Portal vein embolization (PVE) has been a standard procedure used preoperatively to induce growth of the future liver remnant (FLR). The goal with PVE is to ensure that a sufficient liver volume remains after resection to maintain function in patients with liver malignancies intended for extensive resections. However, predicting which patients will achieve an adequate FLR growth after PVE remains challenging. Current prediction methods are based on clinical factors such as baseline liver function, underlying liver conditions (for instance cirrhosis), and other patient factors. Despite their utility, these predictors often lack the precision required for effective individualized treatment planning. Previous studies have highlighted the need for prediction models that are accurate and accessible, without requiring invasive procedures or costly imaging alternatives.
1 Our study aimed to bridge this gap by building a model that integrates clinical data with radiomic features derived from routine CT images to accurately predict insufficient liver hypertrophy following PVE.
2 …