P15.07.b functional brain mapping during awake tumor resections using esm-fmri co-registered functional ultrasound (fus)-imaging

Abstract BACKGROUND Neurosurgical resection of brain tumors resembles a balancing act between maximing extent of tumor resection (efficacy) and minimizing the risk of post-operative neurological deficits (safety). Given the difficulty of this trade-off, it is surprising how limited the neurosurgeon’s intra-operative tools are. To this day, neurosurgeons still resect brain tumors without any form of real-time volumetric functional imaging. Functional Ultrasound (fUS) shows great potential as a next-generation intra-operative functional imaging technique, combining submillimeter-subsecond spatiotemporal resolution, high-depth penetration and large fields of view. In previous work we have successfully shown fUS’ ability to map out hemodynamics-based functional brain activity during language and motor tasks during awake tumor resections. However, there has been no study comparing fUS-acquired functional maps and its gold standard counterpart of BOLD-fMRI imaging or Electrocortical Stimulation Mapping (ESM). Here we present our first results of concomitant ESM-, fMRI- and fUS-imaging in the same human subjects. MATERIAL AND METHODS In N = 3 patients undergoing awake surgery for tumor removal, we performed paired functional testing with pre-operative fMRI and intra-operative fUS and ESM. Using conventional linear ultrasound arrays (GE L8-18I-D (7.8 MHz) and GE 9L-D Logiq 9 (5.3 MHz)) interfaced with our experimental system (Vantage-256, Verasonics), we could acquire fUS-images up to 8 cm in depth with a PRF up to 800 Hz. Building on clinically available neuro-navigation software with optical tracking (Brainlab), we co-registered our intra-operative fUS-maps and ESM-hotspots to pre-operative MRI/CT-data in real-time to study spatiotemporal overlap. RESULTS In motor, language and visual tasks, we were able to demonstrate consistent spatial overlap between the fUS-based and ESM- and fMRI-based functional regions. Our fUS functional maps presented with unprecedented mesoscopic-scale precision (400 µm), even at depths of over > 5 cm. In contrast to fMRI and ESM, fUS was also able to concomitantly visualize the in-vivo microvascular morphology underlying the functional hemodynamics. The current results were only obtainable after we developed a powerful study pipeline including pre-operative fMRI-imaging and ROI-planning, intra-operative integration of experimental Doppler-data and co-registered functional analyses. CONCLUSION The work presented here is the first-ever, in-human comparison between ESM-, fMRI- and fUS-based functional data, serving as a significant milestone towards further clinical maturity of fUS as a new intra-operative tool to improve the safety and efficacy of neurosurgical brain tumor resections. Disclosure: This work was supported by the NWO-Groot grant of The Dutch Organization for Scientific Research (NWO) (Grant no. 108845).