Portal de Periódicos da CAPES

An analytical approach to identify indirect multisensory cortical activations elicited by TMS?

2021; Elsevier BV; Volume: 14; Issue: 2 Linguagem: Inglês

10.1016/j.brs.2021.02.003

1935-861X

Eva Nießen, Martina Bracco, Tuomas P. Mutanen, Edwin M. Robertson,

EEG and Brain-Computer Interfaces

Electroencephalography (EEG) is widely used for detecting transcranial-magnetic stimulation (TMS) evoked responses in the human brain. TMS-evoked potentials (TEPs) may contain both direct activations due to the TMS-induced cortical electric field and indirect cortical activations due to the subsequent multisensory responses to TMS [[1]Ilmoniemi R.J. Kičić D. Methodology for combined TMS and EEG.Brain Topogr. 2009; 22: 233-248Crossref PubMed Scopus (204) Google Scholar]. Distinguishing between direct and indirect sources of cortical activation has largely been attempted using careful experimental designs [2Biabani M. Fornito A. Mutanen T.P. Morrow J. Rogasch N.C. Characterizing and minimizing the contribution of sensory inputs to TMS-evoked potentials.Brain Stimul. 2019; 12: 1537-1552Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar, 3Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar, 4Conde V. Tomasevic L. Akopian I. Stanek K. Saturnino G.B. Thielscher A. Bergmann T.O. Siebner H.R. The non-transcranial TMS-evoked potential is an inherent source of ambiguity in TMS-EEG studies.Neuroimage. 2019; 185: 300-312Crossref PubMed Scopus (84) Google Scholar, 5Gordon P.C. Desideri D. Belardinelli P. Zrenner C. Ziemann U. Comparison of cortical EEG responses to realistic sham versus real TMS of human motor cortex.Brain Stimul. 2018; 11: 1322-1330Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar]. Here, we complement those advances by providing a novel statistical approach. Our results converge with the pattern of indirect activation identified by recent experimental work [[3]Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar], which provides encouraging support for the potential utility of this novel approach. Several different experimental strategies have been used to deal with the indirect response of stimulation. One approach is to suppress the sources of multisensory stimulation and consequently minimize indirect cortical activations. This is powerfully illustrated in a recent study, in which the sound of the coil was very efficiently masked by playing noise-masking via earplugs as well as using ear defenders, and the vibrations of the coil were attenuated by using a foam layer beneath the coil [[3]Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar]. Whether this suppression approach is completely effective is debated (e.g. [[4]Conde V. Tomasevic L. Akopian I. Stanek K. Saturnino G.B. Thielscher A. Bergmann T.O. Siebner H.R. The non-transcranial TMS-evoked potential is an inherent source of ambiguity in TMS-EEG studies.Neuroimage. 2019; 185: 300-312Crossref PubMed Scopus (84) Google Scholar,[6]Belardinelli P. Biabani M. Blumberger D.M. Bortoletto M. Casarotto S. David O. Desideri D. Etkin A. Ferrarelli F. Fitzgerald P.B. et al.Reproducibility in TMS–EEG studies: a call for data sharing, standard procedures and effective experimental control.Brain Stimul. 2019; 12: 787-790Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar]). An alternative experimental approach is to use sham stimulation to mimic, and consequently isolate multisensory sources of indirect cortical activation [3Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar, 4Conde V. Tomasevic L. Akopian I. Stanek K. Saturnino G.B. Thielscher A. Bergmann T.O. Siebner H.R. The non-transcranial TMS-evoked potential is an inherent source of ambiguity in TMS-EEG studies.Neuroimage. 2019; 185: 300-312Crossref PubMed Scopus (84) Google Scholar, 5Gordon P.C. Desideri D. Belardinelli P. Zrenner C. Ziemann U. Comparison of cortical EEG responses to realistic sham versus real TMS of human motor cortex.Brain Stimul. 2018; 11: 1322-1330Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar]. Yet, this requires techniques to accurately reproduce the multisensory response elicited by TMS [[3]Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar,[4]Conde V. Tomasevic L. Akopian I. Stanek K. Saturnino G.B. Thielscher A. Bergmann T.O. Siebner H.R. The non-transcranial TMS-evoked potential is an inherent source of ambiguity in TMS-EEG studies.Neuroimage. 2019; 185: 300-312Crossref PubMed Scopus (84) Google Scholar], which is challenging to achieve. Both of these approaches deal with indirect responses to stimulation using experimental manipulations within-subjects. We offer a complementary approach using between-subjects analysis to indentify the indirect cortical activations due to stimulation. Commonly in TMS–EEG studies, participants receive a stimulation intensity tailored to their individual susceptibility to stimulation. When stimulating the primary motor cortex (M1) this is called the resting-motor threshold (rMT, [[7]Rossini P.M. Burke D. Chen R. Cohen L.G. Daskalakis Z. Di Iorio R. Di Lazzaro V. Ferreri F. Fitzgerald P.B. George M.S. et al.Non-invasive electrical and magnetic stimulation of the brain, spinal cord, roots and peripheral nerves: basic principles and procedures for routine clinical and research application. An updated report from an I.F.C.N. Committee.Clin Neurophysiol. 2015; 126: 1071-1107Crossref PubMed Scopus (1100) Google Scholar]; relative intensity). This corresponds to a percentage of stimulator output (absolute intensity). For example, for one participant 80% of rMT may require 48% of the maximum stimulator output (MSO); while for another participant 80% of rMT will be achieved with only 36% of MSO (Fig. 1A). The administered relative intensity (e.g. 80% of rMT) represents a normalization implying that the magnitude of direct motor cortical activation should be comparable across subjects. By contrast, the indirect activation of a TMS pulse may be closely related to the amount of sensory input, which itself is higher with higher absolute intensity (e.g. 36% vs 48% of MSO; i.e., louder sound, stronger coil vibration, and stronger stimulation of somatosensory fibres). As a consequence, in a between-subjects design, those receiving a higher absolute intensity of stimulation may experience more multisensory stimulation, and in turn display a higher amplitude of indirect cortical activation. We tested how well absolute intensity could explain the between-subject variability of EEG responses, and whether this corresponded to the spatiotemporal pattern of indirect cortical activations identified in earlier work [[3]Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar]. We analysed a pre-existing TMS–EEG dataset (58 participants, 41 females, 23 ± 4 years old (mean ± std); right-handed (defined by Edinburgh Handedness Inventory); neurologically and psychiatrically normal participants, meeting the safety criteria for the use of TMS). Participants received 126 neuro-navigated biphasic single TMS-pulses to left primary motor cortex (M1), while EEG (a 62-channel system) was concurrently recorded. Stimulation intensity was set at 80% of participants' rMT ([[7]Rossini P.M. Burke D. Chen R. Cohen L.G. Daskalakis Z. Di Iorio R. Di Lazzaro V. Ferreri F. Fitzgerald P.B. George M.S. et al.Non-invasive electrical and magnetic stimulation of the brain, spinal cord, roots and peripheral nerves: basic principles and procedures for routine clinical and research application. An updated report from an I.F.C.N. Committee.Clin Neurophysiol. 2015; 126: 1071-1107Crossref PubMed Scopus (1100) Google Scholar]; 54 ± 8% MSO (mean ± std), range from 31 to 77% MSO). Individually adjusted white noise (the volume was gradually increased until participants could not hear the coil click anymore or their threshold of discomfort was reached) was played during the stimulation via padded earplugs. In our analysis, we concentrated on the most commonly assessed dimension of TMS–EEG signal (i.e., TEPs [[3]Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar]; Fig. 1B). Initially, the signal was cleaned from the typical noise and TMS-induced artifacts, via the MATLAB-based toolbox EEGLAB and the TMS–EEG signal analyser plugin (TESA [[8]Rogasch N.C. Sullivan C. Thomson R.H. Rose N.S. Bailey N.W. Fitzgerald P.B. Farzan F. Hernandez-Pavon J.C. Analysing concurrent transcranial magnetic stimulation and electroencephalographic data: a review and introduction to the open-source TESA software.Neuroimage. 2017; 147: 934-951Crossref PubMed Scopus (86) Google Scholar]). For evaluating the spatio-temporal effects of absolute stimulation intensity, we computed TEPs (as the average time-course over trials) for each channel and each subject separately (in the average reference). Then, at each resulting time point (331; i.e. 20 – 350 ms after the TMS pulse) and at each channel (62 channels) we used linear regression to test whether the independent variable, subject-specific absolute intensity, can predict the subject-specific TEP amplitude (FieldTrip [[9]Oostenveld R. Fries P. Maris E. Schoffelen J.-M. FieldTrip: open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data.Comput Intell Neurosci. 2011; 2011: 156869Crossref PubMed Scopus (4073) Google Scholar]). To test for possible significance, we ran independent-samples regression coefficient t-tests (two-tailed, α level p < 0.025). To control the multiple comparison problem (the familywise error rate [[10]Maris E. Oostenveld R. Nonparametric statistical testing of EEG- and MEG-data.J Neurosci Methods. 2007; 164: 177-190Crossref PubMed Scopus (3528) Google Scholar]), clusters were built by assembling neighbouring significant spatio-temporal samples and tested against corresponding permutation statistics (5000 iterations, cluster α level p < 0.025 [[10]Maris E. Oostenveld R. Nonparametric statistical testing of EEG- and MEG-data.J Neurosci Methods. 2007; 164: 177-190Crossref PubMed Scopus (3528) Google Scholar]). We found that absolute stimulation intensity showed a significant regression against TEP amplitude in three different clusters. All three of the clusters occurred at later latencies. One of these was a negative cluster showing a negative TEP deflection (Fig. 1C, bottom row). The first was a positive cluster visible at 174–241 ms after the TMS pulse and mainly involved central electrodes (corrected p = 0.025; Fig. 1C, top row). The second positive cluster occurred later at 264–337 ms and was located above the left sensorimotor cortex (corrected p = 0.034, Fig. 1C, middle row). Finally, the negative cluster was found at 259–303 ms and involved central electrodes as well as electrodes located above left-frontal regions (corrected p = 0.038, Fig. 1C, bottom row). Based on their spatio-temporal patterns, the three clusters may be due to indirect cortical activations elicited by the sensory stimulation caused by TMS (e.g. a combination of auditory and somatosensory processes [[3]Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar]). These results show that the amplitude of late (>170 ms) TEPs were related to the absolute stimulation intensity. This may identify indirect cortical activations, and suggests that in a between-subject design absolute intensity is linked to the multisensory stimulation elicited by delivering TMS. Despite the methodological differences, our results converge both temporally and spatially with the pattern of indirect activation identified by recent experimental work [[3]Rocchi L. Di Santo A. Brown K. Ibáñez J. Casula E. Rawji V. Di Lazzaro V. Koch G. Rothwell J. Disentangling EEG responses to TMS due to cortical and peripheral activations.Brain Stimul. 2021; 14: 4-18Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar]. Hence, our analytical method makes the identification of indirect activations possible – unconstrained by experimental design – allowing new questions to be posed using innovative designs, which were previously impossible. Yet, further work is needed to fully validate this between-subject analytical approach. Nonetheless, it may prove to be a simple yet effective tool to identify and potentially remove multisensory contributions from TMS–EEG data; without the need to include additional experimental conditions within a study. We declare that the research was conducted free from any commercial or financial relationships that could have lead, or be seen to have lead to a conflict of interest. We are grateful to the Air Force Office of Scientific Research ( FA9550-16-1-0191 ; EMR) and the Academy of Finland (Grant No. 321631 ; TPM) for supporting this work.