Abstract Neuroimaging is a cornerstone of patient care for patients with brain injury. High field magnets and access to imaging interpretation have prevented magnetic resonance imaging (MRI) from becoming a universally available tool. Our group has demonstrated the feasibility of acquiring clinically useful images on a portable, low-field MRI. In this proposal, we will validate the use and successful deployment of a portable, mobile MRI into the acute stroke setting. In current practice, all patients, including those that are critically ill, must be transported to a centralized, controlled-access environment to obtain an MRI at a single time point, in a highly inaccessible paradigm. Our hypothesis is that a highly portable, low-field MRI can be deployed into nearly any setting on a platform that provides real-time, automated neuroimaging analysis. Development of this solution incorporates engineering and technological innovation (low field MRI), methodological innovation (imaging reconstruction techniques, machine learning approaches to automated diagnosis), and conceptual innovation (changing clinical workflow to accommodate a non-invasive method capable of point-of-care and serial MRI). Our key rationale is that we can expand already available treatments, facilitate decision making, and inform new approaches to patient care when we reposition the availability of MRI on a near-universal scale. The fundamental insight is that a low field, portable MRI solution, including advanced methods in image quality, reconstruction, and interpretation, can make imaging available to virtually any patient. We will bring MRI technology and interpretation to an individual patient's bedside and in doing so create a platform for MR imaging and analysis on an unprecedented scale. Because the instrument is inexpensive, does not have cooling requirements and operates on a standard 15A 120V electrical source, project success would democratize diagnostic MR imaging for ischemic and hemorrhagic stroke. In this proposal, we will develop, quantify, and validate the measure of certainty required for transition into clinical care. Stroke has been carefully chosen because of the substantial public health burden and existing treatment options that are available but currently limited because of the requirement for acute neuroimaging. We have developed a highly collaborative and multidisciplinary framework, with leading experts in low field MRI, machine learning, stroke, multicenter studies, clinical and translational research. Embedded as well in our team is the expertise to immediately take this exciting solution to a variety of novel settings (e.g. ambulance, low/middle income countries). Our industry partner, Hyperfine, is the first company in the world to develop a truly portable MRI solution, in collaboration with our academic team over the last three years. As we prospectively develop, troubleshoot, and fine-tune our solution at two leading institutions (Yale and MGH), we have assembled a broad scientific team to incorporate technological, clinical workflow, and health systems factors so that our solution is ready to deploy in any clinical setting to improve patient care across human health.

Project Narrative Magnetic resonance imaging (MRI) was a major advance of the 20th century in patient care, but the need for high magnetic field instrumentation limits widespread accessibility. We have made progress with low-field MRI and compute-based imaging reconstruction and interpretation approaches. We propose the deployment of a point-of-care (POC), portable and highly mobile low-field MRI solution in two independent patient care settings. This system will be used at the bedside in individual patients as a model for broader deployment into any healthcare setting.