Listening to Healing: Music Listening Patterns, Speech Biomarkers and Reflexive AI for Depression and Recovery

Young people today are surrounded by sound. Many of them listen to music for hours each day to manage emotions, focus, or simply feel less alone. At the same time, rates of reported depression and social isolation have increased in recent years. This study explores how daily music listening habits and speech patterns might reveal subtle changes in mood and mental health. The project uses the NaHzHaR framework, which stands for Notice, Hold, Heal, and Release. It is a human-centered way of building AI systems that observe, reflect, and support without judging or replacing human care. A six-week study is planned with young adults who will share their playlists and record short speech samples before and after an ethical, therapist-guided playlist intervention. Features like pitch, rhythm, and repetition in both music and speech will be compared with standard mood scales such as the PHQ-9. The goal is to understand how listening and speaking together can show patterns of emotional healing. The study brings together psychiatry, AI, and arts-based therapy, linking the rhythm of neural recovery seen in neuromodulation with the rhythm of everyday life.