TMS-evoked responses are driven by recurrent large-scale network dynamics
Curation statements for this article:-
Curated by eLife
eLife assessment
This manuscript presents a valuable new approach to modelling patterns of brain activity evoked by non-invasive brain stimulation, shedding light on how such stimulation drives neuronal dynamics. The performance of the model is impressive and its validity is supported by solid evidence. This work will be of interest to researchers working in computational neuroscience, neuroimaging, and non-invasive brain stimulation.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
A compelling way to disentangle the complexity of the brain is to measure the effects of spatially and temporally synchronized systematic perturbations. In humans, this can be non-invasively achieved by combining transcranial magnetic stimulation (TMS) and electroencephalography (EEG). Spatiotemporally complex and long-lasting TMS-EEG evoked potential (TEP) waveforms are believed to result from recurrent, re-entrant activity that propagates broadly across multiple cortical and subcortical regions, dispersing from and later re-converging on, the primary stimulation site. However, if we loosely understand the TEP of a TMS-stimulated region as the impulse response function of a noisy underdamped harmonic oscillator, then multiple later activity components (waveform peaks) should be expected even for an isolated network node in the complete absence of recurrent inputs. Thus emerges a critically important question for basic and clinical research on human brain dynamics: what parts of the TEP are due to purely local dynamics, what parts are due to reverberant, re-entrant network activity, and how can we distinguish between the two? To disentangle this, we used source-localized TMS-EEG analyses and whole-brain connectome-based computational modelling. Results indicated that recurrent network feedback begins to drive TEP responses from 100 ms post-stimulation, with earlier TEP components being attributable to local reverberatory activity within the stimulated region. Subject-specific estimation of neurophysiological parameters additionally indicated an important role for inhibitory GABAergic neural populations in scaling cortical excitability levels, as reflected in TEP waveform characteristics. The novel discoveries and new software technologies introduced here should be of broad utility in basic and clinical neuroscience research.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
The authors succeeded in fitting their Jansen-Rit model parameters to accurately reproduce individual TEPs. This is a major success already and the first study of this kind to the best of my knowledge. Then the authors make use of this fitted model to introduce virtual lesions in specific time windows after stimulation to analyze which of the response waveforms are local and which come from recurrent circles inside the network. The methodological steps are nicely explained. The authors use a novel parameter fitting method that proves very successful. They use completely openly available data sets and publish their code in a manner that makes reproduction easy. I really enjoyed reading this paper and suspect its methodology to set a new landmark in the field of brain stimulation simulation. …
Author Response
Reviewer #1 (Public Review):
The authors succeeded in fitting their Jansen-Rit model parameters to accurately reproduce individual TEPs. This is a major success already and the first study of this kind to the best of my knowledge. Then the authors make use of this fitted model to introduce virtual lesions in specific time windows after stimulation to analyze which of the response waveforms are local and which come from recurrent circles inside the network. The methodological steps are nicely explained. The authors use a novel parameter fitting method that proves very successful. They use completely openly available data sets and publish their code in a manner that makes reproduction easy. I really enjoyed reading this paper and suspect its methodology to set a new landmark in the field of brain stimulation simulation. The conclusions of the authors are well supported by their results, however, some analysis steps should be clarified, which are specified in the essential revisions.
We are delighted and flattered by the Reviewer’s positive evaluation of our work, and appreciation of our efforts to maximize its reproducibility. We wish also to thank the Reviewer for their compelling and interesting points, which we have addressed in full, and we believe further enhance the quality of the paper. Thanks again!
Reviewer #2 (Public Review):
Here the authors tackle the problem of identifying which parts of a TMS-evoked response are local to the stimulation site versus driven by reverberant activity from other regions. To do this they use a dataset of EEG recorded simultaneously with TMS pulses, and examine virtual lesions of a network of neural masses fitted to the data. The fitting uses a very recent model inversion method developed by the authors, able to fit time series directly rather than just summary statistics thereof. And it apparently works rather well indeed, at least after the first ~50 ms post-stimulus. I expect many readers will be keen to try this fitting method in their own work.
We are delighted by the Reviewer’s appreciation and consideration of our paper. We have addressed the comments and revisions requested following the flow suggested by the Reviewer’s comments. We would take this opportunity to kindly thank the Reviewer for his/her contribution and for helping us to improve the manuscript.
Reviewer #3 (Public Review):
The manuscript is very well written and the graphics are quite iconic. Moreover, the hypothesis is clear and the rationale is very convincing. Overall, the paper has the potential of being of paramount importance for the TMS-EEG community because it provides a valuable tool for a proper interpretation of several previously published TMS-EEG results.
Unfortunately, in my opinion, the dataset used to train and validate the method does not support the implication and interpretation of the results. Indeed, as clearly visible from most of the figures and mentioned by the authors of the database, the data contains residual sensory artifacts (auditory or somatosensory) that can completely bias the authors' interpretation of the re-entrant activity.
We are most grateful to the Reviewer for their positive evaluation of our manuscript. We also sincerely appreciate all the comments and suggestions raised, and for contributing their evident expertise with TMS-EEG data towards the constructive improvement of this research. We hope the Reviewer will appreciate our efforts made to address their excellent points, and are pleased with the resultant strengthening of the paper.
-
eLife assessment
This manuscript presents a valuable new approach to modelling patterns of brain activity evoked by non-invasive brain stimulation, shedding light on how such stimulation drives neuronal dynamics. The performance of the model is impressive and its validity is supported by solid evidence. This work will be of interest to researchers working in computational neuroscience, neuroimaging, and non-invasive brain stimulation.
-
Reviewer #1 (Public Review):
The authors succeeded in fitting their Jansen-Rit model parameters to accurately reproduce individual TEPs. This is a major success already and the first study of this kind to the best of my knowledge. Then the authors make use of this fitted model to introduce virtual lesions in specific time windows after stimulation to analyze which of the response waveforms are local and which come from recurrent circles inside the network. The methodological steps are nicely explained. The authors use a novel parameter fitting method that proves very successful. They use completely openly available data sets and publish their code in a manner that makes reproduction easy. I really enjoyed reading this paper and suspect its methodology to set a new landmark in the field of brain stimulation simulation. The conclusions of …
Reviewer #1 (Public Review):
The authors succeeded in fitting their Jansen-Rit model parameters to accurately reproduce individual TEPs. This is a major success already and the first study of this kind to the best of my knowledge. Then the authors make use of this fitted model to introduce virtual lesions in specific time windows after stimulation to analyze which of the response waveforms are local and which come from recurrent circles inside the network. The methodological steps are nicely explained. The authors use a novel parameter fitting method that proves very successful. They use completely openly available data sets and publish their code in a manner that makes reproduction easy. I really enjoyed reading this paper and suspect its methodology to set a new landmark in the field of brain stimulation simulation. The conclusions of the authors are well supported by their results, however, some analysis steps should be clarified, which are specified in the essential revisions.
-
Reviewer #2 (Public Review):
Here the authors tackle the problem of identifying which parts of a TMS-evoked response are local to the stimulation site versus driven by reverberant activity from other regions. To do this they use a dataset of EEG recorded simultaneously with TMS pulses, and examine virtual lesions of a network of neural masses fitted to the data. The fitting uses a very recent model inversion method developed by the authors, able to fit time series directly rather than just summary statistics thereof. And it apparently works rather well indeed, at least after the first ~50 ms post-stimulus. I expect many readers will be keen to try this fitting method in their own work.
-
Reviewer #3 (Public Review):
The manuscript is very well written and the graphics are quite iconic. Moreover, the hypothesis is clear and the rationale is very convincing. Overall, the paper has the potential of being of paramount importance for the TMS-EEG community because it provides a valuable tool for a proper interpretation of several previously published TMS-EEG results.
Unfortunately, in my opinion, the dataset used to train and validate the method does not support the implication and interpretation of the results. Indeed, as clearly visible from most of the figures and mentioned by the authors of the database, the data contains residual sensory artefacts (auditory or somatosensory) that can completely bias the authors' interpretation of the re-entrant activity.
-
