Study On Perfusion Heterogeneity Of Glioma Based On Time Shift Analysis

Glioma is the most common primary brain tumor,which has a high malignancy,high invasiveness,and high mortality.At present,the determination of the malignancy of glioma much depends on conventional medical imaging.The advances in imaging technology enable the classification of glioma’s malignancy through analyzing texture,signal characteristics(signal intensity correlation),and brain function(topological properties)base on hierarchical characteristics of multimodal data.However,the process takes longer and is likely to discomfort the patient.Thus,it is essential to study the heterogeneity of glioma to provide referenceable characteristic parameters for determining malignancy.This graduation design constructs time-shift maps using time shift analysis(TSA),classifies the malignancy of glioma by perfusion heterogeneity,and establishes various hierarchical characteristic parameters.After pre-processing the gathered image,the study constructs time-shift maps using global mean and superior sagittal sinus signal as reference sequences respectively.Finally,the study calculates the statistics(TSA mean,standard deviation,variance,skewness,and kurtosis)of both ipsilateral and contralesional hemispheres of test subjects’ region of interest(ROI)and determines the malignancy of glioma using paired-samples T test for each subject and non-parametric statistical analysis for two groups of hemispheres.The study involves 20 glioma patients and six healthy control subjects.The analysis of perfusion heterogeneity of glioma using TSA maps shows hyperemia and ischemia exist in glioma area,and a small range of advanced/hysteresis signals exist in the nearby healthy areas and the contralesional hemispheres of the glioma area.These are in line with the high invasiveness of malignant glioma and that it has a distal,global effect on the brain.The study discovers a higher heterogeneous perfusion in the ROI of the tumor and statistical differences(p<0.05)on all characteristic parameters between the two hemispheres of glioma patients while there are no such differences in those of the healthy subjects.Also,the study discovers TSA mean,variance,and kurtosis is capable of classifying glioma’s invasiveness with enough statistical differences(p<0.05).Using global mean as reference,the TSA mean,variance,and kurtosis of low/high grade glioma are 1.05±0.30/1.41±0.36,1.19±0.67/2.14±0.95,and 2.01±1.98/0.7±2.30.The results while using superior sagittal sinus signal as reference are 1.31±0.36/1.74±0.32,1.85±1.07/3.14±1.01,and1.66±2.22/-0.30±1.04.The research provides supplemental characteristic parameters for the existing resting-state functional magnetic resonance imaging and approaches for precise treatment through heterogeneity detection.