Application Of 1H-MRS Combined With DWI Sequence In The Pathological Grading And Correlation Of Ki-67 Index Of Glioma

ObjectiveTo explore the application of hydrogen proton magnetic resonance spectroscopy(~1H-MRS)combined with diffusion-weighted imaging(DWI)in the pathological grading and prediction of cell proliferation of glioma,so as to improve the accuracy of preoperative diagnosis of glioma.MethodsWe prospectively collected the clinical and imaging data of patients with suspected glioma in our hospital from June 2019 to March 2021,All patients underwent conventional MRI scan and enhanced scan,DWI,multi-voxel ~1H-MRS scan,According to the postoperative pathology,there were 72 patients with complete imaging examination data and the pathological diagnosis was glioma,including 44males and 28 females,with an average age of 47.3 years.The value of NAA/Cr,Cho/Cr,Cho/NAA,ADCt in tumor parenchyma,ADCh value of normal brain parenchyma in the contralateral mirror area,rADC(rADC=ADCt/ADCh)and Ki-67 index were measured as statistical indicators.According to postoperative pathology(refer to the 2016 WHO classification and grading standards for central nervous system tumors),the cases were divided into two groups:low-grade glioma group(WHOⅠ～Ⅱ,n=20)and high-grade glioma group(WHOⅢ～IV,n=52),analyze the differences of the age,gender,NAA/Cr,Cho/Cr,Cho/NAA,ADCt value,ADCh value,rADC and Ki-67 between the two groups of patients,and draw the receiver operating characteristic curve(receiver operating characteristic curve,ROC),find out the best diagnostic threshold of glioma grading and its sensitivity and specificity,objective to evaluate whether DWI combined with ~1H-MRS is superior to DWI or ~1H-MRS in the diagnosis of glioma grading;Finally,Spearman correlation analysis was used to test the correlation between the statistically significant indicators of DWI、~1H-MRS and Ki-67 index.Results1.There were significant differences in age,Cho/Cr,Cho/NAA,ADCt,rADC and Ki-67 between low-grade glioma group and high-grade glioma group(P<0.05).Age,Cho/Cr,Cho/NAA,Ki-67 in the low-grade glioma group[31.2±13.3,1.755(1.28),1.94(1.99),0.05(0.03)]were all smaller than the high-grade glioma group[53.5±15.2,2.83(2.4),4.395(3.53),0.30(0.32)];ADCt value and rADC in low-grade glioma group(1.46±0.51,2.00±0.64)were higher than those in high-grade glioma group(0.97±0.24,1.31±0.32).There was no significant difference in gender,NAA/Cr and ADCh between low-grade glioma group and high-grade glioma group(P>0.05).2.Cho/Cr,Cho/NAA,ADCt and rADC were used as diagnostic indexes to distinguish low-grade glioma group and high-grade glioma.The diagnostic indicators of gliomas can be seen from the ROC curve.The area under the curve(AUC)of Cho/Cr,Cho/NAA,ADCt and rADC values were 0.738,0.767,0.790,and 0.805respectively(all P<0.001),Found out the critical value according to the maximum Youden index,when the values of Cho/Cr,Cho/NAA,ADCt,and rADC were 2.3,2.03,1.39,and 1.73,respectively,the sensitivity and specificity for grading diagnosis of glioma were(75.00%,69.23%),(65.00%,84.62%),(60.00%,98.08%),(70.00%,90.38%).3.The largest area under the curve was the best distinguishing index for low-and high-grade glioma groups,and Cho/NAA was the best identification index of~1H-MRS,its AUC is 0.767,sensitivity and specificity are 65.00%,84.62%,respectively;rADC was the best identification index of DWI,its AUC was 0.805,sensitivity and specificity were 70.00%,90.38%,respectively;Using the best identification index of the two technologies:Cho/NAA combined with rADC as a variable,Logistic regression was used to generate prediction probability(PRE),the ROC curve showed that the area under the curve was the largest,it was 0.824,the sensitivity and specificity were 70.00%and 92.31%respectively.It can be seen that DWI combined with ~1H-MRS was better than the single DWI or ~1H-MRS technology in the grading diagnosis of low-and high-grade gliomas.4.Correlation analysis between the statistically significant indicators of DWI and ~1H-MRS:Cho/Cr,Cho/NAA,ADCt and rADC were correlated with Ki-67(all P<0.01),and the correlation coefficients were 0.409,0.420,-0.496,-0.482,respectively.Conclusions:1.On the basis of conventional MRI examinations,both ~1H-MRS and DWI sequences can improve the accuracy of grading diagnosis of glioma.2.Cho/Cr,Cho/NAA,ADCt and rADC have great clinical value in distinguishing low-and high-grade gliomas.When their values are 2.3、2.03、1.39and 1.73,they can be used as reference values to distinguish low and high grade gliomas.3.DWI combined with ~1H-MRS is better than the single DWI or ~1H-MRS technology in the grading diagnosis of low-and high-grade gliomas.4.Cho/Cr and Cho/NAA are positively correlated with Ki-67 index of glioma,ADCt and rADC are correlated with Ki-67 index of glioma.