CT-Based Radiomics Nomogram: A Potential Tool for Differentiating Hepatocellular Adenoma From Hepatocellular Carcinoma in the Noncirrhotic Liver.

摘要: Rationale and Objectives To evaluate the value of a radiomics nomogram for preoperative differentiating hepatocellular adenoma (HCA) from carcinoma (HCC) in noncirrhotic liver. Materials Methods One hundred thirty-one patients with HCA (n = 46) HCC (n = 85) were divided into training set (n = 93) test (n = 38). Clinical data CT findings analyzed. Radiomics features extracted triphasic contrast images. A signature was constructed least absolute shrinkage selection operator algorithm score calculated. Combined independent clinical factors, developed by multivariate logistic regression analysis. The performance assessed calibration, discrimination usefulness. Results Gender, age, enhancement pattern factors. Three thousand seven sixty-eight reduced to 7 as optimal discriminators build signature. (area under curve [AUC], 0.96; 95% confidence interval [CI], 0.93–0.99) factors model (AUC, 0.93; 95%CI, 0.88–0.99) showed better capability (p = 0.001 0.047) than 0.83; 0.74–0.92) set. In set, 0.94; 0.87–1.00) performed (p = 0.013) 0.75; 0.59–0.91). Decision analysis outperformed terms Conclusion CT-based has potential accurately differentiate