In this study, we aim to provide in vivo evidence of using a novel electrode array for high-resolution deep brain stimulation (DBS) in rats with simultaneous fMRI readouts. This novel development opens up a new avenue to explore and validate functional connectivity in the brain with a resolution and specificity that cannot be achieved by traditional fMRI or fcMRI approach.Functional connectivity MRI (fcMRI) or resting-state fMRI has become one of the leading tools to map neural circuit connectivity and has been playing a pivotal role in several large-scale brain mapping projects worldwide. Despite ongoing technical advancements in MR data acquisition and analysis, we still face tremendous challenges to unequivocally interpret fcMRI data. This prompt us to develop a novel means that allows mapping brain connectivity by directly stimulating a specific brain nucleus at unprecedented spatial resolution. Our group has demonstrated success in using an advanced micromachining approach to fabricate miniature 16-channel electrode arrays¹ In contrast to many platinum-iridium, glass, and silicon-based electrodes, our microelectrode uses a flexible, highly biocompatible, and MR-compatible base substrate – polyimide, which is known to better match the mechanical impedance of the brain than the other materials commonly used. In this study, we perform fMRI at 9.4T to examine the performance of this electrode array for direct mapping of functional connectivity through localized brain stimulation at the ventral posterior thalamic complex (VP) – a small brain region that has very well-known somatotopic organization. In vivo DBS-fMRI experiments were conducted in a group of five Sprague Dawley rats weighing 250-300 g. A 16-channel MR-compatible microelectrode array was stereotactically implanted for into the VP. Rats were first be orotracheally intubated and ventilated under deep isoflurane anesthesia (2-2.5%) for electrode implantation into the VP. During fMRI experiments, the rats were anesthetized by using 1% isoflurane with pancuronium bromide (1 mg/kg/hr). A 72 mm quadrature volume coil was used as the RF transmitter and a 4-channel phase array coil was used as a receiver. fMRI data was acquired with two-shot gradient echo EPI with bandwidth=100kHz, TR/TE=750/14 ms, NR=4, FOV=2.56x2.56 cm², matrix size=128x128, slice number=8 and slice thickness=1mm. Physiological parameters was monitored and recorded using MR-compatible physiological monitoring systems and maintained within normal ranges, similar to those described in our previous publications^2-5. To validate the efficacy of our microelectrode array for selective current injection, bipolar DBS was performed between two adjacent channel in the serial order of 8 combinations (e.g., 1-2, 2-3, … 15-16) in a pseudo-random manner. For each fMRI trial, the stimulation paradigm was 60 s initial rest, 30 s stimulation, followed by 120 s rest. DBS frequency of 20 Hz square-wave, current of 1 mA, and pulse width of 50 µs was used. Three repeated trials were performed for each combination in order to improve functional contrast-to-noise-ratio and measurement precision. All data preprocessing pipeline and analysis followed our published protocols^2,3.

The 16-channel electrode and experimental setup for high resolution DBS of the VP are shown in Fig 1a-b. In most cases, the lowest contact lead was located in the ventroposterior lateral nucleus (VPL) and the highest contact lead was in the ventroposterior medial nucleus (VPM) (Fig 1c). The locations of the electrode were subsequently mapped by using high-resolution spin-echo images and the specific stimulus locations were assigned into three sub-divisions, namely, dorsal, medial and ventral part of VP (Fig 1c-d). Electrical stimulation of VP showed distinct response patterns and well predicted by the respective stimulation sites (Fig 1e). To further elucidate the evoked time-course responses to DBS at different VP subdivisions, we chose three somatosensory areas including barrel (S1BF), jaw/upper lip (S1J-ULp) and forelimb/hindlimb (S1FL/HL)⁶ as regions of interest (Fig 1f). In order to further dissect the respective projections from each VP subdivisions to the cortex, we also employed data-driven statistical approaches to create maps representing differences among DBS locations (Fig 2). This microelectrode array provides a unique solution for implant accuracy, as any pair of electrode channels (flexible from 75 to 1125 µm spacing at various locations, a total of 120 different combinations for 16 channels) within the entire electrode span can be selected for DBS, enabling fine-tuning of stimulation delivery with a single surgical implantation. This novel development opens up a new avenue to explore and validate functional connectivity in the brain with a resolution and specificity that cannot be achieved by traditional fMRI or fcMRI approach.