The use of MRI to image cerebral perfusion is an exciting development in clinical MRI. Since the NMR signal can be made exquisitely sensitive to molecular motion by applying gradients in the magnetic field, this technique is called """"""""diffusion weighted imaging"""""""" (DWI). However, there are several techniques which do not give consistent results. A group at NIH found that a DWI of water in the brain was related to perfusion at the microvascular level. This result was contradicted by a group at MGH. In each case, it is not clear which technique can reliably distinguish intra- from extra-vascular water. Our experiments utilize fluorinated blood substitutes (FBS), which are strictly intravascular. Since FBS is not now approved for human use, the experiments are done only in animals. Our preliminary experiments have demonstrated that DWI shows a loss of signal of the FBS far in excess of what is known to occur as a result of molecular diffusion; i.e., we have demonstrated a perfusion sensitive MR image. However, our results disagree quantitatively with those of the NIH group and require more study to clarify and characterize the effect. Furthermore, it is desirable to understand the influence of the known vasodilator CO2 as well as the influence of NO synthesis from L-Argine, which alters cerebral perfusion in a selective way. Our MRI technique can distinguish coherent from incoherent flow patterns; a major question is over what time interval does capillary flow mimic incoherent (random) flow. We can also selectively weight the received signal towards the arterial or venus vessels to clarify their relative contributions to the effect. Clarification of these points will improve the scientific basis for diagnosis of many pathologic states, such as infarction, tumor, AVM, etc. when proton DWI is used clinically in humans.
| Joseph, P M (1998) Sampling errors in projection reconstruction MRI. Magn Reson Med 40:460-6 |
