BOLD-fMRI response vs. transcranial magnetic stimulation (TMS) pulse-train length: testing for linearity.

Abstract

To measure motor and auditory cortex blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI) response to impulse-like transcranial magnetic stimulation (TMS) pulses as a function of train length. Interleaved with fMRI at 1.5 T, TMS pulses 0.3-msec long were applied at 1 Hz to the motor cortex area for thumb. Six subjects were studied in a TR = 1 second session administering trains of 1, 2, 4, 8, and 16 pulses, and a TR = 3 seconds session administering trains of 1, 2, 4, 8, 16, and 24 pulses. A simple hemodynamic model with finite recovery and saturation was used to quantitatively characterize the BOLD-fMRI response as a function of train length. In both the activations directly induced in motor cortex by TMS and the indirect activations in auditory cortex caused by the sound of the TMS coil firing, the BOLD-fMRI responses to multiple pulses were well described by a summation of single-pulse impulse functions. Up to 24 discrete pulses, BOLD-fMRI response to 1 Hz TMS in both motor cortex and auditory cortex were consistent with a linear increase in amplitude and length with train length, possibly suggesting that stimuli of 1 to 2 seconds may be too long to represent impulses.