Predicting the risk of contracting the Hepatitis B Virus (HBV) is difficult because the disease progresses dynamically. Several variables that affect HBV-related consequences show sudden and slow changes over time, including the onset of cirrhosis, the removal of the Hepatitis C Virus (HCV), aging, portal hypertension, liver stiffness, and platelet count. “Recurrent Neural Networks with Long Short-Term Memory (RNN-LSTM) models,” which are excellent at learning temporal relationships, are used in this study to forecast the evolution of HBV risk to handle this complexity. The accuracy of risk forecasts is increased by using longitudinal patient data to record the temporal fluctuations of predictive parameters. To verify the robustness and dependability of the model, statistical analysis is used for validation. Deep learning and longitudinal modeling are combined in this method to create a strong framework for enhancing HBV risk classification, which offers insightful information for early intervention and individualized treatment of HBV patients.