To produce image information from an object it is subjected to a continuous static magnetic field along a Z axis and to sets of sequences of orthogonal gradients G.sub.x, G.sub.y and G.sub.z to the magnetic field. Spins in a selected plane (the X-Z plane) are excited by selective rf pulses and an associated G.sub.y gradient and the selected spins are subjected to all three gradients of which the G.sub.z gradient provides twist or warp to each column of spins extending along the Z axis to phase-encode the columns. The spin-echo signals are read out in the presence of a G.sub.x gradient. In each set of sequences a different value of Z gradient is employed. The Fourier transformed spin-echo signals obtained from each sequence, when arranged in order of increasing G.sub.z gradient and subjected to a second Fourier transform represent the distribution of spin density in the Z direction, thus giving a two-dimensional image of the selected X-Z plane.
This is a continuation of application Ser. No. 702,126, filed Feb. 15, 1985, now U.S. Pat. No. 4,602,214 issued July 22, 1986, which was a continuation of Ser. No. 320,953, filed Nov. 9, 1981, now U.S. Pat. No. 4,506,222.
The invention relates to a nuclear magnetic resonance tomography method, by which in each sequence after a reading stage in which the nuclear magnetic resonance signals produced at the examination area are recorded, the magnetic gradient fields are switched on and off in such a manner that the time integral across all magnetic gradient fields is equal and different from zero for each sequence.
In a method of distortion correction for gradient non-linearities in a nuclear magnetic resonance tomography apparatus, two auxiliary datasets h(x,), f(x,y) are measured that describe the shift of a measured location (x',y') relative to an actual location (x,y) of a signal origin. A corrected auxiliary dataset h(x+f(x,y), y) is produced from the auxiliary dataset h(x,y). A location correction in y-direction ensues in an image dataset upon employment of the corrected auxiliary dataset h(x+f(x,y), y). A first intensity correction also ensues. A location correction in x-direction subsequently ensues with a second intensity correction. Alternatively, the location correction can also ensue by Fresnel transformation of the raw dataset upon involvement of the corrected auxiliary data h(x+f(x,y), y).
A magnetic resonance signal of a volume of interest within an object is acquired by placing the object in a magnetic field and the application of three successive steps. The first step comprises an rf selective 90.degree. pulse of frequency f.sub.x applied in the presence of an x-gradient magnetic field to select a slice. The magnetization of the slice is then inverted. In the second step an rf selective 90.degree. pulse of frequency f.sub.y is applied in the presence of a y-gradient magnetic field to excite a strip of the slice, in the third step an rf selective 180.degree. pulse of frequency f.sub.z is applied in the presence of a y-gradient magnetic field to refocus the magnetization of a region of said strip. The resulting free induction signal is derived solely from said region.
For MR image acquisition, a sequence of individual measurements that are each composed of an excitation phase and a read-out phase is implemented. In the read-out phase, a nuclear magnetic resonance signal is read out that is allocated to only one point in the k-space defined by a preceding phase-coding gradient. A phase-coding gradient remains activated during a group of individual measurements and changes in size from individual measurement-to-individual measurement. Switching this phase-coding gradient in every individual measurement is thus avoided, so that ramp times can be eliminated and the overall data acquisition procedure is considerably shortened.
Methods for measuring the perfusion of fluid in a substance are shown to include subjecting the fluid to electromagnetic energy so as to cause a response related to the magnetization of the fluid before it enters the substance, performing magnetic resonance measurements on the substance to generate intensity information and processing the intensity information to determine perfusion. In one embodiment of the invention, perfusion is measured by labeling atoms in the fluid at a base point, generating a steady state in the substance by continuing to label atoms until the effect caused by labeled atoms perfusing in the substance, reaches a steady state, generating image information for the substance and processing the image information to determine perfusion. It is preferred to label atoms by applying magnetic resonance perturbation. In one embodiment the labeling of atoms involves saturating spins associated with the atoms. In an especially preferred embodiment labeling involves inverting spins associated with the atoms continuously by adiabatic fast passage. Such inversion is preferably achieved by applying a radio frequency field virtually continuously. The invention is particularly useful where the substance is tissue and wherein the fluid is blood. In such an embodiment, labeling involves labeling the hydrogen atoms of water contained in the blood. It is also preferred for labeling to occur at a point between the heart and the tissue. It is also especially preferred for the generation of magnetic resonance images to involve generating a first image while labeling at the base point, labeling at a remote point, generating a second image while labeling at the remote point and generating a relaxation image. In such an embodiment, all of the images are processed in the determination of perfusion.
