Transcranial magnetic stimulation (TMS) is an increasingly popular neuroscience tool due to its unique ability to noninvasively alter neural activity in targeted regions of the brain 
. Since its introduction in 1985 by Barker and colleagues 
, TMS has been used to probe motor cortex excitability 
, map motor and cognitive functions 
, study anatomical and functional connectivity 
, and modulate brain function with therapeutic aims 
TMS uses a time-varying magnetic field to induce an electrical current through the skull, in a spatially restricted region of the cerebral cortex. The induction of electrical current occurs with minimal attenuation of the magnetic field. Significant currents can be induced without having to apply substantial voltages across the skull, minimizing the activation of pain fibers and pain sensation. The advantage of TMS is also in its temporal (sub-millisecond) and spatial (sub-centimeter) resolution.
Two configurations of TMS coils are commonly used in scientific and clinical research. The figure-of-eight coil (also known as butterfly or double coils) is the most commonly used configuration owing to its superior spatial specificity. The circular coil is less used because while it offers more powerful stimulation and the opportunity to target both motor cortices at the same time with relatively little worry about specific placement or constant positioning, it is also less focused. It has been used in clinical trials that targets large regions of the brain, such as investigations of Parkinson's disease and epilepsy 
and motor physiology studies 
. Its specificity can also be improved when applied to brain regions where the preferred current direction is known, such as the motor and visual cortices 
As with any experimental technique, TMS has its pitfalls 
. Specifically, TMS is accompanied by a number of ancillary effects. The coil emits clicking sounds with each stimulation, and can also stimulate nearby peripheral nerves and muscles. Depending on the location and strength of TMS, this may result in sensations ranging from a light tapping on the scalp to uncomfortable muscle twitches in the face, neck, or shoulders. These sensations can nonspecifically interfere with task performance via distraction or subject biasing, contaminating the results. In clinical research, placebo effects are known to be high 
, especially with medical devices where there is significant patient-investigator contact 
To separate the effects of brain stimulation from those arising from the above artifacts, experimenters can compare results with control conditions in which they either apply sham stimulation or apply real stimulation to a control brain region. These two methods are complementary to one another; one may not be necessary in some studies, and in other studies, stimulation of control brain regions methods may still be necessary in addition to sham TMS (to show specificity of the brain region of interest). Ideally, the experimental and control conditions should differ only by the way in which brain is stimulated, while producing auditory and tactile artifacts that are not easily distinguishable from real stimulation. See Supporting Information Text S1
for detailed discussion about different types of control (including sham) conditions that are available. Furthermore, the conditions should be easily interleaved to allow within-subject comparisons and intermix various conditions trial-by-trial.
The goal of this study was to develop and fully validate a method of delivering several control TMS conditions. Two coils were fabricated; a figure-of-eight coil (Fig8) that has loops of coils in each of the two wings that are driven separately, and a circular coil (Circ) that has two sets of coils stacked on top of another that are also driven separately. An attachment allows the delivery of three types of stimuli in an automated, interleaved manner without switching or moving the coil (single-trial sham TMS). 1) Standard stimuli are delivered when current direction in both loops matches that of the standard coils. 2) Sham stimuli are delivered when current direction in one of the two loops is backwards. 3) Reversed stimuli are delivered when current direction in both loops is backwards. Reversed stimuli reproduce the fields created by coil-flipping, which can be used to increase activation thresholds over brain areas where the preferred stimulus orientation is known, such as motor 
, visual 
and prefrontal cortices 
. In the case of motor and visual areas, these can also be used to preferentially stimulate either hemisphere from a single coil location.
We extend upon Ruohonen et al.'s design of a sham Fig8 coil 
. We add independent control of current direction in both
coil loops so that reverse stimulation is possible in addition to sham and standard stimulation. Further, automated electronic switching of stimulus types can be done within 3 ms with a solid state switch known as thyristors, and we apply the design to both Fig8 and Circ coils. In addition, one can adjust stimulation intensity of each current to achieve complete cancellation of the induced fields (with circular coils, since there is some distance between the two loops of coils, the stimulus intensity necessary to achieve complete cancellation for each loop is different). To enhance the applicability of the design, we implement it in an attachment to Magstim single- and dual-pulse setups, which are in common use in research and clinical settings.
Four types of experiments were performed to validate the Standard, Reversed and Sham TMS delivered from the Fig8 and Circ coils. First, in order to characterize physical properties of the stimuli such as electro-motive force (EMF), we performed mathematical modeling and actual measurements using a search-coil. This included both measurements of single pulses and of successive pulses to ensure that stimulus properties were not contaminated by prior stimuli via residual states in the circuitry. Second, we measured the physiological effects of the stimulus types by comparing thresholds for eliciting motor evoked-potentials (MEPs) when stimulating primary motor cortex. Third, we tested the perceptual effects of the different pulses by testing whether subjects and experienced investigators could differentiate Sham stimuli, and if so whether Standard and Reversed could be differentiated (which may serve as another form of sham TMS). Finally, sound pressure level (SPL), subjective loudness and pain intensity were measured to further characterize their effects on the subjects.