PFI-TT: Enabling More Scans per Machine through in Magnetic Resonance Imaging Data Processing

Award ID: 2044599

The broader impact/commercial potential of this Partnerships for Innovation - Technology Translation (PFI-TT) project is to reduce Magnetic Resonance Imaging (MRI) scan-times and improve patient throughput. This techology will be beneficial for patients as well as healthcare providers. Shorter scan-times will improve patient comfort, especially for patients that are young, elderly, and/or claustrophobic. Improved throughput will increase accessibility, allow patients to receive MRI scans in a timely manner with less wait-time, and contribute to better healthcare outcomes. For healthcare providers and imaging facilities, shorter scans and better throughput will increase operational efficiency, revenue potential, and patient satisfaction. The technology can be leveraged to help accommodate rising healthcare demand from an aging population. The graduate student selected for technology development will obtain research experience and educational training to pursue entrepreneurship following the completion of the project. A team of 10 undergraduate students from underrepresented groups will also participate in the applied research and software development. The proposed project will provide a novel signal processing approach that inputs the poor quality (noisy) images obtained at short scan-times and outputs a noise-free image, similar to what would have been obtained after a long data acquisition time. In MRI, image quality is usually inversely proportional to scan-time. Longer scan-time yields better image quality and vice versa. The algorithm isolates noise from raw MRI data by distinguishing between their distinct characteristics: noise is random while raw MRI data contains patterns/features. There are two novel features of the approach: 1) ability to identify and separate noise; and 2) the application in denoising raw MRI data. Compared to conventional signal processing methods such as filtering methods, the team's proprietary wavelet-shrinkage-based denoising method can process low signal-to-noise ratio signals without the limitations of inadequate noise removal or signal distortion. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Program Director: Jesus Soriano Molla