Research On Gliomas Grading Of MRI With ASL And DTI Techniques

Gliomas, one kind of the most common malignant brain tumors, is seriously harm to human health. It is a hot issue to take advantage of perfusion magnetic resonance imaging (PWI) and diffusion tensor imaging (DTI) to evaluate the state of gliomas in the field of current medical image.Arterial Spin Labeling (ASL) is a completely new noninvasive perfusion MRI method, which could calculate the hemodynamic parameter of the brain tumor——cerebral blood flow (CBF). DTI could be used to analyze diffusion coefficient of water in the brain tissue and obtain the concerned diffusion parameters including fractional anisotropy (FA) and mean diffusion (MD).This paper emphasize on the issue of quantitative assessment of gliomas. The use of multimodality MRI, which combines the technologies of DTI and ASL, has solved several critical problems on gliomas grading. Main achievement and innovation of this study as follows:(1) Programmed and processed the ASL magnetic resonance image, which integrated magnetic resonance parameter calculation algorithm. The image processing procedure included motion correction, denoising, extracting brain, perfusion weighted image calculation. Finally the CBF maps were acquired.(2) Because of the SNR of ASL image is low, it is necessary to acquire images repeatedly. However, it needs a long time to collect images repeatedly. During this long time, the patients felt uncomfortable and moved slightly, which lead to artifacts of the data. A denosing filter method based on non-local means (NLM) was raised. This new filter method realized the destination of higher accuracy of30-40% compared to conventional approach at the condition of less acquiring time.(3) The calculation results of blood flow and diffusion coefficient is highly dependent on the position of the chosen ROI (Region of Interest). A reproducible ROI selection strategy of multimodality MR image was proposed to obtain stable parameter values, with the dependence on human ROI selection decreased.(4) Quantitative assessment on divisibility to gliomas grading of single-modality MR and multimodality MR was analyzed. It is proved that the use of ASL combined with DTI has higher separability value than the single modality imaging technology. The parameters combination with significant separability has been found.It is concluded that the use of multimodality MR method could achieve the goal of grading gliomas, rather than calculate the statistical differences only.