Abstract: Our goal is to develop a cellular level understanding of how transcranial magnetic stimulation (TMS)
may activate neurons in the human cerebellum (Cb) using (i) electrophysiological measurements in an in vitro
turtle Cb, (ii) computational modeling of the induced electric (E) fields in the tissue and (iii) an MEG-EEG-TMS
human saccade study. Since the local microanatomy and electrophysiology of the Cb are evolutionarily
conserved from reptiles to man, it is likely that the fundamental neural dynamics will generalize. In Aim 1a, we
will use our novel high-resolution E-field computation method (BEM-FMM) to develop a thorough
understanding of the induced E-field inside the in vitro tissue (intact, slice). Our simulations will show how the
conductivity boundaries in in vitro preparations can distort, weaken or even reverse the TMS-induced E-field
inside the tissue. The computed E-field values will be verified experimentally with a bipolar electrode. In Aim
1b, we will investigate the optimal bath, tissue and TMS coil geometry for producing the required E-field in the
tissue and design a flux concentrator to provide a uniform strong E-field for slice studies. In Aim 2a, we will
record extracellular and intracellular responses to TMS in the turtle Cb in vitro (intact, slice). Since the neuronal
elements in the Cb (parallel fibers (PF), Purkinje cells (PC) and PC axons) are arranged in mutually orthogonal
directions, we will orient the E-field along each axis so as to selectively stimulate different cell populations and
determine the corresponding E-field threshold. In Aim 2b, we will activate the PC via selective electrical
stimulation of the climbing fibers (CF) (via the inferior olive) and the PF (via the pons, granule cells and the
MF). We will also apply TMS on the afferent bundles in the Cb penduncle and measure the responses intra-
and extracellularly. In Aim 3a, we will use a saccade task developed by one of us to reliably activate two focal
regions in the human Cb and localize and characterize these sources using MEG-EEG. In Aim 3b, we will
compute the E-field in the human Cb using our novel E-field computation method and specify the TMS
parameters for producing an E-field matching that found in Aims 1-2 for PC and PF. In Aim 3c, we will use
TMS-modulation of the saccade by activating the focal Cb regions from Aim 3a. The TMS will be applied at
different latencies, coil orientations and E-field polarities. The effect of TMS on the saccade will be monitored
using EOG, eye tracker and concurrent EEG. The EEG will be compared to TMS-induced responses from
turtle in Aim 2. We expect that this approach will provide a direct link between the BEM simulations (Aim 1), the
turtle studies (Aim 2) and the human Cb study (Aim 3) and result in a greater understanding of the cellular
basis of TMS thus opening new applications of TMS for studying the role of the Cb in human brain function.