This project seeks to deliver unprecedented speed and imaging performance to low-field MRI to demonstrate that low-cost, lightweight implementations of MRI can be developed to allow robust, transportable imaging well suited to diagnose time-sensitive injuries such as head trauma and stroke. Current experimental mobile-sized systems are far too slow and low-resolution for such real-world applications. By innovating upon an already state of the art low-field 6.5 mT MR imaging system through a series of advanced hardware upgrades and pulse sequence optimizations, we expect to achieve 30-fold improvement in imaging speed and thus acquire moderately high-resolution 3D image datasets in about 10 seconds.
Our specific aims are 1) to optimize the design of and install the upgraded gradient hardware, which would provide greater flexibility push our pulse sequences to far higher speeds, and 2) to validate the clinical utility of the optimized system in normal and Traumatic Brain Injury patients and evaluate the low-field system in comparison with gold standard 3T scanning, which will enable us to further optimize our low-field sequences. The technical advances of this low-field project, which are more efficiently developed within an existing larger test-bed system, will be readily translatable into future miniaturized MR systems.
This project is of relevance to public health in that it seeks to develop technology to enable portable MRI scanners that can be operated in mobile environments such as ambulances to rapidly image and diagnose time-sensitive injuries such as head trauma or stroke, where immediate knowledge of the internal state of injury is critical to successful early intervention and improved patient outcomes. This project also pushes forward the movement of low-cost, 'low-field' MR imaging, which may ultimately result in clinically less expensive MRI scanning for many disease types.
| Cohen, Ouri; Zhu, Bo; Rosen, Matthew S (2018) MR fingerprinting Deep RecOnstruction NEtwork (DRONE). Magn Reson Med 80:885-894 |
