Functional magnetic resonance imaging (fMRI) has rapidly become the dominant tool in human neuroscience research, and is poised to become a transformative technology in the areas of psychiatric and neurological diagnostics. In recent years, large-scale ($150M-$1.2B USD) and long-term (10-12 year) international investments (e.g., NIDA Adolescent Brain Cognitive Development Study, NIH Human Connectome Project, White House BRAIN Initiative, UK Biobank, EU Human Brain Project) have expanded the reach of human fMRI to include faster pulse sequences and more complex analytic tools, higher (?7T) field strength, integration with multi-scale experiments and modeling, and an emphasis on integration of data across multiple scanner/study sites. This generation of fMRI studies goes beyond the original simplistic models that focused upon ?activation maps,? to investigate connections, networks, and dynamic nonlinear circuits in the brain. New ways of thinking are being applied to some of our highest-impact areas of societal interest, ranging from addiction, depression, autism, and brain injury, to age-based cognitive degeneration. However, while fMRI research dramatically accelerates, quality assurance protocols for the MRI machines needed to generate these findings have lagged far behind. Thus, today's neuroimaging centers typically use outdated static phantom protocols that are now incapable of targeting quality control issues relevant to current and emerging applications. The Stony Brook Dynamic Phantom is designed to address this urgent need. Building upon a 1st generation working prototype of the phantom (patent pending), as well as engineering improvements in the 2nd generation prototype designed to increase its durability and reliability, here we focus on the next logical steps to commercialization. Phase I focuses on quality control and establishing added value to the market, by showing that our phantom's measure of dynamic fidelity provides a uniquely informative measure of data quality with direct and concrete implications for the interpretation of human data. Incorporating feedback provided by our Senior Advisory Board, Phase II then proceeds towards commercialization in three steps. First, ALA Scientific Instruments will adapt the engineering for mass production. Second, the academic teams at Stony Brook University and Massachusetts General Hospital/Harvard Medical School will develop algorithms that use characterization of scanner noise to clean data of associated artifact, for both single-subject (clinical) applications as well as normalization across scanners for multi-site research. Finally, hardware and software will be then integrated into one product, which will be field-tested by 15 international leaders in the neuroimaging field. Feedback from this group, identifying potential friction points in terms of the initial learning curve and/or day-to-day usage of the phantom, will be implemented in the final design, to ensure that our final manufacturing for the hardware, software, and documentation make the phantom as pleasant to operate and practically useful as possible. At this point, our device will be ready for commercial distribution.
Functional magnetic resonance imaging (fMRI) has rapidly become the dominant tool in human neuroscience research, and is poised to become a transformative technology in promoting our understanding of addiction and neuropsychopharmacology. However, while fMRI research dramatically accelerates, quality assurance protocols for the MRI machines needed to generate these findings have lagged far behind. The Stony Brook Dynamic Phantom is designed to optimize detection sensitivity of newer task-free ?resting-state? brain networks, by providing the first commercial fMRI calibration device (?phantom?) capable of producing a dynamic (?brain-like?) signal. Quantitative characterization, and subsequent correction, of scanner-specific signal distortion not only will markedly improve the detection of clinically relevant biomarkers at the level of the single patient, but also will permit normalization across scanner platforms for large-scale research studies collected across multiple sites.
