EEG Artifact Removal in TMS Studies of Cortical Speech Areas

Real-time optimization to enhance noninvasive cortical excitability assessment in the human dorsolateral prefrontal cortex

Objective

We currently lack a robust noninvasive method to measure prefrontal excitability in humans. Concurrent TMS and EEG in the prefrontal cortex is usually confounded by artifacts. Here we asked if real-time optimization could reduce artifacts and enhance a TMS-EEG measure of left prefrontal excitability.

Methods

This closed-loop optimization procedure adjusts left dlPFC TMS coil location, angle, and intensity in real-time based on the EEG response to TMS. Our outcome measure was the left prefrontal early (20-60 ms) and local TMS-evoked potential (EL-TEP).

Results

In 18 healthy participants, this optimization of coil angle and brain target significantly reduced artifacts by 63% and, when combined with an increase in intensity, increased EL-TEP magnitude by 75% compared to a non-optimized approach.

Conclusions

Real-time optimization of TMS parameters during dlPFC stimulation can enhance the EL-TEP.

Significance

Enhancing our ability to measure prefrontal excitability is important for monitoring pathological states and treatment response.

Changes in Cortical Activation by Transcranial Magnetic Stimulation Due to Coil Rotation Are Not Attributable to Cranial Muscle Activation

Transcranial magnetic stimulation coupled with electroencephalography (TMS-EEG) allows for the study of brain dynamics in health and disease. Cranial muscle activation can decrease the interpretability of TMS-EEG signals by masking genuine EEG responses and increasing the reliance on preprocessing methods but can be at least partly prevented by coil rotation coupled with the online monitoring of signals; however, the extent to which changing coil rotation may affect TMS-EEG signals is not fully understood. Our objective was to compare TMS-EEG data obtained with an optimal coil rotation to induce motor evoked potentials (M1standard) while rotating the coil to minimize cranial muscle activation (M1emg). TMS-evoked potentials (TEPs), TMS-related spectral perturbation (TRSP), and intertrial phase clustering (ITPC) were calculated in both conditions using two different preprocessing pipelines based on independent component analysis (ICA) or signal-space projection with source-informed reconstruction (SSP-SIR). Comparisons were performed with cluster-based correction. The concordance correlation coefficient was computed to measure the similarity between M1standard and M1emg TMS-EEG signals. TEPs, TRSP, and ITPC were significantly larger in M1standard than in M1emg conditions; a lower CCC than expected was also found. These results were similar across the preprocessing pipelines. While rotating the coil may be advantageous to reduce cranial muscle activation, it may result in changes in TMS-EEG signals; therefore, this solution should be tailored to the specific experimental context.

Cleaning up the Brickyard: How Theory and Methodology Shape Experiments in Cognitive Neuroscience of Language

The capacity for language is a defining property of our species, yet despite decades of research, evidence on its neural basis is still mixed and a generalized consensus is difficult to achieve. We suggest that this is partly caused by researchers defining "language" in different ways, with focus on a wide range of phenomena, properties, and levels of investigation. Accordingly, there is very little agreement among cognitive neuroscientists of language on the operationalization of fundamental concepts to be investigated in neuroscientific experiments. Here, we review chains of derivation in the cognitive neuroscience of language, focusing on how the hypothesis under consideration is defined by a combination of theoretical and methodological assumptions. We first attempt to disentangle the complex relationship between linguistics, psychology, and neuroscience in the field. Next, we focus on how conclusions that can be drawn from any experiment are inherently constrained by auxiliary assumptions, both theoretical and methodological, on which the validity of conclusions drawn rests. These issues are discussed in the context of classical experimental manipulations as well as study designs that employ novel approaches such as naturalistic stimuli and computational modeling. We conclude by proposing that a highly interdisciplinary field such as the cognitive neuroscience of language requires researchers to form explicit statements concerning the theoretical definitions, methodological choices, and other constraining factors involved in their work.

Somatosensory input in the context of transcranial magnetic stimulation coupled with electroencephalography: An evidence-based overview

The transcranial evoked potential (TEP) is a powerful technique to investigate brain dynamics, but some methodological issues limit its interpretation. A possible contamination of the TEP by electroencephalographic (EEG) responses evoked by the somatosensory input generated by transcranial magnetic stimulation (TMS) has been postulated; nonetheless, a characterization of these responses is lacking. The aim of this work was to review current evidence about possible somatosensory evoked potentials (SEP) induced by sources of somatosensory input in the craniofacial region. Among these, only contraction of craniofacial muscle and stimulation of free cutaneous nerve endings may be able to induce EEG responses, but direct evidence is lacking due to experimental difficulties in isolating these inputs. Notably, EEG evoked activity in this context is represented by a N100/P200 complex, reflecting a saliency-related multimodal response, rather than specific activation of the primary somatosensory cortex. Strategies to minimize or remove these responses by EEG processing still yield uncertain results; therefore, data inspection is of paramount importance to judge a possible contamination of the TEP by multimodal potentials caused by somatosensory input.

Causal evidence for a coordinated temporal interplay within the language network

Recent neurobiological models on language suggest that auditory sentence comprehension is supported by a coordinated temporal interplay within a left-dominant brain network, including the posterior inferior frontal gyrus (pIFG), posterior superior temporal gyrus and sulcus (pSTG/STS), and angular gyrus (AG). Here, we probed the timing and causal relevance of the interplay between these regions by means of concurrent transcranial magnetic stimulation and electroencephalography (TMS-EEG). Our TMS-EEG experiments reveal region- and time-specific causal evidence for a bidirectional information flow from left pSTG/STS to left pIFG and back during auditory sentence processing. Adapting a condition-and-perturb approach, our findings further suggest that the left pSTG/STS can be supported by the left AG in a state-dependent manner.

Motor potentials evoked by transcranial magnetic stimulation: interpreting a simple measure of a complex system

Transcranial magnetic stimulation (TMS) is a non-invasive technique that is increasingly used to study the human brain. One of the principal outcome measures is the motor-evoked potential (MEP) elicited in a muscle following TMS over the primary motor cortex (M1), where it is used to estimate changes in corticospinal excitability. However, multiple elements play a role in MEP generation, so even apparently simple measures such as peak-to-peak amplitude have a complex interpretation. Here, we summarize what is currently known regarding the neural pathways and circuits that contribute to the MEP and discuss the factors that should be considered when interpreting MEP amplitude measured at rest in the context of motor processing and patients with neurological conditions. In the last part of this work, we also discuss how emerging technological approaches can be combined with TMS to improve our understanding of neural substrates that can influence MEPs. Overall, this review aims to highlight the capabilities and limitations of TMS that are important to recognize when attempting to disentangle sources that contribute to the physiological state-related changes in corticomotor excitability.

Source-based artifact-rejection techniques for TMS-EEG

Neuronal electroencephalography (EEG) signals arise from the cortical postsynaptic currents. Due to the conductive properties of the head, these neuronal sources produce relatively smeared spatial patterns in EEG. We can model these topographies to deduce which signals reflect genuine TMS-evoked cortical activity and which data components are merely noise and artifacts. This review will concentrate on two source-based artifact-rejection techniques developed for TMS-EEG data analysis, signal-space-projection-source-informed reconstruction (SSP-SIR), and the source-estimate-utilizing noise-discarding algorithm (SOUND). The former method was designed for rejecting TMS-evoked muscle artifacts, while the latter was developed to suppress noise signals from EEG and magnetoencephalography (MEG) in general. We shall cover the theoretical background for both methods, but most importantly, we will describe some essential practical perspectives for using these techniques effectively. We demonstrate and explain what approaches produce the most reliable inverse estimates after cleaning the data or how to perform non-biased comparisons between cleaned datasets. All noise-cleaning algorithms compromise the signals of interest to a degree. We elaborate on how the source-based methods allow objective quantification of the overcorrection. Finally, we consider possible future directions. While this article concentrates on TMS-EEG data analysis, many theoretical and practical aspects, presented here, can be readily applied in other EEG/MEG applications. Overall, the source-based cleaning methods provide a valuable set of TMS-EEG preprocessing tools. We can objectively evaluate their performance regarding possible overcorrection. Furthermore, the overcorrection can always be taken into account to compare cleaned datasets reliably. The described methods are based on current electrophysiological and anatomical understanding of the head and the EEG generators; strong assumptions of the statistical properties of the noise and artifact signals, such as independence, are not needed.

Online neurostimulation of Broca's area does not interfere with syntactic predictions: A combined TMS-EEG approach to basic linguistic combination

Categorical predictions have been proposed as the key mechanism supporting the fast pace of syntactic composition in language. Accordingly, grammar-based expectations are formed-e.g., the determiner "a" triggers the prediction for a noun-and facilitate the analysis of incoming syntactic information, which is then checked against a single or few other word categories. Previous functional neuroimaging studies point towards Broca's area in the left inferior frontal gyrus (IFG) as one fundamental cortical region involved in categorical prediction during incremental language processing. Causal evidence for this hypothesis is however still missing. In this study, we combined Electroencephalography (EEG) and Transcranial Magnetic Stimulation (TMS) to test whether Broca's area is functionally relevant in predictive mechanisms for language. We transiently perturbed Broca's area during the first word in a two-word construction, while simultaneously measuring the Event-Related Potential (ERP) correlates of syntactic composition. We reasoned that if Broca's area is involved in predictive mechanisms for syntax, disruptive TMS during the first word would mitigate the difference in the ERP responses for predicted and unpredicted categories in basic two-word constructions. Contrary to this hypothesis, perturbation of Broca's area at the predictive stage did not affect the ERP correlates of basic composition. The correlation strength between the electrical field induced by TMS and the ERP responses further confirmed this pattern. We discuss the present results considering an alternative account of the role of Broca's area in syntactic composition, namely the bottom-up integration of words into constituents, and of compensatory mechanisms within the language predictive network.

Identification and verification of a 'true' TMS evoked potential in TMS-EEG

The concurrent combination of transcranial magnetic stimulation and electroencephalography (TMS-EEG) can unveil functional neural mechanisms with applications in basic and clinical research. In particular, TMS-evoked potentials (TEPs) potentially allow studying excitability and connectivity of the cortex in a causal manner that is not easily or non-invasively attainable with other neuroimaging techniques. The TEP waveform is obtained by isolating the EEG responses phase-locked to the time of TMS application. The intended component in a TEP waveform is the cortical activation by the TMS-induced electric current, free of instrumental and physiological artifact sources. This artifact-free cortical activation can be referred to as 'true' TEP. However, due to many unwanted auxiliary effects of TMS, the interpretation of 'true' TEPs has not been free of controversy. This paper reviews the most recent understandings of 'true' TEPs and their application. In the first part of the paper, TEP components are defined according to recommended methodologies. In the second part, the verification of 'true' TEP is discussed along with its sensitivity to brain-state, age, and disease. The various proposed origins of TEP components are then presented in the context of existing literature. Throughout the paper, lessons learned from the past TMS-EEG studies are highlighted to guide the identification and interpretation of 'true' TEPs in future studies.

Removing artifacts from TMS-evoked EEG: A methods review and a unifying theoretical framework

Transcranial magnetic stimulation (TMS) combined with electroencephalography (EEG) is a technique for studying cortical excitability and connectivity in health and disease, allowing basic research and potential clinical applications. A major methodological issue, severely limiting the applicability of TMS-EEG, relates to the contamination of EEG signals by artifacts of biologic or non-biologic origin. To solve this problem, several methods, based on independent component analysis (ICA), principal component analysis (PCA), signal space projection (SSP), and other approaches, have been developed over the last decade. This article is divided into two parts. In the first part, we review the theoretical background of the currently available TMS-EEG artifact removal methods. In the second part, we formally introduce the mathematics underpinnings of the cleaning methods. We classify them into spatial and temporal filters based on their properties. Since the most frequently used TMS-EEG cleaning approach are spatial filter methods, we focus on them and introduce beamforming as a unified framework of the most popular spatial filtering techniques. This unifying approach enables the comparative assessment of these methods by highlighting their differences in terms of assumptions, challenges, and applicability for different types of artifacts and data. The different properties and challenges of the methods discussed are illustrated with both simulated and recorded data. This article targets non-mathematical and mathematical audiences. Accordingly, those readers interested in essential background information on these methods can focus on Section 2. Whereas theory-oriented readers may find Section 3 helpful for making informed decisions between existing methods and developing the methodology further.

Designing and comparing cleaning pipelines for TMS-EEG data: a theoretical overview and practical example

Combining transcranial magnetic stimulation (TMS) with electroencephalography (EEG) is growing in popularity as a method for probing the reactivity and connectivity of neural circuits in basic and clinical research. However, using EEG to measure the neural responses to TMS is challenging due to the unique artifacts introduced by combining the two techniques. In this paper, we overview the artifacts present in TMS-EEG data and the offline cleaning methods used to suppress these unwanted signals. We then describe how open science practices, including the development of open-source toolboxes designed for TMS-EEG analysis (e.g., TESA - the TMS-EEG signal analyser), have improved the availability and reproducibility of TMS-EEG cleaning methods. We provide theoretical and practical considerations for designing TMS-EEG cleaning pipelines and then give an example of how to compare different pipelines using TESA. We show that changing even a single step in a pipeline designed to suppress decay artifacts results in TMS-evoked potentials (TEPs) with small differences in amplitude and spatial topography. The variability in TEPs resulting from the choice of cleaning pipeline has important implications for comparing TMS-EEG findings between research groups which use different online and offline approaches. Finally, we discuss the challenges of validating cleaning pipelines and recommend that researchers compare outcomes from TMS-EEG experiments using multiple pipelines to ensure findings are not related to the choice of cleaning methods. We conclude that the continued improvement, availability, and validation of cleaning pipelines is essential to ensure TMS-EEG reaches its full potential as a method for studying human neurophysiology.

Disentangling EEG responses to TMS due to cortical and peripheral activations

BACKGROUND

the use of combined transcranial magnetic stimulation (TMS) and electroencephalography (EEG) for the functional evaluation of the cerebral cortex in health and disease is becoming increasingly common. However, there is still some ambiguity regarding the extent to which brain responses to auditory and somatosensory stimulation contribute to the TMS-evoked potential (TEP).

OBJECTIVE/HYPOTHESIS

to measure separately the contribution of auditory and somatosensory stimulation caused by TMS, and to assess their contribution to the TEP waveform, when stimulating the motor cortex (M1).

METHODS

19 healthy volunteers underwent 7 blocks of EEG recording. To assess the impact of auditory stimulation on the TEP waveform, we used a standard figure of eight coil, with or without masking with a continuous noise reproducing the specific time-varying frequencies of the TMS click, stimulating at 90% of resting motor threshold. To further characterise auditory responses due to the TMS click, we used either a standard or a sham figure of eight coil placed on a pasteboard cylinder that rested on the scalp, with or without masking. Lastly, we used electrical stimulation of the scalp to investigate the possible contribution of somatosensory activation.

RESULTS

auditory stimulation induced a known pattern of responses in electrodes located around the vertex, which could be suppressed by appropriate noise masking. Electrical stimulation of the scalp alone only induced similar, non-specific scalp responses in the in the central electrodes. TMS, coupled with appropriate masking of sensory input, resulted in specific, lateralized responses at the stimulation site, lasting around 300 ms.

CONCLUSIONS

if careful control of confounding sources is applied, TMS over M1 can generate genuine, lateralized EEG activity. By contrast, sensory evoked responses, if present, are represented by non-specific, late (100-200 ms) components, located at the vertex, possibly due to saliency of the stimuli. Notably, the latter can confound the TEP if masking procedures are not properly used.

Scalable Implementation of Hippocampal Network on Digital Neuromorphic System towards Brain-Inspired Intelligence

In this paper, an expanded digital hippocampal spurt neural network (HSNN) is innovatively proposed to simulate the mammalian cognitive system and to perform the neuroregulatory dynamics that play a critical role in the cognitive processes of the brain, such as memory and learning. The real-time computation of a large-scale peak neural network can be realized by the scalable on-chip network and parallel topology. By exploring the latest research in the field of neurons and comparing with the results of this paper, it can be found that the implementation of the hippocampal neuron model using the coordinate rotation numerical calculation algorithm can significantly reduce the cost of hardware resources. In addition, the rational use of on-chip network technology can further improve the performance of the system, and even significantly improve the network scalability on a single field programmable gate array chip. The neuromodulation dynamics are considered in the proposed system, which can replicate more relevant biological dynamics. Based on the analysis of biological theory and the theory of hardware integration, it is shown that the innovative system proposed in this paper can reproduce the biological characteristics of the hippocampal network and may be applied to brain-inspired intelligent subjects. The study in this paper will have an unexpected effect on the future research of digital neuromorphic design of spike neural network and the dynamics of the hippocampal network.