The study developed a novel DELU model based on dual elastography to screen for high-risk varices (HRV) in patients with hepatitis B virus (HBV) cirrhosis. The model includes six variables: platelet count (PLT), alanine aminotransferase (ALT), ascites, portal vein thrombosis (PVT), inverse differential moment (IDM), and liver stiffness (liver Vs). AUROC reached 0.822 in the training cohort and 0.779 in the validation cohort, outperforming both RESIST (0.600 and 0.626) and Baveno VI criteria (0.569 and 0.575). DELU saved the most endoscopies (20.7% in training, 11.1% in validation) and maintained performance in subgroups such as Child-Pugh B/C (AUROC 80.3%), Child-Pugh A (79.7%), in antiviral therapy (82.5%) or virologically suppressed patients (84.6%). Performance was not affected by hepatocellular carcinoma, gender, or BMI. The model is a reliable non-invasive tool to exclude HRV, RESIST serves as an alternative without elastography. The study was registered under NCT04640350.