The Application Of Blood Oxygen Level Dependent Imaging And Diffusion Tensor Imaging In Intracranial Tumors

Intracranial tumors which have high morbidity and disability rate, commonly occurs in central neural system. In fact, neurosurgeons are still looking for a proper way to solve the problem: how to resect the tumor sufficiently, protect the normal nerve tissue effectively, at the same time, improve the patients'prognostic quality of life prominently.For the past few years, DWI (diffusion weighted imaging), MRS (magnetic resonance spectroscopy), DTI (diffusion tensor imaging) and BOLD (blood oxygen level dependent imaging) developed rapidly. BOLD with both high temporal resolution and high spatial resolution, is able to display the posittion and area of the cortex activation, which can be fused with anatomy imaging correctly. Newly developed magnetic technology, DTI, which is based on DWI, can be realized to timingly and quantitatively analyze the tissue diffusion properties of water molecules in three dimensions. BOLD and DTI can exactly discover whether the cortex activation and the white matter tract are involved in intracranial tumors or not. Imaging study can provide more appropriate selections of surgical approach and techniques for neurosurgeons, help the patients who are suffering from intracranial tumors.We had studied 27 patients with intracranial tumors. They all underwent conventional MR imaging, BOLD and DTI respectively. The purpose of this article was to study the significance in distinguishing the boundary of the intracranial tumors with BOLD and DTI, to define the relationship between the intracranial tumor and the cortex activation and white matter tract. This study could propose a solution to serve the purpose to decide a surgical approach for surgeons and contribute to improve the patients'quality of life after resection. Methods:A total of 27 patients with gliomas and meningiomas and 3 healthy volunteers underwent conventional MR imaging (T1WI, T2WI and FLAIR),diffusion tensor imaging and blood oxygen level dependent imaging, using SEIMENS 3.0T Trio Tim superconductive magnetic resonance imager. Conventional MR imagings (T1WI, T2WI and FLAIR) were performed with FLASH and turbo spin echo sequence (slice thickness: 5.0mm; slice interval: 1.5mm; FOV: 220mm×220mm; T1WI: TR/TE= 440/2.46ms; T2WI: TR/TE= 5000/93ms; FLAIR: TR/TE= 8000/93ms, TI= 2371.5ms). 20 patients underwent enhanced MRI with magnevist as contrast agent (0.2ml/kg). All patients underwent anatomy (t1-mpr-ns-sag-p2-iso) scan (slice thickness: 1.0mm; slice interval: 0.5mm; FOV: 250mm×250mm; TR/TE=2300/2.53ms). Diffusion tensor imaging was performed with a single shot, spin echo, echo-planar, diffusion-weighted pulse sequence(64 different motion probing gradient directions, TR/TE= 5700/93ms, b=0s/mm2 and b= 1000s/mm2, slice thickness: 3mm, slice interval: 0mm, FOV: 230mm×230mm. When underwent blood oxygen level dependent imaging (slice thickness: 3.0mm; slice interval: 0.75mm; FOV: 192mm×192mm; TR/TE= 3000/30ms), patients performed alternative activation, and the data was collected by 6 groups, each group for 30 seconds. We used Neuro 3D software of the workstation to analyse the data. On FA map, 3 or 4 ROIs were chosen, including solid and cystoid portion of the tumor, edema area and the edge of edema, as well as the contralateral normal brain regions were chosen. A statistic analysis was given after calculating the average FA value of different ROIs. Diffusion tensor tractography was reconstructed by"seed method".Results:On conventional MR images: Gliomas were abnormal in signal intensity, with slight mass effect and a slight edema, with low signal in T1WI and high signal in T2WI/FLAIR. When there was cystoid portion in the tumor, it may be lower signal in T1WI and higher signal in T2WI/FLAIR. The tumors'boundary was vague. Gliomas showed slight enhancement, no enhancement, or complicated enhanced signals according to different grades. Meningiomas which showed sharp boundary were round normally with obvious mass effect, and a base attachment to endocranium, with the same signal with gray matter. Enhancement MRI of meningiomas were uniform bright. On FA images, all giomas and edema areas showed very low signal, cystoid portions showed lower signal. All meningiomas showed low signal in the solid portions. The mean FA value of all gliomas'ROIs was significant lower (P<0.05 or P<0.001) than that of normal brain. A decreasing tendency of mean FA value could be observed as follow: edge of edema areas (0.6527±0.1330) was hightest, followed by edema areas (0.2740±0.2304) and solid portions (0.0889±0.0325) then cystoid portions (0.0602±0.0207). The mean FA value of all meningiomas'solid portions and edema areas'ROIs was significant lower(P<0.001) than that of normal brain, but the mean FA value of edge of edema areas was higher than that of normal brain. We think the reason is that the white matter tract near the tumor is compressed, and some fiber tracts are tighter than that of normal. The mean FA value of all intracranial tumors'ROIs had statistical significance (P<0.05) when they were compared with each other. So we can use FA value to distinguish the boundaries of the intracranial tumors. White matter fibers could be well distinguished by different colors on DEC images. We can observe gliomas on DEC images as follow: Solid portions of the gliomas were dark colors, and fiber tracts were vague; cystoid portions were deep dark colors; edema areas were dark colors too; slightly compressed with normal, bright colors beyond edema areas. DEC images show all the meningiomas as follow: solid portions were uneven dark colors, edema areas were dark colors, the boundaries were vague, the rounding fibers were compressed. 5 cases of all gliomas merely involved CST, 3 cases merely involved corpus callosum, and 4 cases involved both CST and corpus callosum. We describe that as follow: Fiber tracts were destruct and discontinuous in the tumors; the fibers nearby were separate and a little discontinuous, their direction were changed for the tumors'compression (accompany with colors changed). 7 cases of all meningiomas merely involved CST, 8 cases involved both CST and corpus callosum. We describe that as follow: Fibers structure was approximately normal in the tumors, but the direction was changed (accompany with colors changed). The compressed fibers were dense or separate. We can conclude that fiber tracts were destruct and discontinuous in gliomas, and fibers near meningiomas were deviated, dense and rare. We can observed in BOLD that: some intracranial tumors didn't involve the cortex activation at all, so the cortex activations at both sides were symmetry, same area and the signal was consistent. In the article, One case of glioma occurred in the left basal ganglia (glioblastoma, WHO IV grade) had large edema area, and caused the ipsilateral cortex activation reduced. Some meningomas that occurred in cortex activation caused asymmetric of cortex activations, the ipsilateral cortex activation's position was deviated and configuration changed ( according to the tumors'location, shape, oppression in different directions).Conclusion:1. BOLD can display the relationship between the intracranial tumors and the cortex activations, provide more appropriate selections of surgical approach and technique for neurosurgeons by imaging study.2. DTI proposes a solution to select surgical approach, shows the correlation between the white matter tracts and nearby tumors directly.3. FA value contributes positively to study the boundaries of the intracranial tumors.4. The combination of BOLD and DTI is useful for preoperative investigation, especially assisting surgical planning, e.g. deciding on a surgical approach, and also helpful to improve the patients'postoperational quality of life.