AI-driven radiotherapy appointment optimization: A lean management approach

Purpose: Radiotherapy is integral to cancer treatment, but inefficiencies in appointment scheduling often lead to delays, suboptimal resource utilization, and compromised care quality. Leveraging Lean Management methodologies, we identified bottlenecks and inefficiencies in the scheduling process, emphasizing the need for advanced optimization strategies. This study developed an AIdriven system to automate and optimize radiotherapy appointment scheduling, integrating patient-specific and departmental constraints. Materials and Methods: Lean Management tools, including Value Stream Mapping (VSM) and 5S analysis, identified inefficiencies in manual scheduling, such as resource misallocation and delays in urgent cases. An AI-based solution was designed with supervised machine learning algorithms to classify patient urgency and optimize schedules dynamically. The AI system includes data extraction, feature engineering, and real-time optimization modules, ensuring adaptability to workflow variations and emergency cases. Results: Machine learning models were evaluated for urgency classification, with Random Forest achieving the best performance (AUC: 0.81). The system categorized patients into three priority levels (37.1% Group 1, 35.2% Group 2, 27.6% Group 3) and optimized resource utilization by minimizing scheduling conflicts and delays. The automated process improved workflow consistency, reduced physician workload, and ensured timely treatment for high-priority patients. Conclusion: AI-driven optimization, combined with Lean Management insights, offers a transformative solution to radiotherapy scheduling challenges, improving care efficiency and resource management in high-demand clinical settings