Fig. 3From: Preoperative clinical-radiomics nomogram for microvascular invasion prediction in hepatocellular carcinoma using \(^{18}\)F-FDG PET/CTFeatures selection used the LASSO regression model in the features group. A LASSO coefficients produced by the regression analysis. As shown in B, \(\uplambda\) = 0.0240 with log (\(\uplambda\)) = − 3.728 was chosen according to 10-fold cross-validation, where optimal \(\uplambda\) resulted in 45 non-zero coefficients in the features groupBack to article page