A self-learning nurse call system.

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Human-Computer Interaction

Background:

• Increasing complexity in continuous care settings due to aging populations, caregiver shortages, and rising costs.
• Electronic healthcare (eHealth) solutions are introduced but add complexity, requiring caregiver integration and configuration.
• Existing eHealth services often fail to meet specific user needs due to difficulties in capturing environmental nuances during development.

Purpose of the Study:

• To extend an eHealth application with self-learning components for run-time parameter adjustment to user needs and preferences.
• To enhance the ontology-based Nurse Call System (oNCS) for personalized and context-aware caregiver assignment.
• To improve the adaptability and user-centricity of eHealth solutions in complex care environments.

Main Methods:

• Development of self-learning components that gather application usage data.
• Application of data mining techniques, including decision trees and Bayesian networks, to learn and adjust system parameters.
• Integration of learned parameters with associated reliability probabilities into the eHealth application after filtering unreliable values.

Main Results:

• The self-learning components successfully adjusted application parameters to user needs and preferences at run-time.
• The ontology-based Nurse Call System (oNCS) demonstrated efficient discovery of correct parameter values using decision trees and Bayesian networks.
• The system achieved high efficiency, requiring at most 100ms execution time and 20MB memory for a dataset of 1050 instances.

Conclusions:

• Self-learning components can significantly enhance eHealth applications by enabling automatic adaptation to diverse user requirements.
• The developed approach efficiently personalizes eHealth services, improving their suitability for complex healthcare settings.
• This methodology offers a scalable and resource-efficient solution for optimizing eHealth system performance and user satisfaction.