Towards intra-operative guidance in brain tumor surgery using real-time resting-state functional MRI

Executive Summary The long-term goal of this technology development is to improve the clinical utility of resting state fMRI (rsfMRI) for presurgical and intra-operative mapping of resting state networks (RSNs) by developing an analysis tool that can be seamlessly integrated into the presurgical and intra-operative workflow. RsfMRI is a rapidly advancing task-free approach for pre-surgical and intra-operative mapping of RSNs in patients with brain tumors who have difficulties performing tasks. The technology builds on TurboFIRE, a powerful real-time tfMRI and rsfMRI analysis software tool for high-speed fMRI, the success of a prior Phase I STTR study in 12 brain tumor patients that showed clinically acceptable concordance of real-time high-speed multi-echo rsfMRI with tfMRI, and intra-operative electrocortical stimulation (ECS); and our experience with intra-operative rsfMRI in anesthetized brain tumor patients. This technology provides online rsfMRI mapping, eliminating time- consuming postprocessing, and seamless integration into presurgical planning and intra-operative MRI. The objectives of this phase I STTR effort are (a) to develop a fully automated real-time resting state fMRI analysis tool that enables online monitoring of data quality and computation of patient specific maps of resting state connectivity, and (b) assess the performance of this tool for robust presurgical mapping in patients with brain tumors. The goal is to develop real-time functional brain imaging technology with high sensitivity and specificity that is competitive with four state-of-the-art offline toolboxes. The specific aims for Phase II are to (1) develop a real-time rsfMRI analysis tool with online quality control for intraoperative guidance, (2) Validate the technology for presurgical and intra-operative rsfMRI in patients with brain tumors (3) Prepare the documents required by CFR 820.30c for FDA 510(k) clearance. This technology will have significant clinical and commercial potential for a wide range of neurological and psychiatric applications beyond pre-surgical mapping in patients with brain tumors. It opens up functional brain mapping to patient populations that have been difficult to study in the past. Since we submitted the grant application, the FDA-approved offline resting state fMRI analysis tool Quicktome from Omniscent has become available, which, however, does not offer real-time analysis and thus is not suitable for intraoperative mapping. This along with the efforts of several companies, including major MRI manufacturers, who are working on prototypes documents the increasing market needs for resting state fMRI. We have an operational TurboFIRE prototype that was used in two recent feasibility studies in patients with brain tumors. (A) We carried out a feasibility study on non-stationarity of resting-state connectivity in patients with brain tumors in the awake and anesthetized state in preparation for Specific Aim 1b of Phase I and Specific Aim 1a of Phase II, which was submitted to the 2024 ISMRM Annual Meeting (Fig. 1). (B) We carried out a feasibility study on artificial intelligence (AI)-based Figure 2. Peritumoral connectivity using deep learning-based tissue prediction in patient with Figure 1. Group- and scan-averaged test-retest reliability (ICC tumor tissue glioblastoma. (a) Intraoperative view (b) Electro- values) during (a) awake state and (b) anesthesia and (c) difference segmentation cortical stimulation localization (crosshair) during showing increased inter-cortical ICC values during anesthesia. language task inhibition overlaid on T2 MRI (c) Peri- for seed- tumoral connectivity using AI-based seed selection based mapping of peritumoral resting-state connectivity in patients in peri-tumoral edema medial and in proximity to (d) with glioblastomas in preparation for Specific Aim 1a of Phase II, Exner’s language area in tfMRI (verb generation). which was presented at this year’s ISMRM Annual Meeting (Fig. 2). We also started comparing rsfMRI connectivity mapped with two state-of-the-art offline toolboxes in a Human Connectome Project data set. The I-Corps team consists of: Dr. Stefan Posse is a Co-PI of the STTR grant and internationally recognized for the development of real-time fMRI since the late 90s and has over 30 years of experience in biomedical MR research with several patents advancing the field of real-time functional MRI with the TurboFIRE technology. As the Technical Lead/Expert he will be responsible for guiding the technology and patent research of the market discovery strategy. Ms. Jacintha, FACHE is a C-level Corporate Officer (COO/CFO) of Neurinsight LLC who has more than 16 years of experience managing national and international health care organizations, including Sandia National Labs, and 6 years of neuroimaging research experience with PET and fMRI. She will carry out customer discovery and market profitability research and market growth analysis. John Chavez, MBA is a serial entrepreneur and involved in 12 enterprises and startups. He served as President of New Mexico Angels and currently heads the New Mexico Start UP Factory. He will provide guidance with developing the market research strategy and monitor the progress of the market discovery objectives. All three members are committed to the time requirements of the program.