Answers are grounded in this paper and cite their sources.
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.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: