Person-Centric Care for Alzheimer's Patients: A Multi-criteria Decision Support Framework for Activity Recognition
摘要
Alzheimer’s disease significantly impacts the daily lives of affected individuals by diminishing their ability to perform routine tasks and recognize familiar environments, resulting in loss of independence and increased caregiver burden. Human Activity Recognition (HAR) through Human-Computer Interaction (HCI) has emerged as a promising tool to address these challenges, enabling continuous monitoring and support for Alzheimer’s patients by identifying activity patterns and potential behavioral changes. This study proposes a multi-criteria decision-making methodology, combining the TOPSIS and ELECTRE methods, to refine and improve HAR rankings of various activities, assisting in the monitoring and assessment of Alzheimer’s patients’ daily routines. Our approach begins with feature normalization to ensure consistent data processing, followed by TOPSIS calculations to generate an initial ranking of activities based on their closeness to ideal activity patterns for Alzheimer’s care. The ELECTRE method further validates these rankings through a pairwise comparison of activities, refining results to account for detailed behavioral nuances. The closeness coefficient \({C}_{i}\) values generated through TOPSIS indicate that Activity A1, with a coefficient of 0.573, ranks highest as it closely aligns with the ideal routine for Alzheimer’s care, while Activities A2 and A3 rank lower with coefficients of 0.370 and 0.484, respectively. The ELECTRE refinement confirms this ranking, with Activity A1 demonstrating dominance in the aggregate matrix, ensuring reliability in the HAR system. The proposed methodology provides a systematic and accurate tool for HAR in Alzheimer’s care, offering caregivers and healthcare systems a dependable means to monitor patients’ activities and promote safer, more supportive environments tailored to individuals’ needs. Through this dual approach, our study significantly contributes to the development of HCI applications for Alzheimer’s care, paving the way for more effective, compassionate, and patient-centered healthcare solutions.