Learning-Paradigm Based Acquisition Techniques for Hand-Held Magnetic Resonance Imagers.

MIT PI: Jacob White, Department of Electrical Engineering & Computer Science
SkT PI: Dmitry Dylov, Center for Computational and Data-Intensive Science and Engineering

We propose developing a learning-paradigm based acquisition (LPBA) strategies and to demonstrate them in a hand-held magnetic resonance imager (MRI) with real-time capabilities. The very low signal-to-noise ratio (SNR) intrinsic to this objective will require our collective expertise in imaging, machine learning (from CNN's to adversarial networks), optimization algorithms, fast simulation, MRI and hardware design. To initiate our collaboration, MIT will permanently transfer a set of portable imagers to SkolTech, suitable for both classroom and research use, and use them as a platform for our collaboration on LPBA. Success in developing even a small hand-held MR imager would be an important contribution to world health.

Back to the list >>