Comparison of 4 Methods to Estimate the Resting Motor Threshold Using Transcranial Magnetic Stimulation: Rossini-Rothwell, 2-Threshold, Adaptive Threshold Hunting, and Stimulus-Response Curve Approaches

Ashok Jammigumpula; Jithin T Joseph; Abhiram N Purohith; Sonia Shenoy; Samir Kumar Praharaj

doi:10.1097/YCT.0000000000001211

Comparison of 4 Methods to Estimate the Resting Motor Threshold Using Transcranial Magnetic Stimulation: Rossini-Rothwell, 2-Threshold, Adaptive Threshold Hunting, and Stimulus-Response Curve Approaches

J ECT. 2026 Apr 8. doi: 10.1097/YCT.0000000000001211. Online ahead of print.

Authors

Ashok Jammigumpula¹, Jithin T Joseph, Abhiram N Purohith, Sonia Shenoy, Samir Kumar Praharaj

Affiliation

¹ Clinical Research Centre for Neuromodulation in Psychiatry, Department of Psychiatry, Kasturba Medical College, Manipal, Manipal Academy of Higher Education, Manipal, Udupi, Karnataka, India.

PMID: 41949547
DOI: 10.1097/YCT.0000000000001211

Abstract

Background: Estimating the resting motor threshold (RMT) is crucial for both investigative and therapeutic transcranial magnetic stimulation. However, no gold standard exists, as each method involves trade-offs between accuracy, time, and ease of use. This study compared RMT estimates and trial requirements across 4 EMG-based methods-Rossini-Rothwell (RR), maximum likelihood parameter estimation by sequential testing (ML-PEST), Mills-Nithi (MN), supervised parametric estimation (SPE)-and 2 visual methods: RR visual and ML-PEST visual.

Methods: In this observational study, RMT was estimated in 20 healthy participants using 6 approaches: 4 EMG-based (RR, ML-PEST, MN, SPE) and 2 visual (RR and ML-PEST). We assessed agreement among EMG-based methods and compared RMT estimates and trial counts across all methods.

Results: EMG-based methods showed strong agreement (ICC: 0.97, 95% CI: 0.94-0.99). RMT estimates differed significantly (P=0.001), with RR Visual yielding the highest values-significantly higher than RR EMG, ML-PEST EMG, and SPE EMG methods. Trial counts also varied (P<0.001), with ML-PEST requiring the fewest. In subgroup analysis, RR Visual and ML-PEST Visual produced similar RMTs, but ML-PEST Visual needed fewer trials.

Conclusions: Adaptive threshold-hunting methods like ML-PEST offer efficient and accurate RMT estimation while reducing the number of required trials, supporting their use in both clinical and research applications.

Keywords: Mills-Nithi (MN) method; Rossini-Rothwell (RR) method; electromyography (EMG); maximum likelihood parameter estimation by sequential testing (ML-PEST); resting motor threshold (RMT); supervised parametric estimation (SPE); transcranial magnetic stimulation (TMS).