Correlation Between Texture Parameters And Gleason Grade Of Prostate Cancer Based On Biparametric MRI

Objective:To establish a joint prediction model by analyzing the important characteristic parameters of the histogram based on biparametric MRI（bp MRI）combined with the data of key clinical indicators,and to analyze its predictive value for prostate cancer of GS≤3+4（low-risk group）and GS≥4+3（high-risk group）.Methods:99 patients diagnosed with prostate cancer after radical prostatectomy or biopsy,prostate cancer patients obtained according to pathological puncture or radical prostatectomy were divided into GS≤3+4（low-risk group）47 cases and GS≥4+3（high-risk group）52 cases according to GS grading results.All patients underwent preoperative 3.0t-MRI.Texture parameters of axial T2WI and ADC images were extracted by special software,and the extracted texture parameters and clinical data were analyzed by single factor.The single factor predictors meaningful for GS grouping were selected,and then logistics regression analysis was conducted to observe the predictive efficiency of independent predictors and joint indicators by ROC curve.Results:the high-risk group of patients with ADC_Energy,ADCRobust Mean Absolute Deviation,T2WIEnergy,T2WIRobust Mean Absolute Deviation,t-PSA and tumor volume were higher than in low-risk group.Multiple regression analysis through the Logitic get a rise in serum t-PSA ADC_{Robust Mean Absolute Deviation},T2WI_{Robust Mean Absolute Deviation}and associated with GS classification.ROC curve,according to the results when through ADC_{Robust Mean Absolute Deviation},T2WI_{Robust Mean Absolute Deviation}and t-PSA forecast individually,the area under the curve values between 0.751-0.877 range.However,when the three are combined,the area under ROC reaches 0.959.Conclusion:A rise in t-PSA、ADC_{Robust Mean Absolute Deviation、}T2WI_{Robust Mean Absolute Deviation}and associated with the GS classification of PCa patients,Combined application of the three methods can further improve the efficacy of the grading prediction of GS≤3+4 and GS≥4+3 in prostate cancer.