Diffusion imaging has gained importance in the past decade as a valuable means of depicting white matter injury caused by various disease processes. Diffusion imaging holds particular promise for evaluation of individuals who have experienced traumatic brain injury (TBI) because damage to white matter pathways is considered to be an important component in the causation of the many types of neurocognitive impairment that can result from TBI. Diffusion imaging can be performed using a number of imaging techniques, and no single technique is universally recognized as the single best method. As a result, development of large pools of data is hampered by the fact that combining imaging studies obtained by multiple techniques results in an inhomogeneous data set that is difficult to analyze.
If diffusion imaging is to be developed as a means to evaluate Veterans with suspected TBI, a uniform type of image acquisition is needed across the different types of imaging systems available within the VA hospital network. To construct such a system, a means is needed to establish exactly how one scanner differs from another (or from itself over the course of time). Then, modification of imaging sequences and, as needed, hardware and software components, can be performed to allow more uniform data acquisition across scanners. This study uses diffusion imaging phantoms to evaluate differences between scanners with the goal of providing acquisition techniques that will allow data to be compared across different patient groups and combined into large data collections. The objective is to provide a means for the many scanners across the VA hospital system to provide the same imaging answers in a suspected TBI patient.
Inter-scanner differences in diffusion values will be evaluated using two types of novel diffusion phantoms. The first type of phantom will be provided by Michael Boss, PhD at the National Institute of Standards and Technology. These phantoms are composed of various forms of aqueous polymer solutions, which each provide a ground truth value of random microscopic water motion (termed the apparent diffusion coefficient, or ADC) value. The second type of phantom, which will be provided by Walter Schneider, PhD at the University of Pittsburgh, will provide ground truth for other important diffusion imaging parameters, including fractional anisotropy and radial diffusivity. This second type of phantom is composed of micron-scale hollow fiber textiles in various arrays of controlled packing and crossing patterns, which are designed to provide ground truth for axonal patterns similar to those found in the human brain. In addition, the study team will perform diffusion imaging on one to two human volunteers at each of four sites, in conjunction with phantom imaging. Both phantoms and the individual will be scanned multiple times at each site during the study period. Intra- and inter-scanner differences will be measured and, based on these findings, an imaging protocol that will provide optimal uniformity of diffusion results across the sites will be designed.