AI-Driven Detection of Abnormal Activities Among Elderly People Using Acoustic Feedback: Comparison of Key Algorithms

The global population is going through a significant demographic shift and the percentage of individuals aged 65 and over is projected to increase globally. This demographic transition brings forth challenges associated with aging, including an increased prevalence of non-communicable diseases and disabilities. Fall is one of the most common cause of serious injury for the elderly. There is a growing need for comprehensive care and support systems encompassing primary, acute, and end-of-life care, along with assistance in daily activities. Abnormal Activities of Daily Living (AADL) among elderly individuals present significant health risks and challenges. This study explores an innovative approach to detecting such incidents through the integration of artificial intelligence (AI). By analyzing acoustic distress signals, particularly screams, AI algorithms can promptly identify and respond to emergencies. The AI algorithms achieved an average success rate of 80.5% in identifying emergency situations. This method offers a proactive solution to enhance the safety and well-being of the elderly, potentially reducing the time to provide assistance and mitigating the consequences of falls and other abnormal ADL incidents. The implementation of AI in this context underscores its transformative potential in elder care, paving the way for smarter, more responsive health care technologies.