Abstract. Age-related memory impairment often causes significant distress for the elderly, imposing substantial burdens on families and society. It is widely known that auditory stimulation can trigger certain emotions and brain activities, but studies about factors that determine its effectiveness are lacking. This study aims to investigate the potential of music as an intervention for age-related memory impairment and the relative importance of music genre and preference level as factors that impact the effectiveness of music treatment. The research involves 22 participants and evaluates verbal episodic and working memory performance using the ten-word test (TWT) and digit span (DS) test, respectively. Additionally, electroencephalogram (EEG) data is analyzed to examine the impact of music on cortical electrical activity. This study involves the creation of a random forest machine learning model to predict long-term memory outcomes based on short-term data. Based on results comparing preference to music genre, we found that preference is a more important determining factor for the effectiveness of music treatment, and preference level one (most favourite music) treatment yielded the highest average improvements in both the TWT and the DS. Based on genre analysis, classical music stimuli were effective for improving long-term TWT and DS scores, and folk music stimuli effectively improved long-term TWT scores. Following folk music treatment, participants aged 65 and above yielded TWT score enhancements ranging from 25% to 66.7% and DS score enhancements ranging from 20% to 50%. This study also discovered the effectiveness of utilizing the random forest machine learning model to predict long-term TWT and DS outcomes based on short-term results, potentially enhancing the efficiency of future treatment.
Read full abstract