How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!
Q & A With Len Yencharis
An interview with John Van Horn, Ph.D., M.Eng., a Research Associate Professor at the Center for Cognitive Neuroscience, Dartmouth College...
LEN YENCHARIS, A.I.: In the emerging field of cognitive neuroscience, the scientific disciplines required seem to span from not only the pathology of the nervous system but also to pharmacology, clinical assessments and chemistry. What are the benefits from using fMRI over conventional MRI and other medical scanning modalities for understanding the processes in neural development?
DR. JOHN VAN HORN: The advent of functional magnetic resonance imaging (fMRI) of the human brain and the rise of cognitive neuroscience has occurred jointly since the early 1990s. This co-evolution has greatly aided the quest to understand the fundamentals of how the brain processes information, conducts complex reasoning and carries out motor tasks. The advantage of fMRI over other neuroimaging methods, such as positron emission tomography (PET), is in allowing non-invasive examination of brain blood oxygenation-level dependent (BOLD) changes (unlike PET, where subjects are injected with radioisotopes such as oxygen-15 water, H215O). Though fMRI studies have been performed looking at the functional differences across various age groups, we expect that it will be the coupling o fMRI with another magnetic resonance-based measure called Diffusion Tensor Imaging (DTI), where the orientation of white matter fiber tracks may be obtained, that the greatest advancements will occur in understanding how cognitive function changes with brain development. Moreover, these image data can be used to help constrain computational models of functional connectivity between brain regions. Measuring how these connectivity models change as a function of aging will go a long way toward helping to understand the processes involved in cognitive development.
YENCHARIS: Storage requirements is a major concern for these large image files. To participate in your collaborative efforts with other neuroscience centers, how do you plan for what might be an increased use of neuroanatomists or maybe even bioimaging applications beyond your immediate goals? If fMRIDC doubles every nine months, then you have exceeded Moore's Law as we know it. Do you foresee even stronger demand for more storage?
DR. VAN HORN: We are not concerned that Moore's Law limits the growth of the fMRI Data Center (fMRIDC) study data archive, per se. However, we do note that neuroimaging seems particularly affected by Parkinson's Law, which states that "Data expands to fill the space available for storage." Given the opportunity, fMRI researchers will tend to collect as much data as they possibly can from a subject. For instance, brain mapping researchers might purchase more computer disk storage, which permits them to use more storage-intensive image acquisition and analysis techniques, which then leads to the need for even greater storage. As MRI physicists and scanner manufacturers are working continually to improve spatial and temporal resolution of the scanners themselves, brain researchers are adept at looking for ways to capitalize on these technological advantages to obtain more data. This means that the average size of an fMRI study will tend to increase over time. We have several studies at the moment whose size rivals that of the Human Genome Database (~45 GB uncompressed). Though they seem large by today's standards, in time these studies will be considered small. This means that the fMRIDC has to be prepared to receive an ever-increasing amount of information from researchers. To meet this demand, we are putting into place high-end storage solutions from Sun Microsystems. This includes 6 TB of RAID5 disk storage, as well as a Sun StorEdge L700 nearline tape storage system running SAM-FS. Utilizing this leading-edge storage hardware and accompanying Sun software, we are confident that we can grow the archive for many years to come.