This research is based on real-life cases and explores the potential of telemedicine and computational intelligence methods (e.g., artificial neural networks and other machine learning algorithms, evolutionary computation, and hybrid algorithms) in improving mental healthcare service delivery in underserved regions with limited resources. Telemedicine, using electronic tools, can improve services, reduce unnecessary hospital visits, and assign doctors effectively, especially in areas lacking medical professionals. The study highlights the application of mathematical modeling as a decision-making tool for optimizing the timetable of expert doctors for patients using this technology. To achieve this goal a new multi-objective mathematical model is proposed. Given the intricate nature of this model, multi-objective meta-heuristic algorithms are employed to address its complexity and optimize solutions within a reasonable timeframe. Results show that efficient solution representation aids in achieving robust Pareto fronts. Furthermore, the research explores the use of machine learning in diagnosing patients’ conditions. Selecting relevant features to identify patients’ conditions is crucial to achieving the goal of triage processing and patient prioritization. It introduces a new feature selection method combining metaheuristic algorithms and ensemble classifiers to identify and select the most relevant features, consequently enhancing classification accuracy. Results demonstrate this approach is more efficient than conventional methods (e.g., filtering, wrapper, and embedded methods) for feature selection.
