Toward a standard preoperative MRI protocol for functional neurosurgery

Deep Brain Stimulation (DBS) is a well-established approach to treat movement disorders such as Parkinson's Disease, dystonia or essential tremor. For optimal therapy response, accurate electrode placement is critical requiring high signal-to-noise of target areas in preoperative MRI. Currently, imaging protocols vary considerably between DBS centers, making it difficult to compare results or pool data for research purposes. Here, various currently employed MRI sequences from several DBS centers are evaluated regarding their suitability for DBS targeting and a protocol is suggested taking image quality and practical considerations into account. Two healthy subjects (52-year-old female and a 37-year-old male) were each scanned with various sequences (5 T2w, 1 PDw, 4 T2FLAIRw, 2 T2*w, 5 SWI, 2 FGATIR, 1 T1TIR, and 2 QSM techniques) that then were rated by 12 experienced DBS surgeons for their suitability for targeting the subthalamic nucleus (STN), the internal globus pallidus internus (GPi), and the ventrointermediate (VIM) thalamic nucleus. For a subset of sequences, surgeons were asked to identify the optimal DBS target in the STN and GPi. Contrast-to-noise ratios (CNR) were calculated and correlated to intra-rater z-scores and distances of target coordinates. For STN-DBS, surgeons rated T2w, most SWI, QSM, and T2FLAIRw the highest. For GPi-DBS, FGATIR, PDw, and SWI and for VIM-DBS, FGATIR were deemed the most suitable. Higher CNR correlated with higher intra-rater z-scores (R²= 0.29, p < .005) which improved targeting (R²= 0.18, p < .05). Our MRI protocol suggestion is a first step toward standardizing preoperative imaging. All imaging data, MRI sequence parameters, and protocol files are made openly available.