Abstract：Objective To construct a nomogram based on clinical variables and selected CT radiomic features for preoperative
differential diagnosis of clear cell renal cell carcinoma (ccRCC) and fat-poor angiomyolipoma (fp-AML). Methods A total of 87
patients with pathologically identified small renal masses with tumor diameter ≤ 4 cm were involved in this retrospective study,
including 54 patients with ccRCC and 33 patients with fp-AML. All patients underwent three phase CT scans, including unenhanced
phase (UP), corticomedullary phase (CMP) and nephrographic phase (NP). Radiomics features were extracted from the CT images.
Using the Synthetic Minority Oversampling Technique algorithm for sample balancing, we screened the characteristics of radiomics,
and constructed three radiomics models: UP model, CMP model, and NP model. Radiomics features were selected and three
radiomics models were constructed, including UP model, CMP model and NP model. The corresponding radiomics scores were
calculated and the optimal model was screened out. Statistically significant clinical variables were used to contruct clinical model,
and combine them with the optimal model radiomics score to construct joint model and draw the nomogram. The diagnostic efficacy
of clinical models, radiomics models, and joint models was evaluated using receiver operating characteristic curves. The reliability
and clinical value of nomograms were evaluated using correction curves and decision curves, respectively. Results The performance
of the UP model was better than the other two models, and there was strong collinearity among the UP, CMP and NP models; after
Logistic regression analysis, three clinical variables with statistical significance were selected and combined with UP radiomics
model to construct a combined model; the area under the curve of clinical model, radiomics model and the combined model were 0.704
(95% CI: 0.511-0.898), 0.846 (95% CI: 0.698-0.994) and 0.944 (95% CI: 0.876-1.000) respectively; the calibration curve showed
that the predicted values of nomogram were in good agreement with the actual pathological results; the decision curve showed
that nomogram had good clinical application value. Conclusion The nomogram constructed based on preoperative CT radiomics
characteristics and clinical variables has good efficacy in the differential diagnosis of ccRCC and fp-AML.
周鹏，王冬青，刘护丽，张礼荣. 基于CT影像组学列线图对乏脂肪肾血管平
滑肌脂肪瘤与肾透明细胞癌鉴别诊断的价值[J]. 中国医疗设备, 2023, 38(4): 96-101.
ZHOU Peng, WANG Dongqing, LIU Huli, ZHANG Lirong. Value of Preoperative CT-Based Nomogram in the Differential Diagnosis of Fat-Poor
Angiomyolipoma and Clear Cell Renal Cell Carcinoma. China Medical Devices, 2023, 38(4): 96-101.
