sectional methodological survey
Bingle Li1, Xinrui Tang2, Junbo Xu2
1Biomedical Engineering, School of Science and Engineering, University of Dundee, UK; College of Medicine and Biological Information Engineering, Northeastern University, China; School of Biomedical Engineering & State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, China.
Background:
Surface electromyography (sEMG) is widely utilized in biomechanics and human-computer interaction. Modern sEMG systems rely on machine learning models , whose generalization heavily depends on physiological variations linked to sex, age, and ancestry, though these are often unexplored.
Objective:
This survey critically appraises how demographic variables (sex, age, race) are reported in primary sEMG studies and quantifies cohort diversity at both aggregate and individual-study levels.
Methods:
We analyzed a stratified random sample of 377 sEMG articles (January 2000-April 2026) from PubMed and IEEE Xplore. A validated large-language-model workflow extracted demographic data to compute reporting rates, gender imbalance indices, and age spans.
Results:
Overall reporting rates were 87.5% for gender, 89.9% for age, and 2.6% for race. Despite macroscopic sex balance, 49.5% of individual studies exhibited severe gender imbalance. Furthermore, cohorts overwhelmingly concentrated on young adults (19-35 years) with narrow age bands, revealing pronounced homogenization.
Conclusion:
These findings underscore a critical demographic blind spot in current sEMG research. To ensure the safety, robustness, and true generalizability of intelligent neural interfaces across heterogeneous real-world populations, there is an urgent need to mandate structured demographic reporting standards and explicitly define the fairness boundaries of algorithm models.
