TMS of primary motor cortex with a biphasic pulse activates two independent sets of excitable neurones

BACKGROUND Biphasic pulses produced by most commercially available TMS machines have a cosine waveform, which makes it difficult to study the interaction between the two phases of stimulation. OBJECTIVE We used a controllable pulse TMS (cTMS) device delivering quasi-rectangular pulse outputs to investigate whether monophasic are more effective than biphasic pulses. METHODS Temporally symmetric ("biphasic") or highly asymmetric ("monophasic") charge-balanced biphasic stimuli were used to target the hand area of motor cortex in the anterior-posterior (AP) or posterior-anterior (PA) initial current direction. RESULTS We observed the lowest motor thresholds and shortest motor evoked potential (MEP) latencies with initial PA pulses, and highest thresholds and longest latencies with AP pulses. Increasing pulse symmetry tended to increase threshold with a PA direction whereas it lowered thresholds and shortened latencies with an AP direction. Furthermore, it steepened the MEP input-output curve with both directions. CONCLUSIONS "Biphasic" TMS pulses can be viewed as two monophasic pulses of opposite directions, each stimulating a different set of interneurons with different thresholds (PA < AP). At threshold, the reverse phase of an initially PA pulse increases threshold compared with "monophasic" stimulation. At higher intensities, the reverse phase begins to activate AP-sensitive neurones and increase the effectiveness of stimulation above that of a "monophasic" PA pulse. "Biphasic" stimulation with initially AP pulses is dominated at threshold by activation produced by the lower threshold reverse (PA) phase. SIGNIFICANCE The effects of biphasic stimulation are best understood as the summed output of two independent sets of directionally selective neural populations.