Advertisement

Multi-modal investigation of transcranial ultrasound-induced neuroplasticity of the human motor cortex

Open AccessPublished:October 10, 2022DOI:https://doi.org/10.1016/j.brs.2022.10.001

      Highlights

      • tbTUS impacts motor cortex excitability and intracortical circuits.
      • Repetitive tbTUS induces durable alterations in the human motor cortex.
      • tbTUS affects connectivity at the whole brain level and at distinct motor centers.
      • Understanding the mechanisms of tbTUS can inform novel neuromodulation protocols.

      Abstract

      Introduction

      There is currently a gap in accessibility to neuromodulation tools that can approximate the efficacy and spatial resolution of invasive methods. Low intensity transcranial focused ultrasound stimulation (TUS) is an emerging technology for non-invasive brain stimulation (NIBS) that can penetrate cortical and deep brain structures with more focal stimulation compared to existing NIBS modalities. Theta burst TUS (tbTUS, TUS delivered in a theta burst pattern) is a novel repetitive TUS protocol that can induce durable changes in motor cortex excitability, thereby holding promise as a novel neuromodulation tool with durable effects.

      Objective

      The aim of the present study was to elucidate the neurophysiologic effects of tbTUS motor cortical excitability, as well on local and global neural oscillations and network connectivity.

      Methods

      An 80-s train of active or sham tbTUS was delivered to the left motor cortex in 15 healthy subjects. Motor cortical excitability was investigated through transcranial magnetic stimulation (TMS)-elicited motor-evoked potentials (MEPs), short-interval intracortical inhibition (SICI), and intracortical facilitation (ICF) using paired-pulse TMS. Magnetoencephalography (MEG) recordings during resting state and an index finger abduction-adduction task were used to assess oscillatory brain responses and network connectivity. The correlations between the changes in neural oscillations and motor cortical excitability were also evaluated.

      Results

      tbTUS to the motor cortex results in a sustained increase in MEP amplitude and decreased SICI, but no change in ICF. MEG spectral power analysis revealed TUS-mediated desynchronization in alpha and beta spectral power. Significant changes in alpha power were detected within the supplementary motor cortex (Right > Left) and changes in beta power within bilateral supplementary motor cortices, right basal ganglia and parietal regions. Coherence analysis revealed increased local connectivity in motor areas. MEP and SICI changes correlated with both local and inter-regional coherence.

      Conclusion

      The findings from this study provide novel insights into the neurophysiologic basis of TUS-mediated neuroplasticity and point to the involvement of regions within the motor network in mediating this sustained response. Future studies may further characterize the durability of TUS-mediated neuroplasticity and its clinical applications as a neuromodulation strategy for neurological and psychiatric disorders.

      Keywords

      1. Introduction

      Non-invasive brain stimulation (NIBS) tools are used in the research setting to probe the physiologic basis of brain activity, and in the clinical setting to treat various neurologic and psychiatric disorders. However, current NIBS tools are limited in their spatial resolution and ability to target deep brain structures. Low-intensity transcranial ultrasound (TUS) is an emerging NIBS modality with a favorable safety and tolerability profile in humans [
      • Sarica C.
      • Nankoo J.-F.
      • Fomenko A.
      • Grippe T.C.
      • Yamamoto K.
      • Samuel N.
      • et al.
      Human Studies of Transcranial Ultrasound neuromodulation: a systematic review of effectiveness and safety.
      ,
      • Darmani G.
      • Bergmann T.O.
      • Butts Pauly K.
      • Caskey C.F.
      • de Lecea L.
      • Fomenko A.
      • et al.
      Non-invasive transcranial ultrasound stimulation for neuromodulation.
      ]. In comparison to other NIBS approaches such as transcranial magnetic stimulation (TMS) and transcranial electrical stimulation (tDCS), TUS offers superior spatial resolution, relatively low cost and effective targeting of both superficial and deep brain regions [
      • Fomenko A.
      • Neudorfer C.
      • Dallapiazza R.F.
      • Kalia S.K.
      • Lozano A.M.
      Low-intensity ultrasound neuromodulation: an overview of mechanisms and emerging human applications.
      ]. TUS has also been shown to reversibly excite and inhibit targeted brain regions and neural circuits in a variety of model organisms [
      • Darmani G.
      • Bergmann T.O.
      • Butts Pauly K.
      • Caskey C.F.
      • de Lecea L.
      • Fomenko A.
      • et al.
      Non-invasive transcranial ultrasound stimulation for neuromodulation.
      ]. The impact and duration of these neuromodulatory effects depends on stimulation parameters. TUS studies on nonhuman primates, have introduced protocols to induce neuroplastic changes which could range from minutes to hours following sonication [
      • Folloni D.
      • Verhagen L.
      • Mars R.B.
      • Fouragnan E.
      • Constans C.
      • Aubry J.-F.
      • et al.
      Manipulation of subcortical and deep cortical activity in the primate brain using transcranial focused ultrasound stimulation.
      ,
      • Fouragnan E.F.
      • Chau B.K.H.
      • Folloni D.
      • Kolling N.
      • Verhagen L.
      • Klein-Flugge M.
      • et al.
      The macaque anterior cingulate cortex translates counterfactual choice value into actual behavioral change.
      ,
      • Verhagen L.
      • Gallea C.
      • Folloni D.
      • Constans C.
      • Jensen D.E.
      • Ahnine H.
      • et al.
      Offline impact of transcranial focused ultrasound on cortical activation in primates.
      ]. Despite the successful application of these protocols in neuromodulation, the high intensity spatial peak pulse average (ISPPA) and spatial-peak temporal average (ISPTA) used in those studies made a direct and safe translation of those protocols to human studies challenging.
      To address these limitations, an offline theta-burst TUS (tbTUS) protocol has tested and successfully induced consistent changes in excitability of the primary motor cortex (M1) at least 30 min following sonication [
      • Zeng K.
      • Darmani G.
      • Fomenko A.
      • Xia X.
      • Tran S.
      • Nankoo J.-F.
      • et al.
      Induction of human motor cortex plasticity by theta burst transcranial ultrasound stimulation.
      ]. In contrast to offline TUS protocols in nonhuman primate studies, tbTUS uses a lower ISPPA and ISPTA, making its application to the human brain safer and more tolerable. This has implications for the use of TUS as a neuromodulation tool, for example, for movement disorders such as Parkinson's Disease, as well as for motor recovery following traumatic brain injury or stroke. Yet, the impact of tbTUS on neural networks on a global neurophysiologic level beyond M1 has not been systematically explored. To harness the potential of tbTUS, a better understanding of its mechanism is needed.
      Previous studies have investigated the effects of TUS using TMS-evoked electromyographic (EMG) responses, functional MRI (fMRI) [
      • Yoo S.-S.
      • Kim H.
      • Min B.-K.
      • Franck E.
      • Park S.
      Transcranial focused ultrasound to the thalamus alters anesthesia time in rats.
      ], positron emission tomography (PET) [
      • Kim H.
      • Park M.-A.
      • Wang S.
      • Chiu A.
      • Fischer K.
      • Yoo S.-S.
      PETCT imaging evidence of FUS-mediated (18)F-FDG uptake changes in rat brain.
      ], and electroencephalography (EEG) [
      • Legon W.
      • Sato T.F.
      • Opitz A.
      • Mueller J.
      • Barbour A.
      • Williams A.
      • et al.
      Transcranial focused ultrasound modulates the activity of primary somatosensory cortex in humans.
      ,
      • Mueller J.
      • Legon W.
      • Opitz A.
      • Sato T.F.
      • Tyler W.J.
      Transcranial focused ultrasound modulates intrinsic and evoked EEG dynamics.
      ], among other methods [
      • Wattiez N.
      • Constans C.
      • Deffieux T.
      • Daye P.M.
      • Tanter M.
      • Aubry J.-F.
      • et al.
      Transcranial ultrasonic stimulation modulates single-neuron discharge in macaques performing an antisaccade task.
      ,
      • Guo H.
      • Hamilton 2nd, M.
      • Offutt S.J.
      • Gloeckner C.D.
      • Li T.
      • Kim Y.
      • et al.
      Ultrasound produces extensive brain activation via a cochlear pathway.
      ]. While these studies have garnered insight into local TUS-mediated neurophysiologic effects, the impact of TUS on neural networks on a global neurophysiologic scale has not been systematically explored. We hypothesized that: (1) tbTUS over the left motor cortex (M1) not only increases its cortical excitability, but also effects a network of areas anatomically or functionally linked to it; (2) the neuromodulatory effects of TUS are reflected in changes in neuronal rhythms associated with motor control, ultimately impacting functional coupling of the motor network associated to M1. Here, we take advantage of the temporal and spatial properties of magnetoencephalography (MEG) to test these hypotheses in humans.
      MEG represents a unique non-invasive brain mapping tool to evaluate TUS-mediated effects on specific targets. MEG has several advantages over other methodologies. The magnetic fields measured by MEG pass unimpeded through the skull and this technique allows greater detection of high frequency electromagnetic oscillations, such as gamma oscillation, which cannot be reliably detected by EEG. In addition, MEG offers high temporal resolution compared to other techniques such as functional magnetic resonance imaging (fMRI). By combining TMS with MEG in healthy individuals, we describe the landscape of neurophysiologic changes secondary to TUS in humans.

      2. Methods

      2.1 Study subjects and experimental procedure

      A total of 15 healthy subjects (7 females, age range: 20–37 years, mean age 27.1 years, SD: 5 ± 1 years; Supplementary Table 1) were recruited from advertisements posted at the University Health Network (UHN), in accordance with accepted Research Ethics Board protocols. None of the participants had a history of neurological or psychiatric disorders and were not taking any medications. Subjects underwent a neurological examination conducted by a study physician before participating and no deficits were noted in any participant. All subjects participated in three randomly assigned study visits, with one visit each for sham and active tbTUS separated by at least 7 days and one visit for an MRI scan (Supplementary Table 2). The study was approved by the UHN (Toronto, Canada) Research Ethics Board. Written informed consent was obtained from each subject, and all experimental procedures were performed in accordance with the Declaration of Helsinki.
      The experimental procedure is depicted in Fig. 1. We first obtained baseline TMS measurements. Subjects were then taken to the magnetically shielded room (MSR) for a baseline MEG recording at rest and during a motor task (self-paced right index finger abduction/adduction). An accelerometer was attached to the right index finger of each subject to measure acceleration in the three cardinal planes. The plane of maximal acceleration was used to quantify finger movements. Brainstorm v3.220314, a free and open-source software package, was used for event detection. Finger movement was identified as events of peak acceleration greater than three standard deviations above the baseline signal and no closer than 1.5 s apart during 80-s epochs [
      • Tadel F.
      • Baillet S.
      • Mosher J.C.
      • Pantazis D.
      • Leahy R.M.
      Brainstorm: a user-friendly application for MEG/EEG analysis.
      ]. A paired two-tailed t-test was used to assess the difference between the mean number of finger movements between pre-TUS and post-TUS conditions for both active and sham conditions (pre-TUS: measurements taken prior to active/sham TUS; post-TUS: measurements taken after active/sham TUS (Supplementary Fig. 1). Each condition was recorded in duplicate for a duration of 4 min for the first recording and 3 min for the subsequent recording. Following the baseline MEG recording, subjects were taken outside of the MSR for active or sham tbTUS at the left motor cortex. The subjects returned to the MSR for a post-intervention MEG recording at rest and during the motor task. Lastly, post-intervention TMS measurements were obtained.
      Fig. 1
      Fig. 1Experiment workflow diagram. Study subjects presented for two visits and were randomly assigned to receive active tbTUS or sham tbTUS at the first visit and the alternate condition at the second visit. Transcranial magnetic stimulation (TMS) measurements were taken prior to each visit (blue octagonal panels). The first dorsal interosseous (FDI) hotspot was determined using the centre of the TMS coil where TMS effects are believed to be located. The resting motor threshold (RMT) was then measured, followed by 20 pre-motor evoked potential (MEP) trials. Ten trials for each of the following were performed: Test Stimulus (TS), Short interval intracortical inhibition (SICI) and intracortical facilitation (ICF). The subject was then taken to the magnetoencephalography (MEG) suite where two pre-TUS and post-TUS measurements were taken (each at rest and with a finger abduction motor task). Following completion of MEG data acquisition, post-TUS TMS measurements were repeated (PRE = measurements prior to active/sham TUS; POST = measurements after active/sham TUS). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

      2.2 Transcranial magnetic stimulation and hotspot mapping

      TMS was applied to the left primary motor cortex (M1) using a 70 mm figure-eight coil connected to four Magstim 2002 stimulators via a “four-to-one” connection box (Magstim, Whitland, Dyfed, UK). The coil was held tangentially to the skull with the handle pointing backward and laterally at a 45° angle to the sagittal plane. TMS pulses were triggered using external configurations built with Signal 4.07 software (Cambridge Electronic Design, Cambridge, UK).
      Motor cortex excitability before and after ultrasound stimulation was assessed using single-pulse TMS-elicited MEPs. At both collection points, TMS pulses were delivered at an inter-trial interval of 5s and used the stimulator intensity, which induced MEPs of ∼1 mV peak-to-peak amplitude (SI1mV) prior to ultrasound or sham stimulation (mean ± SD %maximum stimulator output [MSO]). Twenty MEPs were recorded for each time point. Intrinsic circuits of M1, including short-latency intracortical inhibition (SICI) and intracortical facilitation (ICF) were probed using paired-pulse TMS at the left M1 before and after ultrasound stimulation. Both procedures consisted of a subthreshold conditioning stimulus (CS) followed by a suprathreshold test stimulus (TS), both delivered using the same TMS coil.
      For SICI, the interstimulus interval (ISI) between CS and TS was 2 ms, and for ICF, the ISI was 10 ms. For both circuits, the CS was set at 80% of the resting motor threshold (RMT), and TS was the stimulator intensity which induced MEPs of ∼1 mV peak-to-peak amplitude (SI1mV). RMT was defined as the lowest stimulator intensity required to elicit MEPs of at least 50 μV in the relaxed FDI muscle in at least 5 of 10 consecutive pulses. As RMT and SI1mV may change after tbTUS, the RMT and SI1mV used for the assessment of SICI and ICF were estimated independently at each time point. For each ISI (2 ms and 10 ms), ten MEPs were collected. In addition, ten trials of TS alone were collected for a total of 30 trials. All trials were tested randomly in the same block. No subjects reported any adverse effects throughout the experiments, supporting the safety profile of the current pulsing schemes. For SICI and ICF trials, the peak-to-peak amplitudes of the conditioned MEPs were expressed as a ratio of the mean peak-to-peak amplitude of the unconditioned MEPs from the TS-only trials. Ratios larger than one indicated facilitation, while ratios smaller than one indicated inhibition.
      To map the target for tbTUS, the TMS coil was held tangentially to the skull with the handle pointing backward and laterally at a 45° angle to the sagittal plane. Stimulation was delivered to the first dorsal interosseous (FDI) “hotspot” in M1, defined as the position evoking the highest and most consistent MEPs on average in the FDI muscle. The hotspot was found using the TMS coil by moving the coil and testing placements in the forward and back direction, and then testing side to side at the best spot in the forward-back axis. The location was then marked with a permanent marker to ensure consistent TMS coil repositioning throughout the experiment and to centre the ultrasound transducer.

      2.2.1 Electromyography recording

      Surface electromyography (EMG) was recorded from the right first dorsal interosseous (FDI) muscle using a pair of 9-mm diameter Ag–AgCl electrodes in a belly-tendon montage. All recordings were made with the FDI muscle at rest. EMG signals were amplified at 1K (Intronix Technologies Corporation Model 2024F, Bolton, Ontario, Canada), bandpass filtered between 20 and 2500 Hz, and digitized at 5 kHz (Micro 1401, Cambridge Electronics Design, Cambridge, United Kingdom). Recordings were stored in a laboratory computer for offline analyses.

      2.3 Transcranial ultrasound stimulation of the motor cortex

      Ultrasound stimulation was delivered using a custom two-element annular array ultrasound transducer (H246, Sonic Concepts Inc., Bothell, Washington) with a fundamental frequency of 0.5 MHz, diameter of 38 mm, and thickness of 10 mm. A programmable radiofrequency amplifier (Transducer Power Output System TPO201-80, Sonic Concepts Inc., Bothell, Washington) delivered the required power to the transducer via a 50Ω impedance matching module. The sonication depth was set as 30 mm according to the scalp-cortex distance to the hand motor area [
      • Stokes M.G.
      • Chambers C.D.
      • Gould I.C.
      • Henderson T.R.
      • Janko N.E.
      • Allen N.B.
      • et al.
      Simple metric for scaling motor threshold based on scalp-cortex distance: application to studies using transcranial magnetic stimulation.
      ] as reported in a previous study [
      • Zeng K.
      • Darmani G.
      • Fomenko A.
      • Xia X.
      • Tran S.
      • Nankoo J.-F.
      • et al.
      Induction of human motor cortex plasticity by theta burst transcranial ultrasound stimulation.
      ]. For sham tbTUS, the transducer was flipped so that the inactive face of the transducer was in contact with the scalp and ultrasonic energy was directed away from the head. No subjects reported any adverse effects throughout the experiments, supporting the safety profile of tbTUS pulsing schemes.
      In accordance with the study parameters for tbTUS based on the International Expert Group on Transcranial Ultrasonic Stimulation Safety and Standards (ITRUSST, Supplementary Table 3). Ultrasound stimulation was targeted to the FDI hotspot, as determined using the centre of the TMS coil where TMS effects are believed to be located, as described above. The theta burst TUS (tbTUS) paradigm consists of an 80s train of pulses with pulse repetition frequency (PFR) of 5 Hz, pulse duration of 20 ms, ultrasonic stimulus duration of 200 ms, and duty cycle of 10%, for a total number of 400 pulses. The power of ultrasound was set as 20W. The acoustic focus of the ultrasound waveform was measured in a previous study using a calibrated fiber optic hydrophone coupled to a preamplifier [
      • Zeng K.
      • Darmani G.
      • Fomenko A.
      • Xia X.
      • Tran S.
      • Nankoo J.-F.
      • et al.
      Induction of human motor cortex plasticity by theta burst transcranial ultrasound stimulation.
      ]. Radial (x and y axes) and longitudinal (z axis) measurements were made in water with the transducer excited at the fundamental frequency (0.5 MHz), focal length of 30 mm, and intensity of 20 W2. Based on the acoustic pressure field, the spatial peak-pulse-average intensity (ISPPA) was 2.26 W/cm2 and the spatial-peak time-average intensity (ISPTA) was 0.23 W/cm2, both well below the United States Food and Drug Administration (FDA) safety standards. Conductive gel (Wavelength® MP Blue Multi-Purpose Ultrasound Gel) was applied to the defined hotspot prior to ultrasound transducer placement on the scalp.

      2.4 Magnetoencephalography acquisition

      MEG recordings were conducted using a 306-channel MEG system (Elekta Neuromag TRIUX, Helsinki, Finland) at a sampling rate of 1000 Hz with an online bandpass filter between 0.1 and 330 Hz. Recording occurred while participants were in a supine position with their eyes closed. To determine the effect of theta-burst TUS on cortical, subcortical and cerebellar activity, participants underwent two separate MEG recording sessions as described above. The two replicate measurements were averaged for source power and connectivity analyses.

      2.5 Magnetoencephalography data processing and analysis

      Artifact suppression in the MEG data was performed using the temporal signal space separation (tSSS) algorithm implemented within the Neuromag MaxFilter system (software version 2.2.12, Elekta, Helsinki, Finland) [
      • Taulu S.
      • Hari R.
      Removal of magnetoencephalographic artifacts with temporal signal-space separation: demonstration with single-trial auditory-evoked responses.
      ,
      • Wennberg R.
      • Del Campo J.M.
      • Shampur N.
      • Rowland N.C.
      • Valiante T.
      • Lozano A.M.
      • et al.
      Feasibility of magnetoencephalographic source imaging in patients with thalamic deep brain stimulation for epilepsy.
      ,
      • Kandemir A.L.
      • Litvak V.
      • Florin E.
      The comparative performance of DBS artefact rejection methods for MEG recordings.
      ] with the default 10-s time window and subspace correlation limit of 0.980. Gradiometer data was processed and analyzed in Matlab (The Mathworks, Natick, MA, USA) using the FieldTrip toolbox [
      • Oostenveld R.
      • Fries P.
      • Maris E.
      • Schoffelen J.-M.
      FieldTrip: open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data.
      ] and custom-made scripts. Data were epoched into 1-s trials, and a high-pass filter of 1 Hz was applied to remove drifts. Filters were also applied at 60, 120, and 180 Hz to remove power line noise and harmonics. All raw data were visually inspected, and segments with signal dropout or artifacts were discarded.
      For source power and coherence analyses, the time-series MEG data were transformed to the frequency domain using a multi-taper method fast Fourier transform, after which the power spectra were generated using a multi-taper frequency transformation [
      • Thomson D.J.
      Spectrum estimation and harmonic analysis.
      ] and discrete prolate spheroidal sequences. To estimate spectral power for specific brain regions, sources were localized using dynamic imaging of coherent sources [
      • Gross J.
      • Kujala J.
      • Hamalainen M.
      • Timmermann L.
      • Schnitzler A.
      • Salmelin R.
      Dynamic imaging of coherent sources: studying neural interactions in the human brain.
      ]. We used the Automated Anatomical Labeling (AAL) atlas to parcellate the single-subject Montreal Neurological Institute (MNI)-space template brain, derived from participants’ structural T1-weighted MRIs (template head models were used for participants without an MRI), such that absolute power estimates were generated for 116 brain regions [
      • Tzourio-Mazoyer N.
      • Landeau B.
      • Papathanassiou D.
      • Crivello F.
      • Etard O.
      • Delcroix N.
      • et al.
      Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain.
      ,
      • Collins D.L.
      • Holmes C.J.
      • Peters T.M.
      • Evans A.C.
      Automatic 3-D model-based neuroanatomical segmentation.
      ]. Frequencies were averaged into delta (1–3 Hz), theta (4–8 Hz), alpha (8–12 Hz), beta (13–29 Hz), low gamma (30–50 Hz), mid gamma (50–70 Hz), and broadband gamma (70–170 Hz) bands. Regions of spectral power desynchronization were depicted on the MNI ICBM 152 non-linear 6th generation symmetric average brain stereotaxic registration model using FSL software (FMRIB Software Library v6.0) [
      • Jenkinson M.
      • Beckmann C.F.
      • Behrens T.E.J.
      • Woolrich M.W.
      • Smith S.M.
      ].
      Next, we analyzed event-related fields given our motor task. We conducted time-locked analysis using ft_timelockanalysis to quantify possible event-related spectral desynchronization (ERSD) and rebound effects for sensorimotor alpha and beta activity. The timing of movements was captured via the accelerometer. Since self-paced movements occurred at a frequency of 1 Hz, we analyzed data 0.5s before and 0.5s after movement in the alpha and beta band, resulting in 60-s epochs centered around movement.
      We employed cluster-based permutation tests on time-frequency data and event-related fields to control alpha inflation across multiple comparisons. We used ft_freqstatistics and ft_clusterplot for visualization of significant sensor-level regions. DICS beamformer source power results were visualized using 3D surface brains. In addition, functional connectivity between AAL-labeled brain regions was estimated using coherence values from 116 virtual channels [
      • Jenkinson M.
      • Beckmann C.F.
      • Behrens T.E.J.
      • Woolrich M.W.
      • Smith S.M.
      ,
      • Hillebrand A.
      • Singh K.D.
      • Holliday I.E.
      • Furlong P.L.
      • Barnes G.R.
      A new approach to neuroimaging with magnetoencephalography.
      ,
      • Van Veen B.D.
      • van Drongelen W.
      • Yuchtman M.
      • Suzuki A.
      Localization of brain electrical activity via linearly constrained minimum variance spatial filtering.
      ,
      • Rosenberg J.R.
      • Amjad A.M.
      • Breeze P.
      • Brillinger D.R.
      • Halliday D.M.
      The Fourier approach to the identification of functional coupling between neuronal spike trains.
      ]. Separate matrices were generated for each condition and consisted of 116x116 brain region pairs. For comparative purposes, source regions were clustered into eight major bilateral brain regions (frontal, sensorimotor, basal ganglia, limbic, parietal, temporal, occipital, cerebellar) and midline vermis. Local functional connectivity was estimated by averaging coherence for all brain region pairs within each of the major brain regions. Interregional (between region of interest (ROI)) connectivity was estimated by averaging coherence for all brain region pairs between two ROIs and included interhemispheric (left vs. right) and intra-hemispheric (left vs. left and right vs. right) comparisons.

      2.6 Statistical analysis

      Parametric statistics were used for MEG data because it is ‘super-averaged’ (e.g., sensorimotor cortical beta source power is an average of its anatomical parcel, 17 frequencies, 300 epochs, and 2 trials) [
      • Kwak S.G.
      • Kim J.H.
      Central limit theorem: the cornerstone of modern statistics.
      ,
      • Kiebel S.J.
      • Tallon-Baudry C.
      • Friston K.J.
      Parametric analysis of oscillatory activity as measured with EEG/MEG.
      ,
      ]. MEP, SICI and ICF were estimated with trials with EMG background within 2.5 standard deviations (SD) above the mean. Single-trial EMG background was estimated as the peak-to-peak EMG activity from 30 ms to 100 ms before TMS pulse. Single-subject EMG background was estimated by averaging the single-trial EMG background across trials. We used a student's paired two-tailed t-test (denoted by the test statistic, t) for the separate analysis and subsequent comparison of pre- and post-TUS conditions (dependent samples). Linear regression using Pearson's coefficient, R, was used to determine the association between continuous variables (e.g., MEP amplitude and SICI value vs. coherence). Significance was set at p < 0.05 for all tests, and corrections were made using false discovery rate with a significance threshold of 5% for analyses involving multiple comparisons [
      • Benjamini Y.
      • Hochberg Y.
      Controlling the false discovery rate: a practical and powerful approach to multiple testing.
      ]. MEG-related statistical analyses were conducted using Matlab (The Mathworks, Natick, MA, USA, http://www.mathworks.com).

      3. Results

      3.1 Theta burst TUS (tbTUS) modulates motor cortical excitability

      Single-pulse TMS measures of motor evoked potential (MEP) peak-to-peak amplitude and resting motor threshold (RMT) were used to assess changes in cortical excitability following active or sham tbTUS to the motor cortex. All analyses were estimated with curated trials, whereby 0.98 ± 1.7 trials, 0.75 ± 1.42 trials, and 0.86 ± 1.53 trials were rejected due to artifacts for MEP amplitude SICI and ICF analyses, respectively. MEP amplitudes significantly increased from 1.03 ± 0.24 mV (mean ± SD) at the baseline to 1.39 ± 0.45 mV after active tbTUS (t(14) = 5.17, p < 0.001; Fig. 2A). By contrast, there was no significant difference in the MEP amplitudes before (1.07 ± 0.28 mV) and after (1.06 ± 0.29 mV) sham tbTUS (t(14) = 0.3, p = 0.77). There is no significant correlation between the variance of baseline MEPs and the increase of MEP after active tbTUS (r = 0.34, p = 0.217), hence, more stable baseline MEP amplitudes did not mean stronger TUS effect. Collectively, the MEP results support the facilitatory effects of tbTUS.
      Fig. 2
      Fig. 2Boxplots demonstrating the effects of theta-burst TUS (tbTUS) on motor cortical excitability. The panels on the left show the results in the active tbTUS condition for motor evoked potential (MEP) amplitude ((a) active and (b) sham), short-interval intracortical inhibition (SICI), and ((c) active and (d) sham), and intracortical facilitation (ICF) ((e) active and (f) sham). Each box depicts the upper and lower quartile of data, with the horizontal line representing the median value. The panels on the right show results for the sham tbTUS condition for the same metrics. Significant MEP facilitation was observed following active tbTUS but not sham tbTUS. For SICI and ICF, data is plotted as the ratio to unconditioned MEP amplitude (test stimulus, TS). Values below 1 denote inhibition, and values above 1 denote facilitation. Active tbTUS significantly decreases the SICI but not ICF. Sham tbTUS has no effect on both SICI and ICF (∗∗ and ∗∗∗: p < 0.01 and p < 0.001, respectively).
      Paired-pulse TMS was used to investigate the neurophysiologic effects of tbFUS on intracortical circuits of the motor cortex. Active tbTUS significantly decreases SICI (t(14) = 3.27, p = 0.005, Fig. 2B) from (0.40 ± 0.21) to (0.52 ± 0.22), but did not alter ICF (1.42 ± 0.46 vs 1.49 ± 0.56). As expected, sham tbTUS has no effect on both SICI (0.42 ± 0.20 vs 0.43 ± 0.19, Fig. 2C) and ICF (1.44 ± 0.48 vs 1.43 ± 0.46). The s SICI and ICF after TUS to baseline values also shows that active tbTUS significantly decreases SICI (t(14) = 2.64, p = 0.019) compared to sham tbTUS, but did not change ICF (t(14) = 0.17, p = 0.868). The baseline SICI and ICF in active tbTUS are significant correlated with that in sham tbTUS, confirming the validity of our findings related the modulation of intracortical circuits.
      The observed paired and single pulse effects were durable over an average of 72.4-min timeframe (SD = 9.5 min) from baseline measurements and an average of 26.7 min (SD = 5.3 min) from the completion of tbTUS. In all, the MEP amplitude and SICI were significantly modulated by tbTUS, however, there is no significant correlation (r = 0.28, p = 0.321) between the changes of MEP and SICI before and after active tbTUS. There were no differences observed in the baseline RMT between the active and sham conditions, and stimulus intensity that elicited MEP peak-to-peak amplitudes of 1 mV (SI1mV) were comparable (RMT: active tbTUS vs sham tbTUS (50.28 ± 10.1% vs 50.16 ± 10.8%); SI1mv: active tbTUS vs sham tbTUS (61 ± 14.5% vs 60.1 ± 14.1%)). The baseline MEP amplitudes between active and sham tbTUS were also similar (active tbTUS vs sham tbTUS (1.03 ± 0.24 mV vs 1.07 ± 0.28 mV)). These comparisons suggest that changes observed following tbTUS, but not sham tbTUS, were not due to differences in baseline cortical excitability or TMS stimulator intensity. In addition, we also examined the correlation for the baseline SI1mV and RMT between active and sham tbTUS. The results showed that the baseline SI1mv (r = 0.98, p < 0.001) and RMT (r = 0.92, p < 0.001) are significant correlated between the two visits, validating the robustness of these measurements.

      3.2 tbTUS impacts alpha and beta spectral power in distinct anatomical regions during a motor task

      Analysis of MEG-derived signals reveals that active, but not sham, tbTUS administration to the left motor cortex produced changes in alpha (8–12 Hz) and beta (13–29 Hz) spectral power during the right index finger abduction-adduction task, when compared to the resting state condition (Fig. 3; Supplementary Fig. 3; Supplementary Table 4). Movement frequency did not differ between the two conditions (Supplementary Fig. 1). Cluster-based permutation tests on time-frequency data were performed to control for alpha inflation across multiple comparisons. In particular, these analyses reveal left sensorimotor/fronto-temporal alpha spectral power increases with tbTUS at rest (Fig. 3A), as well as left sensorimotor/temporo-parietal beta increase with tbTUS during movement (Fig. 3B). No significant changes in spectral power for mid-gamma (50–70 Hz) and broadband gamma (70–170 Hz) frequency ranges were observed. Alpha spectral power alterations were most strongly detected as decreases in parcellated brain regions including the left limbic regions and left cerebellum, as well as bilateral sensorimotor cortices, with the greatest changes in alpha spectral power detected in the right supplementary motor region (Fig. 3B, pRSupplementaryMotor=0.0029). Changes in the beta frequency were detected at the whole-brain level and most prominently as decreased beta spectral power in the right parietal region (Fig. 3C, Supplementary Table 1; pRParietal=0.0021) and right basal ganglia (Supplementary Table 4; pRBasalGanglia = 0.0073). Significant changes in beta spectral power were also found in the bilateral sensorimotor cortices, bilateral limbic regions, left basal ganglia, left cerebellum and right occipital regions. There were no significant changes in spectral power for delta (1–3 Hz) or theta (4–8 Hz) bands. These movement-associated findings suggest that, in the alpha and beta frequency ranges, movement measured immediately following motor cortex tbTUS are associated with activation in distinct anatomical regions. Time-locked analysis to quantify possible event-related spectral desynchronization (ERSD) and rebound effects for sensorimotor alpha and beta activity did not reveal distinct significant differences in spectral power. As such, our results, in aggregate, represent an overall average of the oscillatory changes detected using MEG during a motor task following tbTUS administration.
      Fig. 3
      Fig. 3Regional changes in alpha and beta spectral power with movement-associated tbTUS. (AB) Cluster-based permutation testing. (A) Right sensorimotor/frontal alpha (centered around 10 Hz) increase with movement-associated tbTUS, p = 0.022. (B) Right sensorimotor/parieto-occipital beta (centered around 24 Hz) increase with movement-associated tbTUS, p = 0.044. Significant clusters, represented as ‘x’, for TUS/SHAM interaction were calculated as ([post-TUSMOVE – post-TUSREST] – [post-SHAMMOVE – post-SHAMREST]). Yellow (positive T-statistic values) represent an increase in spectral power with movement-associated tbTUS. Blue (negative T-statistic values) represent a decrease in spectral power with movement-associated tbTUS. (C-D) 3D representation of brain regions with the greatest statistically significant changes in (C) alpha and (B) beta spectral power. Statistically significant regions of changes in spectral power following active tbTUS with movement (normalized to resting state) as shown. Changes in spectral power in the regions shown were not observed using the same comparisons in the sham tbTUS condition. (C) Depiction of changes in alpha spectral power (red), with the most significant changes in the left limbic regions (LL) and left cerebellum (LC), as well as left and right sensorimotor cortices (LSM, RSM); (i) right, (ii) anterior, and (iii) superior views. (D) Beta spectral power changes (blue), with the most statistically significant changes in the right basal ganglia (RBG) and right parietal regions (RP); (iv) right, (v) anterior, and (vi) superior views. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

      3.3 Changes in local and interregional functional coupling are detected during a motor task in response to tbTUS

      To evaluate how tbTUS mediates spectral power changes, coherence was computed as a measure of functional connectivity between brain region pairs for all frequency bands assessed with MEG (Fig. 4A–E). Overall, tbTUS increased local and interregional functional coupling. Changes in connectivity profiles for delta (1–3 Hz) and theta (4–8 Hz) bands were similar, with local increases in parietal and cerebellar regions and widespread increases in interregional coupling involving the frontal, parietal, temporal, occipital, sensorimotor, basal ganglia, and cerebellar regions. Changes in alpha (8–12 Hz) coherence only involved interregional increases in coupling of the bilateral temporo-parietal regions, bilateral cerebelli, and left basal ganglia-cerebellum (p < 0.001). Lastly, increases in low-gamma (30–50 Hz) coupling were observed between bilateral fronto-parietal regions and left parieto-occipital regions, and locally, within the right cerebellum (p < 0.001). The sole instance of decoupling occurred in the beta band within the right basal ganglia (Fig. 4D). Significant changes in functional connectivity were not observed in mid-gamma (50–70 Hz) and broadband gamma (70–170 Hz) bands, as was the case for spectral power analysis.
      Fig. 4
      Fig. 4Local and interregional functional coupling is modulated by tbTUS (ae). a-e) Functional connectivity maps demonstrating changes in local and interregional coherence following motor cortex tbTUS for (a) delta (1–3 Hz), (b) theta (4–8 Hz), (c) alpha (8–12 Hz), (d) beta (13–29 Hz), and (e) low gamma (30–50 Hz) frequency bands. Each panel depicts the differences in local and interregional coherence between pre- and post-TUS during movement. Statistically significant increases in connectivity with tbTUS are represented in red (‘functional coupling’), decreases in connectivity in blue (‘decoupling’). Colour intensity indicates the degree of significance (t-statistic for FDR-corrected p < 0.05). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

      3.4 tbTUS-associated MEP and SICI changes correlate with changes functional coupling

      We next investigated whether tbTUS impacts functional coupling. Changes in coherence with tbTUS were related to changes in MEP amplitude and SICI by specifically investigating the aforementioned regions of significant changes in local and inter-regional connectivity (Fig. 5). MEP positively correlated with local and interregional coherence, such that subjects with the greatest tbTUS-mediated MEP amplitude increases (increased excitability) demonstrated the greatest increases in coupling (increased activation) (Fig. 5A, delta right parietal region, R = 0.55, p = 0.032; Fig. 5B, theta left frontal-left parietal connection, R = 0.55, p = 0.035).
      Fig. 5
      Fig. 5TUS-mediated changes in functional connectivity and correlates to changes in MEP and SICI (ad). MEP and SICI correlate to functional coupling. MEP positively correlated with (a) local and (b) interregional coherence, such that subjects with the greatest TUS-mediated MEP amplitude increases (i.e., increased excitability) had the greatest increases in coupling (i.e., increased activation). SICI positively correlated with (c) local and (d) interregional coherence, such that subjects with the greatest TUS-mediated SICI value increases (i.e., decreased inhibition) had the greatest increases in coupling (i.e., increased activation). Pearson's linear regression was used with a sample size of N = 15 participants. MEP, motor evoked potential; SICI, short-interval intracortical inhibition, RP = right parietal, LF = left frontal, LP = left parietal, RC = right cerebellum, RO = right occipital, RT = right thalamic.
      SICI also positively correlated with local and interregional coherence, such that subjects with the greatest tbTUS-mediated SICI value increases (decreased inhibition) had the greatest increases in coupling (increased activation) (Fig. 5C, delta right cerebellar region, R = 0.53, p = 0.043; Fig. 5D, alpha right temporal-right occipital connection, R = 0.55, p = 0.033). In addition, we observed a trending positive correlation (p = 0.080) between Delta RP and Theta LF-LP coherence, such that high Delta RP local connectivity is associated with high Theta LF-LP connectivity (Supplementary Fig. 2A). We observed a positive correlation (p = 0.034) between Delta RP and Alpha RO-RT coherence, such that high Delta RP local connectivity is associated with high Theta LF-LP connectivity (Supplementary Fig. 2B).

      4. Discussion

      An understanding of the mechanisms of transcranial ultrasound modulation is essential to inform its clinical applications. We obtained direct measures of neuroplasticity induced by tbTUS of the motor cortex in humans using multimodal outcome measures. To our knowledge, this is the first study to utilize MEG to assess the neuromodulatory effects of tbTUS in humans. Moreover, this study is unique in its integration of MEG with a motor task, source localization techniques, and connectivity analyses.
      The proposed mechanisms of TUS have primarily been speculated on at the cellular level. Leading hypotheses include (1) the formation of bubbles inside neuronal membranes impacting their capacitance; (2) lipid rafts sequestering membrane-associated enzymes thereby limiting their interaction with membrane-bound substrates; and (3) thermal or (4) mechanical forces disrupting neuronal membranes leading to increasing enzyme-substrate interactions or opening of mechanosensitive ion channels [
      • Darmani G.
      • Bergmann T.O.
      • Butts Pauly K.
      • Caskey C.F.
      • de Lecea L.
      • Fomenko A.
      • et al.
      Non-invasive transcranial ultrasound stimulation for neuromodulation.
      ]. However, there are currently limited neurophysiologic studies of the mechanisms of TUS. By combining MEG analysis with TMS-elicited measures of cortex excitability, we validated prior studies demonstrating that tbTUS can induce lasting changes in excitability [
      • Zeng K.
      • Darmani G.
      • Fomenko A.
      • Xia X.
      • Tran S.
      • Nankoo J.-F.
      • et al.
      Induction of human motor cortex plasticity by theta burst transcranial ultrasound stimulation.
      ]. The mechanisms of electrical, magnetic, and ultrasound stimulation differ and therefore we do not anticipate the results of theta burst TUS to mirror those of other NIBS modalities, such as TMS. For example, LTP was induced in the hippocampal slices by theta burst electrical stimulation while LTD-like plasticity was induced by cTMS in human motor cortex [
      • Larson J.
      • Wong D.
      • Lynch G.
      Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation.
      ].
      In addition, our study describes the global and regional changes in neural oscillations detected following application of tbTUS to the motor cortex. Such effects are comparable to those seen using other forms of non-invasive brain stimulation [
      • Arif Y.
      • Spooner R.K.
      • Heinrichs-Graham E.
      • Wilson T.W.
      High-definition transcranial direct current stimulation modulates performance and alpha/beta parieto-frontal connectivity serving fluid intelligence.
      ]. Investigation of tbTUS activity using MEG in the present study demonstrates movement-related decreases in alpha and beta spectral power within motor regions including the bilateral supplementary motor cortices, basal ganglia, and cerebellum. Previous fMRI studies have shown that electrical activity in the beta frequency is associated with resting state motor network activity [
      • Mantini D.
      • Perrucci M.G.
      • Del Gratta C.
      • Romani G.L.
      • Corbetta M.
      Electrophysiological signatures of resting state networks in the human brain.
      ]. The discrete regions of alpha and beta spectral power may converge on functionally connected regions of the resting and task-related motor networks. Activation of bilateral supplementary motor cortices in response to a finger movement task has been previously reported, and functional connectivity between the left and right motor cortices is well characterized [
      • Brookes M.J.
      • Woolrich M.
      • Luckhoo H.
      • Price D.
      • Hale J.R.
      • Stephenson M.C.
      • et al.
      Investigating the electrophysiological basis of resting state networks using magnetoencephalography.
      ]. In addition, the most significant regions identified in the present study correspond to functionally connected regions of the resting-state motor network including a bihemispheric network comprising premotor regions, parts of the somato-sensory and superior partietal cortex and post-central gyrus [
      • Pool E.-M.
      • Rehme A.K.
      • Eickhoff S.B.
      • Fink G.R.
      • Grefkes C.
      Functional resting-state connectivity of the human motor network: differences between right- and left-handers.
      ]. Although such functional networks were derived using fMRI, resting state connectivity patterns using MEG mirror results derived from fMRI studies [
      • Brookes M.J.
      • Hall E.L.
      • Robson S.E.
      • Price D.
      • Palaniyappan L.
      • Liddle E.B.
      • et al.
      Complexity measures in magnetoencephalography: measuring “disorder” in schizophrenia.
      ]. Finally, anatomical regions of activation and may be associated with TUS-mediated increases in functional coupling. For example, local decoupling of the right basal ganglia in the beta frequency range may also increase activation and movement-related desynchronization by reducing inhibitory output [
      • Larson J.
      • Wong D.
      • Lynch G.
      Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation.
      ]. Future studies involving patient populations with motor deficits may better elucidate a role for tbTUS in modulating movement execution.
      Our findings demonstrate changes in functional coupling following administration of tbTUS to the motor cortex. Overall increases in coupling promote the activation of areas as represented by movement-related alpha and beta spectral power changes within motor regions. In human EEG studies, TUS has been shown to modulate the activity of primary somatosensory cortex in the beta but not gamma frequency range [
      • Legon W.
      • Sato T.F.
      • Opitz A.
      • Mueller J.
      • Barbour A.
      • Williams A.
      • et al.
      Transcranial focused ultrasound modulates the activity of primary somatosensory cortex in humans.
      ,
      • Mueller J.
      • Legon W.
      • Opitz A.
      • Sato T.F.
      • Tyler W.J.
      Transcranial focused ultrasound modulates intrinsic and evoked EEG dynamics.
      ]. This corresponds to the lack of TUS-mediated changes in gamma spectral power and functional connectivity in our MEG study. These findings are consistent with TUS-fMRI studies in primates demonstrating that repetitive TUS modulates inter-regional networks in addition to the direct target [
      • Verhagen L.
      • Gallea C.
      • Folloni D.
      • Constans C.
      • Jensen D.E.
      • Ahnine H.
      • et al.
      Offline impact of transcranial focused ultrasound on cortical activation in primates.
      ]. Specific local decoupling of the right basal ganglia in the beta frequency range is thought to reduce inhibitory output [
      • Moran A.
      • Stein E.
      • Tischler H.
      • Bar-Gad I.
      Decoupling neuronal oscillations during subthalamic nucleus stimulation in the parkinsonian primate.
      ], thereby increasing activation and movement-related desynchronization.
      The connectivity analysis showed similar widespread regional changes with both increases and decreases in coupling. LFPs recorded from a rat demonstrated low-frequency delta and theta power increased with tbTUS [
      • Liu Y.
      • Wang G.
      • Cao C.
      • Zhang G.
      • Tanzi E.B.
      • Zhang Y.
      • et al.
      Neuromodulation effect of very low intensity transcranial ultrasound stimulation on multiple nuclei in rat brain.
      ]. Another rat study used EEG to show tbTUS-related increases in alpha, beta, and gamma activity [
      • Yu K.
      • Sohrabpour A.
      • He B.
      Electrophysiological source imaging of brain networks perturbed by low- intensity transcranial focused ultrasound.
      ]. Since our study involved a motor task, decreases in alpha and beta power were most prominent. Moreover, theta event-related desynchronization/synchronization does not have a significant role in the cortical dynamics of voluntary movement [
      • Storti S.F.
      • Formaggio E.
      • Manganotti P.
      • Menegaz G.
      Brain network connectivity and topological analysis during voluntary arm movements.
      ]. It should be noted, however, that spectral power and functional connectivity represent different approaches to assessing cortical dynamics and results may diverge if interregional coupling is modulated without significant changes in regional activation [
      • Storti S.F.
      • Formaggio E.
      • Manganotti P.
      • Menegaz G.
      Brain network connectivity and topological analysis during voluntary arm movements.
      ,
      • Gerloff C.
      • Richard J.
      • Hadley J.
      • Schulman A.E.
      • Honda M.
      • Hallett M.
      Functional coupling and regional activation of human cortical motor areas during simple, internally paced and externally paced finger movements.
      ]. Finally, future studies may be directed at better understanding the relationship between baseline cortical activity and responsiveness to transcranial ultrasound stimulation.
      Some study limitations should be noted. The total MEG recording time was limited to ensure participant compliance and quality of the data, particularly during motor tasks. Despite this, neuromodulation-induced changes in beta activity measured with MEG have been investigated using as little as 20 s of data and our recording times were orders of magnitude greater [
      • Kuhn A.A.
      • Kempf F.
      • Brucke C.
      • Gaynor Doyle L.
      • Martinez-Torres I.
      • Pogosyan A.
      • et al.
      High-frequency stimulation of the subthalamic nucleus suppresses oscillatory beta activity in patients with Parkinson's disease in parallel with improvement in motor performance.
      ]. In addition, our time-locked analyses revealed that the rate of movement during the motor task may have been suboptimal in our ability to detect event-related spectral desynchronization and rebound effects in alpha and beta bands in the sensorimotor region. Future studies may be aimed at specifically investigating the temporal dynamics of neural oscillations following tbTUS. Finally, the long-term durability of tbTUS-mediated neuroplastic changes remains to be determined; however, it is promising that MEP changes following tbTUS are greater than those seen with TMS alone [
      • Zeng K.
      • Darmani G.
      • Fomenko A.
      • Xia X.
      • Tran S.
      • Nankoo J.-F.
      • et al.
      Induction of human motor cortex plasticity by theta burst transcranial ultrasound stimulation.
      ]. Future studies assessing temporal changes may provide greater insight into the regimens needed to exert persistent effects.

      5. Conclusion

      The findings from this study demonstrate that repetitive tbTUS can induce alterations in the human motor cortex. The application of TMS and MEG to study the neurophysiologic basis of these effects led to unique insights into the regional and functional connectivity changes induced by tbTUS. Future applications of this technique may aid in the development of new neuromodulation protocols for neurologic and possibly psychiatric disorders including movement disorders, and motor rehabilitation following traumatic brain injury and stroke.

      CRediT authorship contribution statement

      Nardin Samuel: Conceptualization, Methodology, Investigation, Formal analysis, Writing – review & editing, Visualization. Ke Zeng: Conceptualization, Methodology, Investigation, Formal analysis, Writing – review & editing, Visualization. Irene E. Harmsen: Conceptualization, Methodology, Investigation, Formal analysis, Writing – review & editing, Visualization. Mandy Yi Rong Ding: Validation, Formal analysis, Writing – review & editing, Resources. Ghazaleh Darmani: Validation, Formal analysis, Writing – review & editing, Resources. Can Sarica: Validation, Formal analysis, Writing – review & editing, Resources. Brendan Santyr: Validation, Formal analysis, Writing – review & editing, Resources. Artur Vetkas: Validation, Formal analysis, Writing – review & editing, Resources. Aditya Pancholi: Validation, Formal analysis, Writing – review & editing, Resources. Anton Fomenko: Validation, Formal analysis, Writing – review & editing, Resources. Vanessa Milano: Validation, Formal analysis, Writing – review & editing, Resources. Kazuaki Yamamoto: Validation, Formal analysis, Writing – review & editing, Resources. Utpal Saha: Validation, Formal analysis, Writing – review & editing, Resources. Richard Wennberg: Conceptualization, Methodology, Writing – review & editing, Supervision, Project administration, Funding acquisition. Nathan C. Rowland: Conceptualization, Methodology, Writing – review & editing, Supervision, Project administration, Funding acquisition. Robert Chen: Conceptualization, Methodology, Writing – review & editing, Supervision, Project administration, Funding acquisition. Andres M. Lozano: Conceptualization, Methodology, Writing – review & editing, Supervision, Project administration, Funding acquisition.

      Declaration of competing interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Appendix A. Supplementary data

      References

        • Sarica C.
        • Nankoo J.-F.
        • Fomenko A.
        • Grippe T.C.
        • Yamamoto K.
        • Samuel N.
        • et al.
        Human Studies of Transcranial Ultrasound neuromodulation: a systematic review of effectiveness and safety.
        Brain Stimul. 2022; 15: 737-746https://doi.org/10.1016/j.brs.2022.05.002
        • Darmani G.
        • Bergmann T.O.
        • Butts Pauly K.
        • Caskey C.F.
        • de Lecea L.
        • Fomenko A.
        • et al.
        Non-invasive transcranial ultrasound stimulation for neuromodulation.
        Clin Neurophysiol. 2022; 135: 51-73https://doi.org/10.1016/j.clinph.2021.12.010
        • Fomenko A.
        • Neudorfer C.
        • Dallapiazza R.F.
        • Kalia S.K.
        • Lozano A.M.
        Low-intensity ultrasound neuromodulation: an overview of mechanisms and emerging human applications.
        Brain Stimul. 2018; 11: 1209-1217https://doi.org/10.1016/j.brs.2018.08.013
        • Folloni D.
        • Verhagen L.
        • Mars R.B.
        • Fouragnan E.
        • Constans C.
        • Aubry J.-F.
        • et al.
        Manipulation of subcortical and deep cortical activity in the primate brain using transcranial focused ultrasound stimulation.
        Neuron. 2019; 101 (e5): 1109-1116https://doi.org/10.1016/j.neuron.2019.01.019
        • Fouragnan E.F.
        • Chau B.K.H.
        • Folloni D.
        • Kolling N.
        • Verhagen L.
        • Klein-Flugge M.
        • et al.
        The macaque anterior cingulate cortex translates counterfactual choice value into actual behavioral change.
        Nat Neurosci. 2019; 22: 797-808https://doi.org/10.1038/s41593-019-0375-6
        • Verhagen L.
        • Gallea C.
        • Folloni D.
        • Constans C.
        • Jensen D.E.
        • Ahnine H.
        • et al.
        Offline impact of transcranial focused ultrasound on cortical activation in primates.
        Elife. 2019; 8e40541https://doi.org/10.7554/eLife.40541
        • Zeng K.
        • Darmani G.
        • Fomenko A.
        • Xia X.
        • Tran S.
        • Nankoo J.-F.
        • et al.
        Induction of human motor cortex plasticity by theta burst transcranial ultrasound stimulation.
        Ann Neurol. 2022; 91: 238-252https://doi.org/10.1002/ana.26294
        • Yoo S.-S.
        • Kim H.
        • Min B.-K.
        • Franck E.
        • Park S.
        Transcranial focused ultrasound to the thalamus alters anesthesia time in rats.
        Neuroreport. 2011; 22: 783-787https://doi.org/10.1097/WNR.0b013e32834b2957
        • Kim H.
        • Park M.-A.
        • Wang S.
        • Chiu A.
        • Fischer K.
        • Yoo S.-S.
        PETCT imaging evidence of FUS-mediated (18)F-FDG uptake changes in rat brain.
        Med Phys. 2013; 40033501https://doi.org/10.1118/1.4789916
        • Legon W.
        • Sato T.F.
        • Opitz A.
        • Mueller J.
        • Barbour A.
        • Williams A.
        • et al.
        Transcranial focused ultrasound modulates the activity of primary somatosensory cortex in humans.
        Nat Neurosci. 2014; 17: 322-329https://doi.org/10.1038/nn.3620
        • Mueller J.
        • Legon W.
        • Opitz A.
        • Sato T.F.
        • Tyler W.J.
        Transcranial focused ultrasound modulates intrinsic and evoked EEG dynamics.
        Brain Stimul. 2014; 7: 900-908https://doi.org/10.1016/j.brs.2014.08.008
        • Wattiez N.
        • Constans C.
        • Deffieux T.
        • Daye P.M.
        • Tanter M.
        • Aubry J.-F.
        • et al.
        Transcranial ultrasonic stimulation modulates single-neuron discharge in macaques performing an antisaccade task.
        Brain Stimul. 2017; 10: 1024-1031https://doi.org/10.1016/j.brs.2017.07.007
        • Guo H.
        • Hamilton 2nd, M.
        • Offutt S.J.
        • Gloeckner C.D.
        • Li T.
        • Kim Y.
        • et al.
        Ultrasound produces extensive brain activation via a cochlear pathway.
        Neuron. 2018; 98 (e4): 1020-1030https://doi.org/10.1016/j.neuron.2018.04.036
        • Tadel F.
        • Baillet S.
        • Mosher J.C.
        • Pantazis D.
        • Leahy R.M.
        Brainstorm: a user-friendly application for MEG/EEG analysis.
        Comput. Intell. Neurosci. 2011; : 1-13https://doi.org/10.1155/2011/879716
        • Stokes M.G.
        • Chambers C.D.
        • Gould I.C.
        • Henderson T.R.
        • Janko N.E.
        • Allen N.B.
        • et al.
        Simple metric for scaling motor threshold based on scalp-cortex distance: application to studies using transcranial magnetic stimulation.
        J Neurophysiol. 2005; 94: 4520-4527https://doi.org/10.1152/jn.00067.2005
        • Taulu S.
        • Hari R.
        Removal of magnetoencephalographic artifacts with temporal signal-space separation: demonstration with single-trial auditory-evoked responses.
        Hum Brain Mapp. 2009; 30: 1524-1534https://doi.org/10.1002/hbm.20627
        • Wennberg R.
        • Del Campo J.M.
        • Shampur N.
        • Rowland N.C.
        • Valiante T.
        • Lozano A.M.
        • et al.
        Feasibility of magnetoencephalographic source imaging in patients with thalamic deep brain stimulation for epilepsy.
        Epilepsia Open. 2017; 2: 101-106https://doi.org/10.1002/epi4.12027
        • Kandemir A.L.
        • Litvak V.
        • Florin E.
        The comparative performance of DBS artefact rejection methods for MEG recordings.
        Neuroimage. 2020; 219117057https://doi.org/10.1016/j.neuroimage.2020.117057
        • 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; 156869https://doi.org/10.1155/2011/156869
        • Thomson D.J.
        Spectrum estimation and harmonic analysis.
        Proc IEEE. 1982; 70: 1055-1096
        • Gross J.
        • Kujala J.
        • Hamalainen M.
        • Timmermann L.
        • Schnitzler A.
        • Salmelin R.
        Dynamic imaging of coherent sources: studying neural interactions in the human brain.
        Proc Natl Acad Sci U S A. 2001; 98: 694-699https://doi.org/10.1073/pnas.98.2.694
        • Tzourio-Mazoyer N.
        • Landeau B.
        • Papathanassiou D.
        • Crivello F.
        • Etard O.
        • Delcroix N.
        • et al.
        Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain.
        Neuroimage. 2002; 15: 273-289https://doi.org/10.1006/nimg.2001.0978
        • Collins D.L.
        • Holmes C.J.
        • Peters T.M.
        • Evans A.C.
        Automatic 3-D model-based neuroanatomical segmentation.
        Hum Brain Mapp. 1995; 3: 190-208https://doi.org/10.1002/hbm.460030304
        • Jenkinson M.
        • Beckmann C.F.
        • Behrens T.E.J.
        • Woolrich M.W.
        • Smith S.M.
        FSL. Neuroimage. 2012; 62: 782-790https://doi.org/10.1016/j.neuroimage.2011.09.015
        • Hillebrand A.
        • Singh K.D.
        • Holliday I.E.
        • Furlong P.L.
        • Barnes G.R.
        A new approach to neuroimaging with magnetoencephalography.
        Hum Brain Mapp. 2005; 25: 199-211https://doi.org/10.1002/hbm.20102
        • Van Veen B.D.
        • van Drongelen W.
        • Yuchtman M.
        • Suzuki A.
        Localization of brain electrical activity via linearly constrained minimum variance spatial filtering.
        IEEE Trans Biomed Eng. 1997; 44: 867-880https://doi.org/10.1109/10.623056
        • Rosenberg J.R.
        • Amjad A.M.
        • Breeze P.
        • Brillinger D.R.
        • Halliday D.M.
        The Fourier approach to the identification of functional coupling between neuronal spike trains.
        Prog Biophys Mol Biol. 1989; 53: 1-31https://doi.org/10.1016/0079-6107(89)90004-7
        • Kwak S.G.
        • Kim J.H.
        Central limit theorem: the cornerstone of modern statistics.
        Kor. J. Anesthesiol. 2017; 70: 144-156https://doi.org/10.4097/kjae.2017.70.2.144
        • Kiebel S.J.
        • Tallon-Baudry C.
        • Friston K.J.
        Parametric analysis of oscillatory activity as measured with EEG/MEG.
        Hum Brain Mapp. 2005; 26: 170-177https://doi.org/10.1002/hbm.20153
      1. Friston K.J. Statistical parametric mapping: the analysis of funtional brain images. first ed. Elsevier/Academic Press, Amsterdam ; Boston2007
        • Benjamini Y.
        • Hochberg Y.
        Controlling the false discovery rate: a practical and powerful approach to multiple testing.
        J Roy Stat Soc B. 1995; 57: 289-300https://doi.org/10.1111/j.2517-6161.1995.tb02031.x
        • Larson J.
        • Wong D.
        • Lynch G.
        Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation.
        Brain Res. 1986; 368: 347-350https://doi.org/10.1016/0006-8993(86)90579-2
        • Arif Y.
        • Spooner R.K.
        • Heinrichs-Graham E.
        • Wilson T.W.
        High-definition transcranial direct current stimulation modulates performance and alpha/beta parieto-frontal connectivity serving fluid intelligence.
        J Physiol. 2021; 599: 5451-5463https://doi.org/10.1113/JP282387
        • Mantini D.
        • Perrucci M.G.
        • Del Gratta C.
        • Romani G.L.
        • Corbetta M.
        Electrophysiological signatures of resting state networks in the human brain.
        Proc Natl Acad Sci USA. 2007; 104: 13170-13175https://doi.org/10.1073/pnas.0700668104
        • Brookes M.J.
        • Woolrich M.
        • Luckhoo H.
        • Price D.
        • Hale J.R.
        • Stephenson M.C.
        • et al.
        Investigating the electrophysiological basis of resting state networks using magnetoencephalography.
        Proc Natl Acad Sci USA. 2011; 108: 16783-16788https://doi.org/10.1073/pnas.1112685108
        • Pool E.-M.
        • Rehme A.K.
        • Eickhoff S.B.
        • Fink G.R.
        • Grefkes C.
        Functional resting-state connectivity of the human motor network: differences between right- and left-handers.
        Neuroimage. 2015; 109: 298-306https://doi.org/10.1016/j.neuroimage.2015.01.034
        • Brookes M.J.
        • Hall E.L.
        • Robson S.E.
        • Price D.
        • Palaniyappan L.
        • Liddle E.B.
        • et al.
        Complexity measures in magnetoencephalography: measuring “disorder” in schizophrenia.
        PLoS One. 2015; 10e0120991https://doi.org/10.1371/journal.pone.0120991
        • Moran A.
        • Stein E.
        • Tischler H.
        • Bar-Gad I.
        Decoupling neuronal oscillations during subthalamic nucleus stimulation in the parkinsonian primate.
        Neurobiol Dis. 2012; 45: 583-590https://doi.org/10.1016/j.nbd.2011.09.016
        • Liu Y.
        • Wang G.
        • Cao C.
        • Zhang G.
        • Tanzi E.B.
        • Zhang Y.
        • et al.
        Neuromodulation effect of very low intensity transcranial ultrasound stimulation on multiple nuclei in rat brain.
        Front Aging Neurosci. 2021; 13656430https://doi.org/10.3389/fnagi.2021.656430
        • Yu K.
        • Sohrabpour A.
        • He B.
        Electrophysiological source imaging of brain networks perturbed by low- intensity transcranial focused ultrasound.
        IEEE Trans Biomed Eng. 2016; 63: 1787-1794https://doi.org/10.1109/TBME.2016.2591924
        • Storti S.F.
        • Formaggio E.
        • Manganotti P.
        • Menegaz G.
        Brain network connectivity and topological analysis during voluntary arm movements.
        Clin EEG Neurosci. 2016; 47: 276-290https://doi.org/10.1177/1550059415598905
        • Gerloff C.
        • Richard J.
        • Hadley J.
        • Schulman A.E.
        • Honda M.
        • Hallett M.
        Functional coupling and regional activation of human cortical motor areas during simple, internally paced and externally paced finger movements.
        Brain. 1998; 121: 1513-1531https://doi.org/10.1093/brain/121.8.1513
        • Kuhn A.A.
        • Kempf F.
        • Brucke C.
        • Gaynor Doyle L.
        • Martinez-Torres I.
        • Pogosyan A.
        • et al.
        High-frequency stimulation of the subthalamic nucleus suppresses oscillatory beta activity in patients with Parkinson's disease in parallel with improvement in motor performance.
        J Neurosci. 2008; 28: 6165-6173https://doi.org/10.1523/JNEUROSCI.0282-08.2008