Item

Abstract

An important practical challenge in interleaved TMS/fMRI is the accurate positioning of the coil inside the scanner. We describe a novel method which allows accurate coil placement inside the MR scanner using pre-planned coil positions previously determined with a neuronavigation system. The method was successfully used in a pilot study on the motorcortex which served to demonstrate the viability of our overall TMS/fMRI setup.
Coil positions-of-interest are initially determined using a neuronavigation system
(BrainView, Fraunhofer IPA, Stuttgart, Germany) and saved in respect to the
coordinate system defined by a high-resolution T1 image. Inside the scanner, the
position of the subject‘s head is determined using a fast FLASH image, which is
automatically coregistred to the high resolution image. Based on the coregistered
images, the software automatically determines the parameters for a coil holding
device for targeting the pre-planned position. The accuracy of the method was tested in phantom measurements. Mean deviation between pre-planned coil positions
marked on a spherical phantom and the position given by the setup was 3.895 mm.
This demonstrates a good spatial accuracy in the range previously reported for offline
neuronavigation systems1.
In the pilot study, paired pulse TMS on the motor cortex was investigated in 6 subjects. Whole-brain echo planar images were acquired at 3T. The coil was positioned above the hot spot of the motor cortex (M1) which had previously been determined using the neuronavigation system. Subjects participated in four runs of supra motor threshold stimulation (120% MT), three runs at 80% MT stimulation and one run with volitional movement (acoustically triggered by pulses at 50% MT). A run took 8 minutes in which 27 paired pulses were applied. The group activation maps are in good concordance with previous findings2,3. In particular, for 120% MT stimulation, significant responses were observed in the primary auditory cortex, M1/S1, the supplementary & cingulate motor areas and the thalamus. The observed activation pattern showed robust overlap with the results of the volitional finger movements.