| Objectives: 1 to discuss methodology of CSI of H-MRS and representation of CSI on image of normal brain; 2 to study representation of CSI of 'H-MRS of brain radiation injury caused by radiotherapy of NPC (nasopharyngeal carcinoma) and determine its clinical value; 3 to investigate relativity between pathological changes of canine brain caused by radiation and MRS.Materials and methods: A software package supplied by Siemens Co. was used to analyze spectrum of NAA, Cr and Cho. The values of NAA, Cr and Cho on the horizontal coordinate were 2.0pprn, 3.0ppm and 3.2ppm respectively. Sums of areas under the curves stood for peak values. With the base image of transect, FOV was readjusted with matrix of 64X64 and matrix of 16X32 by means of K space zero fill up. FOV readjusted with matrix of 64 X 64 was used for metabolic graphs of NAA, Cr and Cho. FOV readjusted with matrix of 16 X 32 was used for graphs of their peak values. Data of both temporal lobes and midbrain were collected for analysis. Position of each pixel was checked on corresponding three-dimension spectral positioning image. No pixel with air, bone or much cerebrospinal fluid on a positioning image was used to collect data. A spectral curve was map out for each eligible pixel. Pixels of both temporal lobes in the control group were got from images for analysis. According to images, patients in the case group were divided into three groups: image-negative group, RE group and metastases group. Pixels of midbrain in the case group were got from the images for analysis. Pixels of both temporal lobes in the image-negative group were got from the images for analysis. Pixels in foci, at margins and periphery of foci in the RE group were got from the images for analysis. When a focus appeared at one temporal lobe, eligible pixels of the other lobewere used. Pixels in tumor, at margins of tumor, in edema region, at margins and periphery of edema and eligible pixels of the other lobe in the metastases group were got from the images for analysis. The right temporal parts of three healthy mongrels were irradiated once by a linear accelerator at a dose of 80Gy. SPSS software package was used for statistical analysis. When a sample was composed of two groups, its data were T-tested. When a sample was composed of several groups, its data were analyzed by variance of single factor. P less than 0.05 stood for significant statistical difference. Averages of data with significant statistical difference were compared.Results: 1 From images of both temporal lobes of the control group, 198 pixels were got. Peak values of NAA, Cr and Cho were 112.7 ?9.3, 58.5?4.1and 68.2+28.8 respectively. Ratios of Cho/NAA, Cr/NAA and (Cho+Cr)/NAA were 0.63?.29, 0.53+0.22 and 1.16?.43 respectively. From images of the midbrains, 47 pixels were got. Peak values of NAA, Cr and Cho were 160.2?36.1,60.1 ?4.5 and 63.1 ?4.0 respectively. Peak values and ratios of both temporal lobes showed no significant statistical difference among all subjects. Peak value of NAA of both temporal lobes and the midbrains showed significant statistical difference. That of the midbrains was bigger. Other peak values and ratios showed no significant statistical difference.2 From images of both temporal lobes of the image-negative group, 483 pixels were got. Peak values of NAA, Cr and Cho were 96.7 ?6.3, 75.7 + 26.3 and 82.9 ?29.9 respectively. Ratios of Cho/NAA, Cr/NAA and (Cho+Cr)/NAA were 0.95 ?.55, 0.88 ?.53 and 1.83 ?.99 respectively. From images of the midbrains, 102 pixels were got. Peak values of NAA, Cr and Cho were 129.2?8.7, 86.1 + 29.5 and 58.3?6.0 respectively. Ratios of Cho/NAA, Cr/NAA and (Cho+Cr)/NAA were 0.56 ?.20, 0.50?.17 and 0.94+0.38 respectively. Peak values and ratios of the temporal lobes of the control and the image-negative group showed significant statistical difference (p<0.01). Peak values of Cho and Cr of the image-negative group werebigger and that of NAA was smaller. Ratios of Cho/NAA, Cr/NAA and (Cho+Cr)/NAA of the image-negative group were...
|
PDF Full Text Request