Akademik Veri Yönetim Sistemi
5th International Conference on Informatics and Software Engineering, IISEC 2026, Ankara, Türkiye, 5 - 06 Şubat 2026, ss.267-271, (Tam Metin Bildiri)
This study proposes an intelligent triage system for use in hospital emergency services for safe and consistent patient triage. The system comprises two main components. The first component is a structured text-based intake system that gathers required triage information. It has the desired features of using plain language, validating measurements, managing missing or non-applicable data, supporting multilingual use, and exporting standardized records with timestamps and audit trails. The second component applies a CatBoost classifier to triage data with minimal preprocessing. After resolving data quality issues and fine-tuning learning parameters, the model achieved 85% balanced accuracy. For the critical ESI-1 class, which the patients with the severest conditions belong to, it reached a precision value of 0.995, recall of 0.928, and F1 of 0.960, demonstrating high reliability in identifying critical cases. The modular architecture supports standardized validation, feature mapping, classification, audit logging, and deployment to ONNX and Core ML for low-latency inference. Continuous monitoring, retraining, and prospective evaluation in multilingual real-world settings will guide future development, including improved fairness, usability, and data collection specifically for triage applications.