Magnetic resonance imaging (MRI) systems are capable of safely generating detailed images of the human brain with a rich assortment of intrinsic tissue contrast mechanisms, making MRI machines powerful diagnostic and research tools.
Iron accumulation in midbrain structures such as the substantia nigra (SN) and red nucleus is associated with the normal aging process, but also with pathologies such as Parkinson’s disease, a serious and progressive movement disorder. Iron, stored as ferritin, is a highly paramagnetic material and perturbs the uniform magnetic field required for imaging. This results in faster than normal signal decay over time, which can be measured quantitatively with advanced imaging techniques. Unfortunately, current clinical MRI techniques are not used extensively to diagnose Parkinson’s disease as they lack sufficient contrast and resolution to delineate subtle iron accumulations, which are though to serve as a disease biomarker. Scanners operating at very high magnetic field strengths (3 Tesla and above) can generate very high-resolution images and are known to be more sensitive to brain iron stores than lower magnetic field strength magnets (1.5 Tesla and below). As shown in the figure below, high field imaging systems can resolve sub-fields of structures relevant to Parkinson’s disease and provide outstanding contrast in these areas; these systems may provide insight into the cause of Parkinson’s disease and eventually serve as a diagnostic tool for early detection.
(A) The substantia nigra (SN) highlighted on an illustration of the human brain. (B) Enlarged illustration of the SN sub-fields and the red nucleus. (C) MRI of the red nucleus and SN acquired at 3 T with parameters chosen to emphasize iron contrast. (D) MRI at 4.7 T of the same region provides outstanding resolution (structures as small as 350 micrometers can be seen) and displays superb contrast. We hope to measure, and understand, signal variations in images such as this as they relate to Parkinson’s disease.
Is it Secure?
We aim to develop models describing the high field MRI signal from iron rich brain tissues in hopes of relating MR image intensity values to iron concentration and distribution. Monte Carlo simulations tracking nuclear spin states during an MRI experiment are being performed on the OpenMacGrid. The combined resources of the Mac community make it feasible to simulate signal decay rates resulting from the ensemble average of a very large number (at least 10000!) of these nuclear spins diffusing through ferritin containing tissues. The OpenMacGrid serves as the ideal computational environment for this work as these simulations require few system resources, other than processor time, and are optimized for speed on the Mac. Sets of simulations that would take months on a single desktop or small lab cluster can be completed in several days thanks to those who have donated their spare computation power.