Comparing electromyography, accelerometry, and visual inspection to assess the resting motor threshold for transcranial magnetic stimulation

Introduction : Electromyography (EMG) remains the gold standard for estimating the Resting Motor Threshold (RMT) in Transcranial Magnetic Stimulation (TMS) studies, but its cost and limited accessibility often lead researchers to use visual inspection (VIS). However, VIS may introduce variability and systematic bias. Accelerometry (ACC) offers a cost-effective, objective alternative to capture TMS-evoked responses. Objective : To compare the RMT as estimated using EMG, ACC, and VIS. Methods : Five participants underwent TMS while EMG, ACC, and video recordings were collected. Separately, 64 observers judged hand movement in videos to estimate RMT via VIS. RMTs were compared across the three methods using Bayesian model comparison, Bland-Altman analyses, and Intraclass Correlation Coefficients (ICCs). Results : RMTs estimated via EMG were lower than those obtained using either ACC or VIS. Compared to EMG, VIS tended to overestimate RMT (mean bias = 5.23%, 95%CI = [1.00–11.00]), while ACC and VIS estimates were more closely aligned (mean bias = 0.43%, 95%CI = [–4.00 – 5.00]). ICC (2,1) values indicated moderate reliability for VIS vs EMG (mean = 0.580, 95%CI = [0.389 – 0.748]), and good-to-excellent reliability for VIS vs ACC (mean = 0.845). However, bootstrapped 95% confidence intervals identified significant variability in the estimates provided by visual inspection, ranging from +1 to +11 for VIS vs EMG, but as low as -4 to +5 for VIS vs ACC. Conclusions : EMG remains the most sensitive technique for estimating the RMT, but when EMG is not feasible, accelerometery provides a quantifiable, more objective, and less variable alternative than visual inspection.