The Predictive Value Of MRI Image Features And Texture Analysis In Gleason Grading Of Prostate Cancer

objective To evaluate the predictive value of MRI imaging features in Gleason grading of low,middle and high grade prostate cancer(Prostaticcancer,PCa).Methods 54 patients with 63 lesions of PCa confirmed by biopsy or radical surgery in the first Hospital of Lanzhou University from January2018 to June 2020 were analyzed retrospectively.The age of the patient,PSA(Prostatespecificantigen,PSA)and its related indexes,as well as the quantitative and qualitative image features of the lesions on T2 WI images were included.Independent sample t-test or Mann-Whitney U was used to test the differences of age,PSA and its related indexes,MRI quantitative image features between low and medium-high risk groups PCa,and chi-square test was used to analyze the differences of qualitative image features between groups.The area(AUC),under the curve was calculated by the receiver operating characteristic(ROC)curve to evaluate the diagnostic efficiency.Results There were significant differences in PSA,f PSA,PSDA,tumor maximum diameter,relative lesion diameter,tumor volume,lesion shape,prostate capsule,focus capsule,signal and boundary among PCa groups(P < 0.05 or P < 0.01).The results of ROC analysis showed that the tumor volume and AUC of PSAD were the largest.Conclusion The imaging features of MRI and PSA and its related indexes can improve the diagnostic value of low,middle and high grade PCa and contribute to the non-invasive assessment of the risk of prostate cancer before operation.objective to evaluate the value of ADC texture analysis in predicting the Gleason grade of low,medium and high grade prostate cancer(Prostaticcancer,PCa).Methods 63 lesions of PCa confirmed by puncture biopsy or radical operation in the first Hospital of Lanzhou University from January 2018 to June 2020 were analyzed retrospectively.The histogram texture feature parameters(entropy,skewness,kurtosis,non-uniformity,standard deviation,maximum,minimum,average,median and 10%,25%,50%,75%,90% quantile pixel values)were obtained by manually drawing ROI,layer by layer on the ADC diagram by using Fire Voxel software.Independent sample t-test or Mann-Whitney U test was used to compare the differences of entropy,skewness,kurtosis,standard deviation,maximum,minimum,average,median,non-uniformity and texture parameters such as 10%,25%,50%,75%,90% quantile pixel values between low and medium-high risk groups PCa.The diagnostic efficacy was evaluated by using the receiver operating characteristic(ROC)curve to calculate the area(AUC),under the curve,and the combined diagnostic efficacy of it and MRI quantitative imaging parameters in the differential diagnosis of PCa in low and medium-high risk groups was evaluated.Logistic regression analysis was used to screen the independent risk predictors of texture parameters for prostate cancer in medium-and high-risk groups.Results There were significant differences in entropy,skewness,minimum,average,median and 10%,25%,50%,75% quantile pixel values among PCa groups(P < 0.05).The results of ROC analysis showed that the AUC of skewness and entropy were larger,and the diagnostic efficacy of skewness and entropy combined with tumor volume and relative lesion diameter was stronger than that of single use.Entropy,median,10% quantile and 50% quantile were selected as independent risk predictors of prostate cancer in medium-high risk group by logistic regression analysis.Conclusion The texture parameters of ADC map may be a valuable tool for predicting PCa in medium-and high-risk groups,and the combination of specific texture parameters extracted from ADC images and specific quantitative features of MRI images may be an additional tool for predicting PCa in medium-and high-risk groups.