Tu-SMD-03 - Fatigue Detection During Wall-Sit Exercise Using Statistical Features of Chest ECG-Derived EMG

Event

EUROSENSORS 2026
2026-09-06 - 2026-09-09
Zurich

Band

Poster

Chapter

Sensing For Medical Diagnostics And Monitoring

Author(s)

C. Park, C. Huh, H. J. Park - Electronics and Telecommunications Research Institute (ETRI),Daejeon (Republic of Korea)

Pages

608 - 609

DOI

10.5162/eurosensors2026/Tu-SMD-03

ISBN

978-3-910600-12-6

Price

free

Abstract

Muscle fatigue monitoring during isometric exercise can support safer training and rehabilitation, but conventional limb sEMG setups can be inconvenient in daily use. This work investigates four-class fatigue detection from the myoelectric component extracted from chest ECG during wall-sit exercise. Six healthy adults performed wall-sit until task failure, and the signal was segmented into non-overlapping 1 s windows labeled as rest, low fatigue, moderate fatigue, and high fatigue. Standard deviation, skewness, and kurtosis were used to train a bagging tree classifier. The proposed approach achieved an overall accuracy of 67.5% in 10-fold cross-validation and demonstrates that simple distribution-based features from chest ECG-derived EMG can support low-complexity fatigue sensing.