“Non-invasive” brain stimulation is not non-invasive

INTRODUCTION The functions of the healthy brain can be studied in two main ways. Firstly, the changes in the brain’s state can be measured using techniques such as EEG or functional MRI. Secondly, the activity of the brain can be disrupted through the use of brain stimulation. The famous experiments of Wilder Penfield and colleagues in the 1950s showed the power of brain stimulation in people whose brain was exposed in surgery, and highlighted the possibility of inducing changes in the brain’s state to demonstrate the involvement of specific brain areas in particular functions (Jasper and Penfield, 1954). Two main techniques are available for human brain stimulation: transcranial magnetic stimulation (TMS) and transcranial current stimulation (tCS). More recently, it has been suggested that TMS and tCSmight be used to enhance brain function, as well as to disrupt activity. These techniques have collectively become known as “non-invasive brain stimulation.” We argue that this term is inappropriate and perhaps oxymoronic, as it obscures both the possibility of sideeffects from the stimulation, and the longer-term effects (both adverse and desirable) that may result from brain stimulation. We also argue that the established tendency for the effects of TMS and tCS to spread from the target brain area to neighboring areas is in itself contrary to the definition of non-invasiveness. We argue that the traditional definition of an invasive procedure, one which requires an incision or insertion in the body, should be re-examined, and we propose that it be widened to include targeted transcutaneous interventions. TYPES OF BRAIN STIMULATION An electric current travelling through a coiled wire creates a magnetic field. This property is used in TMS to create brief magnetic pulses which easily traverse the skull and other matter overlaying the brain. The pulses generate electrical potentials in the brain, depolarizing neurons and thereby triggering action potentials (Di Lazzaro et al., 2004). A single pulse of TMS will have two effects: firstly the generation of action potentials in the targeted brain areas underlying the coil; secondly a refractory “silent period” in those same cells as the ion balance is restored. While the effects of a single TMSmay only last on the order of a few milliseconds, multiple pulses may induce long term potentiation or depression in the target cells. For example, trains of pulses delivered at 1Hz result in reduced excitability in the target area for a prolonged period of time, on the order of tens of minutes after the end of the stimulation. A recent development has been the use of rapid bursts of pulses such as thetaburst stimulation (TBS), which can have opposing effects on excitability depending on the temporal pattern of the bursts (Huang et al., 2005). TMS may be used either “online,” to affect the brain during a task, or “offline,” to compare task performance after vs. before longer periods of stimulation. tCS is a term that covers several techniques, principally involving direct or alternating current (tDCS or tACS). In a typical tDCS experiment, the participant performs a task to establish a baseline performance level. Then a pair of electrodes is placed on the head, one (or both) of which overlie a target brain area. The experimenter delivers a small electrical current for around 10–20min. Following this the participant performs the task a second time to establish whether the stimulation has had an effect on behavior. The effect depends on a number of factors, including current amplitude and duration (higher currents delivered for more time usually induce a greater effect), the polarity of the electrode over the target area (typically the negative electrode, or cathode, will worsen performance while the positive, anodal, electrode will enhance it), and the brain area and task under study (Nitsche and Paulus, 2000, 2001). tACS is less well studied, however the technique offers the possibility of exploring the casual involvement not only of a target brain area, but also of a particular frequency band. For example the beta range (15–35Hz) is known to be associated with human motor control, however it has only recently been possible to show the causal involvement of beta frequencies in maintaining motor state (Pogosyan et al., 2009; Fuerra et al., 2011). These latter studies used tACS to increase the power of the beta band while the motor system was under study, giving somewhat contradictory results (Davis et al., 2012). The timescale over which the effects of brain stimulation are seen can vary from milliseconds to weeks. At the briefest level, a single pulse of TMS lasts for 100–200μs, during which time an electric field is induced in the target area. This is enough to generate action potentials in these target cells, and to induce a refractory silent period following the initial burst. Conversely the instantaneous effects of tCS are under-explored, and much of our knowledge of the effects of the electric field on the brain comes from modeling