Electromyography (EMG) Signal Processing to Evaluate Low-Frequency Tremors.

Objective quantification of tremor remains a challenge in Parkinson's disease (PD) assessment, with current clinical assessments relying largely on subjective scale ratings. This study evaluates the feasibility and signal behaviour of integrating surface electromyography (sEMG) with MDS-UPDRS-aligned tasks in a healthy adult cohort, with the aim of establishing normative low-frequency muscle activation profiles. Thirty-two healthy participants (mean age 27.6 ± 5.3 years) completed seven upper-limb tasks derived from the MDS-UPDRS while sEMG data were recorded from antagonistic forearm muscles. Signals were normalised using maximum voluntary contraction, filtered at 14 Hz, and analysed using frequency-domain (FFT) and time-frequency (STFT) methods. Significant task-dependent differences were observed in both frequency occurrence and magnitude (p < 0.05), particularly within the 3.5-9 Hz range. Finger tapping elicited increased low-frequency activity compared to baseline, while pronation-supination produced the most stable and consistent muscle activation across participants. Frequencies above 12 Hz showed minimal task discrimination. These findings demonstrate that low-frequency tremor-like activity can occur during specific MDS-UPDRS tasks in healthy individuals and may require further validation before being considered suitable for PD staging. This work establishes normative sEMG benchmarks to support future clinical validation and PD cohort comparisons.