Vol. 79, pp. 3523–3526, June 1982 | T. R. BROWN*, B. M. KINCAID*, AND K. UGURLIB†
A method is presented for obtaining the three-dimensional distribution of chemical shifts in a spatially inhomogeneous sample using Fourier transform NMR. The method uses a sequence of pulsed field gradients to measure the Fourier transform of the desired distribution on a rectangular grid (x, y, z) space. Simple Fourier inversion then recovers the original distribution. An estimated signal/noise ratio of 20 in 10 min is obtained for an “image” of the distribution of a 10 mM phosphorylated metabolite in the human head at a field of 20 kG with 2-cm resolution. The method determines the frequency (chemical shift) distribution at each spatial point with an optimum S/N ratio. By suitably pulsing magnetic gradients across a specimen contained within a single pick-up coil, an “image” can be constructed consisting of high-resolution NMR frequency distribution averaged over the resolution volume. This is possible because a pulsed gradient encodes positional information in the initial phases of the free induction decay but does not affect the resonant frequency distribution in space after the gradient has been turned off. Thus, by sampling the free induction decay after a gradient pulse, information about spatial variation can be separated from information about frequencies. The net effect is to measure the Fourier transform of the spatial and frequency distribution function of the spins. This is then inverted to obtain the spatial distribution of frequencies (chemical shifts) over the sample.
The method is demonstrated using a simple one-dimensional phantom consisting of two parallel cylinders, one containing P and the other containing fructose 6-phosphate (Fru-6-P). The demonstration was restricted to one dimension because of computational size limitations; with sufficient memory, three dimensions could be done with the same algorithms. The results show that the method can resolve the two separate components even in the case in which the physical regions overlap. The method is expected to produce an image of a phosphorylated metabolite of 10 mM concentration in a human head in a 20-kG static field in 10 min with a resolution of 2 cm and a S/N ratio of 10-20. The method is of general applicability and can be used to obtain images of the major phosphorylated metabolites in intact animals and humans with spatial resolutions of a few centimeters and S/N ratios of approximately 10. The use of surface coils to image only regions of interest will reduce the number of different k values needed to recover an image and therefore result in faster data acquisition. The method is safe for human observations, with brief exposures to static fields of 20 kG causing no harmful effects. The method is expected to be useful for in vivo biochemistry and 31P imaging.A method is presented for obtaining the three-dimensional distribution of chemical shifts in a spatially inhomogeneous sample using Fourier transform NMR. The method uses a sequence of pulsed field gradients to measure the Fourier transform of the desired distribution on a rectangular grid (x, y, z) space. Simple Fourier inversion then recovers the original distribution. An estimated signal/noise ratio of 20 in 10 min is obtained for an “image” of the distribution of a 10 mM phosphorylated metabolite in the human head at a field of 20 kG with 2-cm resolution. The method determines the frequency (chemical shift) distribution at each spatial point with an optimum S/N ratio. By suitably pulsing magnetic gradients across a specimen contained within a single pick-up coil, an “image” can be constructed consisting of high-resolution NMR frequency distribution averaged over the resolution volume. This is possible because a pulsed gradient encodes positional information in the initial phases of the free induction decay but does not affect the resonant frequency distribution in space after the gradient has been turned off. Thus, by sampling the free induction decay after a gradient pulse, information about spatial variation can be separated from information about frequencies. The net effect is to measure the Fourier transform of the spatial and frequency distribution function of the spins. This is then inverted to obtain the spatial distribution of frequencies (chemical shifts) over the sample.
The method is demonstrated using a simple one-dimensional phantom consisting of two parallel cylinders, one containing P and the other containing fructose 6-phosphate (Fru-6-P). The demonstration was restricted to one dimension because of computational size limitations; with sufficient memory, three dimensions could be done with the same algorithms. The results show that the method can resolve the two separate components even in the case in which the physical regions overlap. The method is expected to produce an image of a phosphorylated metabolite of 10 mM concentration in a human head in a 20-kG static field in 10 min with a resolution of 2 cm and a S/N ratio of 10-20. The method is of general applicability and can be used to obtain images of the major phosphorylated metabolites in intact animals and humans with spatial resolutions of a few centimeters and S/N ratios of approximately 10. The use of surface coils to image only regions of interest will reduce the number of different k values needed to recover an image and therefore result in faster data acquisition. The method is safe for human observations, with brief exposures to static fields of 20 kG causing no harmful effects. The method is expected to be useful for in vivo biochemistry and 31P imaging.