Dr. Matt Rosen is a physicist, tool-builder, and inventor whose research bridges the spectrum from fundamental physics to applied bio-imaging work in the field of MRI. He has built a career around an ability to identify big-picture needs and respond with creative solutions that span academic disciplines and push the boundaries of innovation.  His laboratory is known for innovative unconventional approaches to MRI, including world-leading results at ultra-low field (6.5 mT). This work established the field of robust low-cost MRI, and led directly to the development and deployment of the first commercially available low-cost low field point-of-care clinical MRI scanner for neuroimaging in critical care settings by Hyperfine, Inc. Work published in JAMA Neurology in 2020 demonstrated the world’s first point-of-care bedside MRI neuroimaging using a safe low-field (64 mT) scanner in complex clinical care settings for the evaluation of critically ill patients in the neuro ICU and COVID19 wards.
 
Hand-in-hand with MRI at ultra-low magnetic field, the Rosen lab is deeply involved in opportunities provided by hyperpolarization including in vivo Overhauser DNP, SABRE, and spin-exchange optical pumping—all physics-based strategies for transferring spin angular momentum between systems to improve the attainable sensitivity and contrast of NMR and MRI. This work includes such wide-ranging approaches including those using nanodiamonds, aqueous free radicals, and hyperpolarized vitamins.
 
Dr. Rosen leads a significant effort developing new strategies for the acquisition and the reconstruction of imaging data, including network (NN) deep learning (DL) based approaches such as AUTOMAP (Automated Transform by Manifold Approximation). AUTOMAP was the first approach of its kind to learn the optimal sparse transform from sensor data to image data. This approach allows imaging systems to automatically find the best computational strategies to produce clear, accurate images in a wide variety of imaging scenarios. This has led to several new projects involving machine learning to solve inverse problems, such as DRONE for MRF, AUTOSEQ for pulse sequence discovery, and highly multiparametric CEST MRF. Recently, their NN-based SynthSR super-resolution method was combined with automated segmentation to accurately determine brain morphological metrics at 64 mT.

Dr. Rosen is a Fellow of the American Physical Society, a Fellow of the International Society of Magnetic Resonance in Medicine and was named Distinguished Investigator by the Academy for Radiology & Biomedical Imaging Research in 2023.

He is the Kiyomi and Ed Baird MGH Research Scholar, and currently an Associate Professor of Radiology at Harvard Medical School. He Directs the Low-field MRI and Hyperpolarized Media Laboratory and is the Co-Director of the Center for Machine Learning at the MGH/Martinos Center for Biomedical Imaging. He is the Founder of five companies including Hyperfine, which has developed the world’s first portable MRI scanner which can be used at the patient bedside by virtue of its operation at low magnetic field. He has served on the scientific advisory boards of nine companies since 2014.