The Value Of CT Texture Analysis In The Differential Diagnosis Of Renal Chromophobe Cell Carcinoma And Renal Oncocytoma

Objective: To investigate the value of radiomics combined with renal enhanced CT texture analysis and machine learning in the identification of chromophobe cell renal carcinoma(CCRC)and renal oncocytoma(RO).Materials and methods: CT images of 64 cases of CCRC and 31 cases of RO renal tumor lesions confirmed by pathology after the first affiliated hospital of China medical university from January 2013 to July 2018 were retrospectively analyzed.ITK-SNAP version 4.11.0 software was used to delineate the region of interest and A.K.Version v3.0.0.r software was used to extract texture features.Random forest model is established by texture features included in random forest algorithm.Logistic regression was used to evaluate the discriminative efficacy of the established model in CCRC and RO,A p ? 0.05 was considered statistically significant.Results: The results of AUC for the characteristics of mixed selection in corticomedullary phase,nephrographic phase and combination of the two phases were 0.73,0.74 and 0.83,respectively.Logistic regression was used to evaluate the texture parameters with the highest AUC values in the first 20 of the weighted values after mixing,and the AUC values were 0.876,0.861 and 0.945,respectively.Conclusion: The study of texture analysis based on CT images has clinical value in the differential diagnosis of CCRC and RO.