Magnetic Resonance Imaging Technology Development

National Institute Of Neurological Disorders And Stroke

Investigators

Linked publications & trials

Abstract

The AMRI section at NINDS carries the main responsibility for the development of NIHs 11.7T human MRI scanner, a uniquely powerful system that will allow in-vivo neuroimaging with unprecedented clarity. Over the last year, AMRI has continued preparation for re-installation of the 11.7T human MRI system, which got delayed because of the COVID-19 crisis. Together with the clinical trials unit at NINDS, it has nearly completed an application for an investigational device exemption, required by the FDA for exposing humans to magnet fields above 8T. In parallel, we have collaborated with scientists of the Engineering core in LFMI/NINDS to develop RF transmission and detection hardware for operation at 11.7T. A major goal of the high field MRI development is to classify brain structures and lesions based on a combination of MRI contrasts, and to quantify certain tissue constituents. In this regard, the main targets are iron and myelin, both of which are critical to healthy brain function, and both of which provide strong MRI contrast. To allow their quantification, we are identifying to what extend they contribute to MRI contrast. This requires combination of a variety of techniques, including high resolution in-vivo MRI on humans and animals, MRI of brain samples, and histology, and the development of biophysical models of MRI contrast generation. With the goal of improving quantitation and reproducibility, we further investigated how iron and myelin affect MRI in order to derive accurate models to be used in data analysis and possibly improve measurement methods. Using pulsed magnetization transfer (MT) and inversion recovery-type T1 techniques, we previously determined that the magnetization in water trapped between the layers of myelin has a much shorter lifetime than commonly assumed, averaging about 13 ms. As a result, a 2-pool model of exchange suffices to explain most T1 and MT contrast, and robust estimates of brain semisolid fraction (and possibly myelin) can be obtained with a simplified measurement approach. Using this model, we published our finding that the T1 relaxation time of myelin hydrogen protons (MP) is strongly dependent on field strength (Magnetic Resonance in Medicine, 2020). Using NIH MRI scanners operating at field strengths (B0) of 0.5, 1.5, 3, and 7T, we were able to characterize this dependence as linear according to T1MP=0.08*B01.0. In collaboration with the Bandettini group at NIMH we used this knowledge to improve GM-WM contrast for cortical segmentation (published in Neuroimage, 2021) During an MRI brain scan, voluntary head motion is a common occurrence and affects image quality, in particular at high field. Previously AMRI developed a unique approach to correct for the effect of head motion in order to recover MRI quality. The approach simultaneously corrects for spatial encoding errors associated with head motion as well as accompanying magnetic field changes. In collaboration with the group of Daniel Reich, we evaluated the effectiveness of this approach for lesion visualization in MS patients and published this work in Investigative Radiology in 2021. AMRI is currently working on modifications of this technology to allow higher spatial resolutions and full brain coverage.