The Application Of Models Based On Multiple B-value DWI In Fat-poor Angiomyolipoma And Clear Cell Renal Cell Carcinoma

Objective:To evaluate the diagnostic value of fractional order calculus model（FROC）,intravoxel incoherent motion model（IVIM）and diffusion kurtosis imaging model（DKI）based on multiparameter diffusion weighted imaging（DWI）to differentiate pathology grade of clear cell renal cell carcinoma（ccRCC）.Materials and Methods:A total of 62 ccRCCs who were confirmed by pathology underwent T2WI and 11b-values（0-3000s/mm²）DWI.The standard apparent diffusion coefficient（ADC）,the FROC parameters(Diffusion coefficient D_froc,fractional order parameterβ,microstructural quantityμ),the DKI parameters（mean kurtosis MK,and mean diffusivity MD）and the IVIM parameters（true diffusivity D,pseudo-diffusion coefficient D*,perfusion fraction f）were calculated and compared between high-grade//and low-grade ccRCC.The values to diagnose pathological grading were assessed by receiver operating characteristic（ROC）analysis.Results:It is worth that high-grade ccRCC showed significantly lower ADC,Dfroc,β,D and MD（P<0.05）,while higherμand MK（P<0.05）than low-grade ccRCC.However,D*（P=0.922）and f（P=0.217）were not markedly difference between the high and low grade ccRCC.The FROC parameterμperformed noticeably well with optimal area under the curve（AUC=0.843）,followed by Dfroc（AUC=0.837）,β（AUC=0.803）,MK（AUC=0.781）,ADC（AUC=0.744）,D（AUC=0.727）and MD（AUC=0.712）.The combination of Dfroc,μandβproduced added diagnostic efficiency with the AUC increased to 0.887 compared withμ.Conclusion:Diffusion-dependent parameters(ADC,D_froc,β,μ,MD,MK and D)could quantificationally differentiate high and low grade ccRCC.Neither of perfusion-related parameters（D*and f）exhibit statistical significance.The FROC parameterμprovide most optimal value for grading ccRCC,especially combined D andβ.We also found that compared with the DKI and IVIM,the FROC diffusion model may have the best diagnostic efficacy to high-and low-grade ccRCC.Objective:To discuss the diagnostic value of fractional order calculus model（FROC）based on multiparameter diffusion weighted imaging（DWI）to differentiate clear cell renal cell carcinoma（ccRCC）from fat-poor angiomyolipoma（fp AML）.Materials and Methods:A total of 43 ccRCCs and 16 fp AMLs who were confirmed by pathology underwent 11 b-values（0–3000s/mm2）DWI before surgery.The standard apparent diffusion coefficient（ADC）,and the FROC parameters（Diffusion coefficient Dfroc,fractional order parameter β,microstructural quantity μ）,were calculated and compared between ccRCC and fp AML.The values to diagnose ccRCC and fp AML were assessed by receiver operating characteristic（ROC）analysis.Results:It is worth that ccRCC showed significantly higher ADC,Dfroc and β,while lowerμ than fp AML [（1.92±0.41）×10^-3mm²/s vs.（1.28±0.43）×10^-3mm²/s],[（1.49±0.32）×10-3mm²/s vs.（0.94±0.22）×10^-3mm²/s],[（0.61±0.12）vs.（0.49±0.14）] and[（7.94±1.13）×10^-3mm vs.（8.51±0.32）×10^-3mm,P < 0.001].The FROC parameter Dfroc performed noticeably well with optimal area under the curve（AUC = 0.912）,followed by ADC（AUC = 0.855）,μ（AUC = 0.823）,β（AUC = 0.758）.When Dfroc was combined with β and μ produced top diagnostic efficiency with the AUC increased to0.929,the sensitivity and specificity were 93.0% and 87.0% respectively.Conclusion:Diffusion-related parameters ADC,D_froc,β and μ could quantificationally differentiate ccRCC and fpAML.The FROC parameter D_froc provide most optimal value for distinguishing ccRCC from fp AML,when Dfroc combined with β and μ,the diagnostic efficiency was further improved.