We have measured the T2* signal response associated with cortical activation due to finger motion at 1.5 Tesla. Both thin slice 2D and 3D images show signal intensity changes which vary from 2% to 32% depending on volunteer, echo time, slice thickness, and in-plane resolution. The largest signal change occurred for the thinnest slices and highest resolution (2 mm3). This is consistent with reducing partial volume effects and a simple difference in phase between the intravascular signal and surrounding parenchyma. No inflow enhancement was seen on the 2D or 3D scans, confirming the nature of the signal difference for this approach was due to local field inhomogeneity effects. Using 3D imaging, multiple effects can be seen simultaneously. With a 3D MRA method, it was possible to locate the vessel that was the source of the T2* behavior; it was in each case a vein on the surface of the cortical parenchyma.

Recent emphasis on high resolution gradient echo studies in functional imaging has led to the conclusion that there are likely three domains of response to the blood circulation in the brain when considering field inhomogeneity effects of the venous blood pre- and during activation. The first is a coherent effect due to large or macroscopic vessels on the order of the size of the voxel (ca 200-500 microns in most studies). These can lead to very large signal changes (up to 100%). The second is at the venule level (ca 50-200 microns) and is associated with smaller parenchymal changes (usually ca 10% or less). The third is at the capillary level and is associated with much smaller signal changes at 1.5 T and even up to 4 T. The actual signal changes depend on field strength and sequence design. In this paper, we present our experience in detecting the first two domains with 2D and 3D gradient echo imaging at 1.5 T. We find that high resolution enables visualization of the larger small veins in motor cortex studies and that, on occasion, it is possible to see the venule effects as well. We suggest a simple model to explain the large signal changes based on susceptibility changes and partial volume effects. Comparisons of the functional imaging data to this model and to MR angiographic studies are also shown as evidence of the venous sources of the susceptibility changes.

Measurements of nuclear magnetic relaxation times for protons of water in living skeletal frog muscle show the transverse relaxation time, T(2), increases when a muscle contracts isometrically. This result and other experimental data suggest that a fraction of the intracellular water molecules have restricted rotational freedom and that this fraction decreases when contraction occurs.

To estimate the feasibility of measuring in vivo CSF protein, oxygen, or other solutes through their effect on proton relaxation times, the T1 and T2 of CSF protons has been measured within the human lateral ventricles. T1 was measured at 6.25, 25.4, and 60.1 MHz with a two-point method. T2 was measured at 6.25 and 25.4 MHz using the CPMG sequence to acquire 8 echo images. The T1 was 4.3 s with no evidence of field dependence. The T2 was 2 s. Although these values approach those for water at the same temperature it is possible that the T1 is influenced by the normal oxygen concentration. Calculations based on the relaxivity of dissolved protein indicate that the use of these methods for the detection of elevated levels of CSF protein would be less sensitive than existing methods.

The potential utility of H2(17)O as a contrast agent has been demonstrated in biological solutions and isolated tissues but its use has been impaired by the need to run heavily T2-weighted spin-echo images. By choosing an appropriate steady-state free precession experiment sensitive to T1/T2, we have improved the available contrast-to-noise per unit time by more than a factor of 5. This allows easy measurement of the proton effects for concentrations as low as 0.4% H2(17)O in less than 1 min. Injection into small animals produces a marked reduction in the overall image intensity. Consecutive imaging at the rate of one every 52 s has been used to follow the rate of change in brain image intensity immediately after injection.