The Discussion Of Correlation Between MR Apparent Diffusion Coefficient And Gleason Score Of Prostate Cancer

Part one.The analysis of correlation between apparent diffusion coefficient and apparent diffusion coefficient ratio and Gleason score of prostate cancer and their evaluation of preliminary applicationObjective:To discuss the value and diagnostic efficiency of the MR apparent diffusion coefficient and apparent diffusion coefficient ratio for discriminating low and intermediate/high Gleason score of prostate cancer.Materials and Methods:A total of 50 patients went transrectal ultrasound guided prostate systemic biopsy and pathologically confirmed as prostate cancer who had routine MRI and DWI-MRI（include b values 100 and 1000）scan at 3.0T MR from Jan.2015 to Dec.2017 were retrospectively analyzed.Upload the DWI data to the workstation,use b value 100,1000 to reconstruction ADC map for measure and analysis.ROI take the largest level of lesion,along the cancer contour manual sketch.Sketch other ROI intransition zone and peripheral zone with PI-RADS score≤2 and pathological report ofnormal tissue,bladder and pelvic floor muscles,all ROI measurements averaged three times.Calculate the ratio of ADC to other normal parts of the lesion,,acquire r ADC lesions/transitional zone,r ADC lesions/peripheral zone,r ADC lesions/urine,r ADC lesions/pelvic floor muscles,r ADC lesions/offside（if the lesion located in the transitional zone,r ADC lesions/offside is equal to r ADC lesions/transitional zone,if located in the peripheral zone,then r ADC lesions/peripheral zone）.Using the Spearman rank correlation analysis to evaluate the of correlation between ADC/ADC ratios and the Gleason scores of prostate cancer,at the same time in the value of the analysis results,select the parameters related to Gleason scores of prostate cancer.The diagnostic performance of identifying low GS and intermediate/high GS Gleason score of prostate cancer were compared by receiver operating characteristic（ROC）curve and area under curve（AUC）.Results:A total of 50 patients with prostate cancer,Gleason score 6 was 18,Gleason score 7 was 16,Gleason score 8 was 9,Gleason score 9 was 7.ADC value and ADC ratios were inverse correlated to Gleason score of prostate cancer（P<0.05）.The correlation coefficient was ADC value-0.609,r ADC cancer/transition zone-0.612,r ADC cancer/peripheral zone-0.464,r ADC cancer/pelvic floor muscle-0.481,r ADC cancer/urine-0.589,r ADC cancer/offside-0.568.The ADC value of prostate cancer and the ratio of ADC in each group identified the diagnostic efficacy of low GS and intermediate/high GS prostate cancer by comparing the area under the curve.The differential diagnosis threshold was ADC value less than 7.123×10^-4,r ADC cancer/transitionzone<0.466,r ADC cancer/peripheral zone<0.367,r ADC cancer/pelvic floor muscle<0.638,r ADC cancer/urine less than 0.252,r ADC The cancer/offside<0.478,the ADC value of the prostate cancer and the AUC value of the ADC ratio of each group were 0.798,r ADC cancer/transition zone 0.778,r ADC cancer/peripheral zone 0.682,r ADC cancer/pelvic floor muscle 0.771,r ADC cancer/urine 0.757,r ADC cancer/offside 0.738.The ADC value of prostate cancer and the ratio of ADC in each group could identify the low GS and intermediate/high GS prostate cancer.The ADC value was better than the r ADC cancer/peripheral zone（Z=2.036,p<0.05）,and there was no statistical difference between ADC value and the ADC ratio of the other groups.Conclusions:This study shows that the ADC value and the ratio of the prostate cancer foci are correlated with the Gleason score.There is no significant difference between the ADC value and the ADC ratio of the other groups in the differential diagnosis of low GS and intermediate/high GS prostate cancer except the ADC value is superior to the r ADC cancer/peripheral zone.Both of them are helpful to evaluate the Gleason score of prostate cancer before operation.Part two.MR apparent diffusion coefficient map texture analysis quantitative parameters for identifying low GS and intermediate/high GS prostate cancerObjective: Compare the differences of ADC map texture analysis quantitative parameters between low GS and intermediate/high GS prostate cancer,and discuss their value in identifying low GS and intermediate/high GSprostate cancer.Methods: 50 patient with histologically confirmed as prostate cancer were examed using DWI performed at 3T employing an extend b-value range from 0 to 2000 s/mm2（b=0,10,20,50,100,200,500,1000s/mm2）.Tumors were classified into two groups: Gleason score≤6（n=18）,Gleason score≥7（n=32）.According to the pathological results of each case,the prostate cancer area was predefined in its corresponding ADC map,the region of interest was set up,and the region was delineated along the edge of the cancer area（including the whole cancer range,but not beyond the range of the cancer range）,and the three-dimensional volume of interest（VOI）was obtained.Fire Voxel software was used for texture analysis of the obtained VOI,and the quantitative analysis parameters such as mean value,non-uniformity,peak value,skewness and entropy were obtained.The independent sample t test was used to compare the differences in the quantitative parameters of the ADC map between the low GS and intermediate/high GS group of the prostate cancer.The area under the curve was obtained by using the working characteristic curve of the subjects,and the quantitative parameters of the texture analysis were determined to identify the threshold and the diagnostic efficiency of the low GS and intermediate/high GS group of the prostate cancer.Results: 1.ADC map texture analysis results showed that the mean value,inhomogeneity,skewness,kurtosis and entropy of low GS prostate cancer were 0.6798±0.8823,0.1871±0.6318,0.4857±0.4417,0.2302±0.8196 and 3.9781±0.1482,respectively,while the mean value,inhomogeneity,skewness,kurtosis,entropy of intermediate/high GS prostate cancer were 0.5872±0.8321,0.1813±0.7643,0.6407±0.4786,0.9203±1.3100 and 3.8096±0.2943.There were significant differences in mean,kurtosis and entropy between prostate cancer in low GS and intermediate/high GS（T values were 3.694,-2.019,2.265,P < 0.05）.There was no significant difference in the degree of heterogeneity and skewness between the low GS group and the intermediate/high GS prostate cancer（t = 0.270,-1.129,P > 0.05）.2.The quantitative parameters of ADC map texture analysis were used to identify the AUC value and diagnostic efficacy of low GS and intermediate/high GS prostate cancer respectively: the mean value was 0.759,the sensitivity was 0.759%and the specificity was 88.89%.Kurtosis discriminating low GS andintermediate/high GS prostate cancer AUC value is 0.759,sensitivity is 71.87% and specificity is 61.11%.The AUC value of entropy for discriminating low GS and intermediate/high GS prostate cancer was 0.646,sensitivity 37.50% and specificity 94.44%.Compared one by one,there was no significant difference in mean / kurtosis,mean / entropy or kurtosis / entropy.Conclusion: The quantitative parameters of ADC map texture analysis in the differential diagnosis of low GS and intermediate/high GS prostate cancer showed that the average value and entropy of the low GS prostate cancer were higher than those in the intermediate/high GS group,and the kurtosis was lower than that of the intermediate/high GS group.The differences were statistically significant.There was no significant difference in AUC between prostate cancer and intermediate/high GS group.Therefore,the mean value,kurtosis and entropy in the quantitative parameters of ADC map texture analysis are different between low GS and middle and high GS prostate cancer,which is helpful for the differential diagnosis of low GS and intermediate/high GS prostate cancer.