Objective: Sepsis is a potentially fatal systemic response to infection, in which early clinical intervention is critical to reduce mortality. This study presents a hybrid deep learning model that combines temporal and structural information from clinical data to improve early sepsis prediction.

Methods: We used data from the PhysioNet/Computing in Cardiology Challenge 2019 to predict sepsis onset up to 12 hours in advance. We developed a hybrid model integrating Long Short-Term Memory (LSTM) networksand Graph Attention Networks (GAT) to capture temporaldynamics and inter-variable relationships. Performance was compared with three baseline models. To ensure robustness, all models were trainedusing five repeatedtrain-test splits with different random seeds.

Results: The dataset includes40,336 adult ICU patients. Of all the patients, 2,932 developed sepsis during their stay. Each patient’s data includes hourly data on 40 clinical variables,including vital signs, laboratory results, and demographic information. The LSTM-GATmodel achieved an AUROC of 0.853 ± 0.005, F1-score of 0.627 ± 0.006, and specificity of 0.872 ± 0.007, outperforming baselinemodels. Despite being trained on fixed temporal windows, the model generalized well across multiple prediction horizons without retraining.

Discussion: By integrating temporal and structural representations, the proposed approach achieves improved predictive performance compared with baseline. This capability may support earlier identification of high-risk patients and enhance timely clinical decision-making in critical care environments.

Conclusions: The proposed model demonstrates the advantage of combining sequence and graph-based methods. It offers a promising tool for real-time clinical decision support in sepsis detection.