State-dependency of transcranial magnetic stimulation

Abstract

Transcranial magnetic stimulation (TMS), a tool that allows noninvasive modulation of cortical neural activity, has become an important tool in cognitive neuroscience and is being increasingly explored in neurotherapeutics. Amongst the factors that are likely to influence its efficacy, the importance of the baseline cortical activation state on the impact of TMS has not received much attention. However, this state-dependency is important as the neural impact of any external stimulus represents an interaction with the ongoing brain activity at the time of stimulation. The effects of any external stimulus are therefore not only determined by the properties of that stimulus but also by the activation state of the brain. Here we review the existing evidence on the state-dependency of TMS and propose how its systematic study can provide unique insights into brain function and significantly enhance the effectiveness of TMS in investigations on the neural basis of perception and cognition. We also describe novel approaches based on this state-dependency which can be used to investigate the properties of distinct neural subpopulations within the stimulated region. Furthermore, we discuss how state-dependency can explain the functional mechanisms through which TMS impairs perception and behavior.

Fig. 1

A schematic of the state-dependency of TMS effects, depicting the principle of TMS preferentially facilitating attributes encoded by the less active neural populations. The same outcome is observed when the initial cortical activation state has been manipulated by either adaptation or priming prior to application of TMS. (a) After adaptation to leftward motion, neurons tuned to this direction are less active than neurons tuned to rightward motion. Perceptually, this adaptation is manifested as a bias to perceive a subsequent moving stimulus moving in the opposite, rightward direction (panel 2). Application of TMS over the visual area V5/MT reverses this bias, with subjects more likely to report a test stimulus moving in the adapted direction. (b) After priming to leftward motion, neurons tuned to this direction are more active than neurons tuned to rightward motion. Perceptually, this priming is manifested as a bias to perceive a subsequent motion stimulus to move in the primed direction (panel 2). Application of TMS over the visual area V5/MT reverses this bias, with subjects now more likely to report a test stimulus moving in the nonadapted (rightward) direction

Fig. 2

The TMS-adaptation paradigm. (a) Each test block begins with a period of adaptation, with the objective of differentially affecting the initial states of functionally distinct neural populations. In this study by Cattaneo and Silvanto (2008b), subjects were adapted to either leftward or rightward moving motion stimulus. Adaptation is followed by a block of experimental trials; in this study subjects were presented with motion stimuli moving either in the adapted direction (i.e., congruent targets) or targets moving in the opposite targets (i.e., incongruent trials). At least 24 experimental trials can be run after each period of adaptation without TMS significantly weakening the strength of adaptation (Cattaneo and Silvanto 2008b). (b) The state-dependent effect of TMS. In the No TMS condition, subjects were worse in detecting the adapted direction (i.e. congruent trials) relative to the nonadapted direction (i.e., incongruent trials), demonstrating that adaptation was behaviorally effective. This effect was reversed when TMS was applied over the motion-selective area V5/MT (a region known to contain motion-selective neurons), with TMS facilitating the detection of the adapted stimuli relative to the nonadapted stimuli. As neurons encoding the adapted direction of motion were less active than neurons encoding other attributes at the time of stimulation, this finding implies that TMS behaviorally facilitates the less active neural populations

Fig. 3

TMS-priming paradigm (adapted from Cattaneo et al. in press). On each experimental trial, a priming stimulus is presented with the objective of differentially affecting the initial states of functionally distinct neural populations prior to application of TMS and presentation of the test stimulus. In this study by Cattaneo et al. (in press), subjects were primed to one of four letters (either V, A, E, or F) and one of these letters was also presented as the target stimulus, Subjects were asked to indicate whether the target letter was a vowel or a consonant. A single-pulse of TMS was applied at stimulus onset on each experimental trial. (b) The mean reaction times (n = 12) for primed vs. unprimed letters in the letter discrimination task (Error bars depict standard error of the means). In the No TMS condition, subjects were significantly faster in detecting primed letters than to unprimed letters, demonstrating that priming was effective. Left PPC TMS reversed the effects of priming: subjects’ RTs to the unprimed letters (i.e., congruent trials) were faster than their RTs to the primed letters (i.e., incongruent trials). This reversal occurred by TMS facilitating the detection of the unprimed letters (rather than by impairing the detection of the primed letters). The critical statistical comparisons are indicated with asterisks. As neurons tuned to the nonprimed targets were less active during application of TMS than neurons encoding the primed targets, this result shows that, as in the TMS-adaptation paradigm, TMS perceptually facilitates the attributes encoded by the less active neural populations