3.0T MR Spectroscopy Imaging In The Evaluation Of Brain Tumors

3.0T MR Spectroscopy imaging in the evaluation of brain tumors Part Ⅰ: 3.0T MR Spectroscopy imaging techniquePurpose: To search for the MRS sequences fit for the examination of cerebral tumor by comparing the different imaging parameters systematically. Materials and methods: Using a GE signa VH/i 3.0T super conduct MR unit, 36 cases of patients whose possible diagnoses are cerebral tumors were performed with 2-D multi-voxel PRESS 144ms and STEAM35ms, and single-voxel PRESS 144ms at the same time.6 cases were performed with 2-D multi-voxel PRESS30ms. Different size of ROI were compared to find out the change of the relative quantification. Functool software package was used for post-processing of spectrum. Results: 1 The baseline and the FWHM of multi-voxel PRESS 144ms are better than that of multi-voxel STEAM35ms, but the latter one show higher SNR than the prior one. 2. No significant difference can be revealed between PRESS30ms and STEAM35ms. 3. No significant difference of relative quantification can be revealed with different size of ROI. Conclusion: 1. Multi-voxel PRESS 144ms is the MR spectroscopy sequence fit for the examination of cerebral tumor. 2. Appropriate location and size of ROI is benefit to the MRS examination. PartⅡ: Relative quantitative 3.0T MR Spectroscopy imaging in normative humanPurpose: Using MRSI to evaluate the brain metabolite of volunteers, and obtain the normal comparison for cerebral tumor. Materials and methods: Using a GE signa VH/i 3.0T super conduct MR unit, 44 cases of volunteers with the same period of age as the cerebral tumor were performed with 2-D multi-voxel PRESS 144 to explore the metabolite of the brain. Then compare the parenchyma of hemisphere and the gray/white matter by the relative quantification and pseudocolor images of NAA, Cho, Cr, Glx, mI. Functool software package was used for post-processing of spectrum. Results: 1.Significant differences can be revealed in the Cho/Cr, NAA/Cr between white matter of occipital lobe in male adult, and Cho/NAA, NAA/Cr between the semioval center in female adult, and Cho/NAA between white matter of the frontal lobe in female adult. The difference between gray and white matter in the same lobe in male is the same as in female adult. 2. Pseudocolor image can identifythe spatial distribution of the brain metabolite directly. Conclusion: 1. By spectrum, relative quantification can be used to evaluate the difference of brain metabolite. Consistency can be revealed between the anatomy and the MRSI in the asymmetry of the brain. Partni: 3.0T MR Spectroscopy imaging in the evaluation of cerebral astrocytic tumorsPurpose: To reveal the biological feature of astrocytic tumors in MR spectrum and explore the possibility of grading the malignancy of astrocytic tumors, then find out a threshold for the differential diagnosis between astrocytic tumors and other leisons. Materials and methods: Using a GE signa VH/i 3.0T super conduct MR unit, 51 cases of patients(including 1 case of pilocytic astrocytoma, 23 cases of diffuse astrocytoma, 3 cases of anaplastic astrocytoma and 24 cases of glioblastoma) were performed with 2-D mutlti-voxel and single-voxel PRESS 144ms while 2-D mutlti-voxel STEAM35ms were performed on 18 cases. Functool software package was used for post-processing of spectrum. The metabolites of low and high malignant astrocytic tumor were quantified in the different part of the tumors including parenchyma, edematous region surrounding the tumor and opposite side. For the differential diagnosis, 6 cases of subacute cerebral infarction were compared with low malignant astrocytic tumor, 6 cases of inflammation and 7 cases of lymphoma were compared with high malignant astrocytic tumor. Results: 1.No significant difference can be seen in the contrast-enhanced part and the non-enhanced part in low malignant astrocytic tumor while significant difference can be obtained in Cho/Cr, Cho/NAA> NAA/Cr between central part and peripheral part of the tumor. 2. Significant difference can be obtained in Cho/NAA, NAA/Cr between contrast-enhanced part and non-enhanced part in high malignant astrocytic tumor. 3. Significant difference obtained in Cho/Cr.. Cho/NAA between edematous region 1 surrounding low malignant astrocytic tumor and the opposite side was not revealed between edematous region 2> 3 and the opposite side. 4. Significant difference can be obtained between edematous region K 2, 3 and the opposite side in high malignant astrocytic tumor. 5. Significant difference Cho/Cr* Cho/NAA can be obtained between low and high malignant astrocytic tumor in parenchyma and the edematous region. In parenchyma, the value is 1.965±0.101 and 4.037±0.497 in the prior one while3.513±0.299 and 7.103±1.033 in the latter one. 6. Significant difference can be obtained in Cho/NAA among the subacute cerebral infarction, inflammation, low and high malignant astrocytic tumor. No significant difference can be obtained between the edematous part of the lymphoma and high malignant astrocytic tumor. Conclusion: 1 .Metabolic difference can be revealed between the peripheral and the central part of the low malignant astrocytic tumor, while not revealed between contrast-enhanced and non-enhanced part. 2. Significant difference in Cho/NAA -. NAA/Cr can be obtained between contrast-enhanced and non-enhanced part in the high malignant astrocytic tumor. 3. Degree of infiltration in the edematous region surrounding tumor can be used to differentiate low and high malignant astrocytic tumor. 4Considering 2.5 as a threshold of CNI is benefit for differential diagnosis among subacute cerebral infarction, inflammation, low and high malignant astrocytic tumor. S. Differential diagnosis between lymphoma and high malignant astrocytic tumor is difficult in some degree. PartIV: 3.0T MR Spectroscopy imaging in the evaluation of cerebral meningiomasPurpose: To reveal the biological feature of benign and malignant meningioma in MR spectrum and to evaluate the value of MRSI in the differential diagnoses between benign meningioma % malignant meningioma and haemangiopericytoma. Materials and methods: Using a GE signa VH/i 3.0T super conduct MR unit, 26 cases of patients (including 20 cases of benign meninigioma and 6 cases of malignant meningioma) were performed with 2-D mutlti-voxel and single-voxel PRESS 144ms and 13 cases were performed with STEAM35ms. Differential diagnoses were performed among the tumors by comparing the MR spectrum of parenchyma and edematous region surrounding the tumor. Functool software package was used for post-processing of spectrum. Results: l.Ala peak is characteristic of the menigioma, while no significant difference of Ala/CrO is revealed among meningiomas. 2.No significant difference were revealed between fibrous and meningothelial meningioma except Lipl.3/CrO. 3. Significant difference was revealed in Cho/CrO between parenchyma and edematous region of benign and malignant meningioma. In parenchyma, Cho/CrO is 1.898±0.112 in the prior one and 2.501±0.187 in the latter one. 4. No Ala peak is revealed in haemangiopericytoma. Significant difference of...