Characterizing and Modulating Brain Circuitry through Transcranial Magnetic Stimulation Combined with Electroencephalography

Abstract

The concurrent combination of transcranial magnetic stimulation (TMS) with electroencephalography (TMS-EEG) is a powerful technology for characterizing and modulating brain networks across developmental, behavioral, and disease states. Given the global initiatives in mapping the human brain, recognition of the utility of this technique is growing across neuroscience disciplines. Importantly, TMS-EEG offers translational biomarkers that can be applied in health and disease, across the lifespan, and in humans and animals, bridging the gap between animal models and human studies. However, to utilize the full potential of TMS-EEG methodology, standardization of TMS-EEG study protocols is needed. In this article, we review the principles of TMS-EEG methodology, factors impacting TMS-EEG outcome measures, and the techniques for preventing and correcting artifacts in TMS-EEG data. To promote the standardization of this technique, we provide comprehensive guides for designing TMS-EEG studies and conducting TMS-EEG experiments. We conclude by reviewing the application of TMS-EEG in basic, cognitive and clinical neurosciences, and evaluate the potential of this emerging technology in brain research.

Keywords: TMS-EEG; biomarker discovery; brain mapping; electroencephalography (EEG); experiment design; neuromodulation; signal processing; transcranial magnetic stimulation.

Figure 1

The basic principle of transcranial magnetic stimulation. Figure depicts the schematics of TMS-evoked potentials (TEP)s (A), TMS-induced descend volleys (B), and TMS-induced motor-evoked potential (MEP) (C) when TMS is applied to the motor cortex. (A) The waveform illustrates the average TEPs recorded through electroencephalography (EEG) from a hypothetical EEG sensor close to the vertex. When applied to the motor cortex, several negative (N) and positive (P) TEP components are reported at different latencies (in ms) relative to the time of TMS delivery, including the N15, P30, N45, P60, N100, P180, and N280 (reviewed in Komssi and Kahkonen, 2006). Several of these peaks are associated with activation of specific excitatory and inhibitory neural circuitries. The TMS coil schematic also depicts that the TMS-induced magnetic field (B) is perpendicular to the plane of the TMS coil. The induced electric field in the tissue (E) is in turn perpendicular to the magnetic field. The direction of the induced current in the tissue is anti-parallel to the direction of the current in the coil [E arrow going into the page or coming out]. (B) Waveform illustrates the schematic of the TMS-induced descending volleys that can be recorded from the spinal cord, and the direct (D) and indirect (I) waves that are induced in the corticospinal tract depending on the TMS pulse intensity and coil orientation. The D and I waves are associated with direct or transsynaptic activation of pyramidal neurons, respectively. (C) The waveform depicts an MEP recorded from a peripheral muscle through electromyography (EMG). The latency and the peak-to-peak amplitude of the MEP are conventionally employed to examine the integrity of the corticospinal tract. (D) Figure depicts the simplified system diagram of a TMS coil (L) attached to a TMS device main unit. The TMS main unit often consists of a Voltage (V) source, Switch (S), Capacitor (C), Diode (D), Resistor (R), and Thyristor (T). (E) Waveforms are schematics of monophasic (black) vs. biphasic (blue) TMS pulse shapes, here illustrated by the current (Ampere) in the coil. (F) Figures demonstrate two popular coil shapes, the figure-of-eight (top) and the circular (bottom) coil shape. Figures also illustrate the relationship between the direction of current in the coil (black dotted arrows) and the direction of the current induced in the brain tissue (red arrow) which is anti-parallel to the coil current. Please note that for the circular coil placed on the vertex that has anti-clockwise coil current, the induced current in the tissue would be clockwise. Therefore, current in the tissue would be anteriorly oriented on the left hemisphere and posteriorly oriented on the right hemisphere.

Figure 2

A system diagram guiding the design of TMS-EEG studies. The figure presents a system diagram of three main types of parameters that can be selected when designing a TMS-EEG experiment: TMS input parameters (location, protocol, and time), EEG output parameters (location, measures, and time), and brain state parameters (developmental, behavioral, dynamical, and disease states). Several possible choices are listed for each parameter and where applicable mechanism or neurobiology associated with each parameter is specified (e.g., for TMS input protocol and EEG output measures). The figure also depicts the possibility of using the EEG output parameters to guide the TMS input parameters, either offline or through online feedback systems.

Figure 3

A summarized step-by-step guideline for carrying out a TMS-EEG experiment. The figure is a summary of major steps taken in conducting a TMS-EEG experiment. These steps include: (1) selection of TMS input parameters (e.g., input location, protocols, and time) as depicted in Figure 2 and described in Section TMS Parameters, (2) making note of or controlling the brain state as described in Section Brain State, (3) choosing TMS-compatible EEG systems as described in Section TMS-EEG Equipment, (4) proper EEG cap preparation for minimizing induction of TMS artifacts as described in Section Sensor Placement, Sensor Wire Arrangement and Orientation, and Sensor-Skin and Skin-Skin Impedance, (5) incorporating appropriate experimental conditions as control conditions described in Section Controlling for TMS Click, Controlling for Scalp Sensation, Controlling for Attention, and Other Control Conditions (e.g., including sham conditions, masking TMS loud clicking sound, and vibration), (6) following several considerations during data acquisition to prevent induction of noise (e.g., placement of a layer of foam between coil and electrodes) or to properly adapt TMS protocols and enhance signal to noise ratio by using large number of stimuli per condition as described in Section Implementation of TMS Protocols during EEG Recording and Preventing TMS-Related EEG Artifacts, (7) following recommended guidelines for data processing, involving first removing large amplitude TMS-related artifacts such as TMS pulse artifact or TMS induced EMG before application of filters as described in Section Correcting TMS-EEG Artifacts, (8) selection of appropriate EEG output parameters (e.g., input location, protocols, and time) as depicted in Figure 2 and described in section EEG Analysis, and finally (9) choosing an appropriate statistical framework suitable for the characteristics of the multidimensional TMS-EEG outcomes (e.g., based on the data dimensions, or distribution characteristics; Frehlich et al., 2016) also described in Section Correcting TMS-EEG Artifacts.

Figure 4

A schematic diagram of translational value of TMS-EEG. (A) The method of TMS-EEG provides a means to non-invasively assess the integrity and characteristics of numerous brain circuitries in the intact human brain using an input (TMS)-output (EEG) approach. Equivalent in vitro, animal, genetic, or computational modeling studies can further provide insight into the link between genes, brain function, and behavior. (B) The blue waveform illustrates schematics of typical TMS-evoked potential (TEP) when suprathreshold single-pulse TMS is applied to the primary motor cortex. Various characteristics of TEP such as amplitude or latency of components [e.g., negativity at latency 15 ms (N15), positivity at latency 30 ms (P30), N45, P60, N100, P180, N280] are highlighted. The scatter plots are schematic illustrations of the link between TMS-EEG features and genetic variations (left panel) or behavior (right). (C) The waveforms highlight change in TEPs for two hypothetical brain states (e.g., before and after an intervention). The scatter plots are schematic illustrations of the link between change in TMS-EEG features and genetic variations (left panel) or change in behavior (right panel).