Efficient MRI through Improved Encoding & Reconstruction

Abstract:  MRI has demonstrated ability to provide exquisite contrast for non-invasive imaging. What limits its efficiency and sensitivity are the tradeoffs between scan time, resolution and signal-to-noise ratio (SNR). Our research is devoted to breaking this stalemate by developing new acquisition and reconstruction methods that simultaneously exploit MR physics, signal processing, deep learning and hardware capabilities. We develop these technologies and pursue their applications to enable,

I. high-resolution neuroimaging at the mesoscale to examine structural and functional architecture of the human cortex in vivo,

II. efficient quantitative imaging to map biophysical tissue parameters and probe tissue composition,

III. open source imaging for harmonization of data acquisition, reconstruction and analysis, and easier dissemination,

IV. rapid and motion-robust abdominal and fetal imaging, and

V. synergistic combination of scan-specific deep learning with physics-based image reconstruction.

The overarching goal in creating these techniques is to push the limits of spatial/temporal resolution and SNR of MRI to make it a better neuroscientific and clinical tool, and improve efficiency of MRI exams to make them more cost effective and more widely used in the clinic.

 

Biography: Berkin is an Assistant Professor in Radiology at the Martinos Center for Biomedical Imaging, Massachusetts General Hospital, and Harvard Medical School, where his group develops data acquisition and reconstruction techniques for MRI.