TMS Motor Threshold (MT) Estimation App

Stochastic Approximator of MT (SAMT)

Initialize procedure

Enter final TMS intensity for motor hot spot determination % MSO

and move to next section.

Settings, instructions, and method descriptions.

Motor threshold estimate

Start procedure first.


Estimation history

or data

Back to top.

image

Back to top.

Information about this app

Settings

If estimation history is not saved, automatically trigger download when (1) the procedure ends after pulse No. 29 or is reset earlier, or (2) the page is closed or refreshed:

Back to top.

Instructions and method description

For fast estimation, start from an intensity that is close to the likely motor threshold (MT) and preferably suprathreshold (i.e., an intensity that produces motor responses more often than not). The intensity used to locate the motor hot spot is a good starting point.

The MT is estimated using a stochastic approximation method with a digital control sequence whose step size follows harmonic convergence with adaptive stepping (DCS-HA) and an initial step size of 4.2% MSO. For adaptive stepping, the control sequence is adjusted only when the response changes from suprathreshold to subthreshold or vice versa. With a good initial intensity, the estimator takes on average 20 pulses to determine the MT with median relative error |⁠δ⁠| of less than 1.5%.

Note: The fixed stepping method DCS-H, which was available up to version 1.7.4, is now deprecated, because adaptive stepping is more robust with respect to subject characteristics, target muscle, starting TMS amplitude, as well as TMS device and coil.

Known limitations and bugs:

  • The tool is provided as is. In particular, the description of the quality of the MT estimation is based on an approximate algorithm which may prematurely declare the estimation “good”. The final judgement on the MT must always be made by the operator!
  • The stochastic approximation algorithm may attempt to step to impossible intensities for the MT estimate (below 0% or above 100% MSO). The implementation in this app limits the next intensity between 1% and 100% MSO and the MT estimate between 0% and 101% MSO.
  • The copy button does not work on some mobile devices. Please copy the text manually if the copy button fails.

All processing is done by the app locally. No data are shared online.

This HTML/JS application is written by Boshuo Wang and Lari M. Koponen, with contribution from Vedarsh U. Shah, Shivum Vaishnavi, and Yiwen Zhang.

If you have questions or feedback, please email Boshuo Wang at boshuo.wang@duke.edu.

Back to top.

References

The stochastic approximation MT methods are described in the following manuscript, which explores the performance of different control sequences, stepping adaptiveness, and initial step size:

  • B. Wang, A. V. Peterchev, and S. M. Goetz, “Three Novel Methods for Determining Motor Threshold with Transcranial Magnetic Stimulation Outperform Conventional Procedures”, J. Neural Eng., vol. 20, no. 5, 056002, Oct. 2023.
    DOI: 10.1088/1741-2552/acf1cc.

The choice of initial stimulation intensity close to the MT and the description of the quality of the current estimation follow suggestions in:

  • L. M. Koponen and A. V. Peterchev, “Preventing misestimation of transcranial magnetic stimulation motor threshold with MTAT 2.0”, Brain Stim., vol. 15, no. 5, pp. 1073–6, Sep. 2022.
    DOI: 10.1016/j.brs.2022.07.057.

Validation of SAMT in clinical trials is described here:

  • B. Wang, V. U. Shah, L. M. Koponen, S. M. Goetz, A. D. Neacsiu, J. Y. Choi, Z. J. Daskalakis, P. B. Fitzgerald, L. G. Appelbaum, I. Hadas, H. Daniels, K. Rodrigues, E. S. Gotsis, N. W. Bailey, J. Raveendran, A. T. Gallo, S. Brinley, and A. V. Peterchev, “Stochastic Approximator of Motor Threshold for Transcranial Magnetic Stimulation (SAMT): Performance in Clinical Trials”, 2024 NYC Neuromod. Conf., New York, NY, USA, July 31–Aug. 4, 2024. Neuromodec J., vol. 4, no. 2, Nov. 2024.
    DOI: 10.31641/nmj-DIHO5587.

Back to top.

© 2022, Boshuo Wang, Lari M. Koponen, Stefan M. Goetz, Angel V. Peterchev, at Duke University, the University of Birmingham. All rights reserved.

The copyrights of this software are owned by Duke University and the University of Birmingham. As such, two licenses for this software are offered:

  1. An open-source license under the GPLv2 license for non-commercial use.
  2. A custom license with Duke University, for commercial use without the GPLv2 license restrictions.

As a recipient of this software, you may choose which license to receive the code under. Outside contributions to the Duke-owned code base cannot be accepted unless the contributor transfers the copyright to those changes over to Duke University. To enter a custom license agreement without the GPLv2 license restrictions, please contact the Digital Innovations Department at the Duke University Office for Translation & Commercialization (OTC) (otc.duke.edu/digital-innovations) at otcquestions@duke.edu with reference to “OTC File No. 8063” in your email.

Please note that this software is distributed AS IS, WITHOUT ANY WARRANTY; and without the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Back to top.

Number of visits: