Supplementary Tables – Developmental artificial neural network model to evaluate the preoperative safe limit of future liver remnant volume for HCC combined with clinically significant portal hypertension

  

Supplement Table 1. Comparison of the incidence of grade B/C PHLF of patients between non-CSPH and mild-CSPH subgroup. 

  

Supplement Table 2. Comparison of the incidence of grade B/C PHLF of patients between non-CSPH and severe-CSPH subgroup. 

  

Supplement Table 3. Univariable analyses to identify predictors for grade B/C PHLF in the training cohort.

Supplement Table 4. Accuracy of the prediction of the ANN model for estimating the risk of grade B/C PHLF.

  

Supplement Table 5. Comparison of different models in predicting severe PHLF

  

Supplement Table 6. Comparison of different models in predicting grade B/C PHLF in defferent subgroups

  

Supplement Table 7. Comparison of the incidence and grade of PHLF among different risk groups in different subgroups