The intersection of AI and music therapy in dementia care

Dementia can affect the quality of communication and connection with family and those caring for them.  Severe symptoms such as agitation, distress and confusion can add to the strain. Yadzia’s MATCH technology harnesses AI-driven music therapy to detect agitation and distress, tailoring music interventions in real time to restore moments of calm, connection and dignity to those who need it most.

The need

Today, 55 million people worldwide live with dementia, a figure expected to reach 78 million by 2030. Families, caregivers and the aged care sector are struggling to cope, particularly with the agitation and distress symptoms that often accompany the disease. These symptoms often lead to increased hospitalisations, reliance on medications and a diminishing quality of life.

While music therapy has consistently proven effective in reducing agitation and improving emotional wellbeing, access to trained music therapists is limited.

“We know music therapy works, but there simply aren’t enough music therapists in the world to meet the growing demand. MATCH isn’t here to replace humans or music therapists, but to scale the impact of music so that more people living with dementia can benefit,” Professor Felicity Baker, Founder of Yadzia, said.

The research

As a clinician-turned-researcher, Professor Baker has spent decades demonstrating the therapeutic benefits of music for people living with dementia. Her clinical trials proved that music therapy reduces agitation, lowers anxiety and depression, and improves emotional wellbeing.

Yet one challenge remained – access. While music therapy works, it requires trained professionals, and as a result of the lack of them, widespread implementation is limited.

Determined to change this, Professor Baker assembled a team of experts to develop MATCH (Music Attuned Technology – Care via eHealth). The idea? Combine AI-driven music therapy with wearable sensors to predict and detect agitation and automatically deliver personalised music interventions in real time.

After conducting large-scale clinical trials across Australia, Norway, Germany, Poland and the UK, the team refined the MATCH caregiver training program to tailor interventions to individual patient needs.

Building on this, MATCH Plus integrates music attunement with AI to predict, detect and regulate agitation and distress via its bespoke wearable app. What once required a trained therapist could now be delivered through AI, making personalised music therapy accessible to millions.

Going to market

“I always knew MATCH should be commercialised, but I naively thought someone else would take it forward. Now I see that if I don’t drive it, it won’t go anywhere,” Professor Baker said.

Bringing a research-backed intervention to market required a shift in thinking. Used to academic rigor, Professor Baker and her team had to navigate financial forecasting, regulatory pathways and product development.

“The speed of commercialisation is in conflict with how we do things in research. In academia, we want everything to be perfect before launching, but in commercialisation, you need to focus on building a viable product that evolves over time,” Professor Baker said.

Yadzia plans to launch MATCH in three stages. First, the MATCH App will provide a caregiver training tool, helping families and aged care workers understand the emotional and behavioural needs of people with dementia while guiding them to implement personalised music interventions.

Next, MATCH Plus will introduce a wearable sensor that detects biometric indicators such as heart rate, movement and vocal distress, automatically triggering personalised music interventions based on real-time feedback. Finally, MATCH Pro aims to deliver a centralised dashboard for aged care providers, enabling staff to monitor multiple residents in real time, track distress levels and implement proactive care strategies.

With $2 million in funding from Google.org’s AI for the Global Goals initiative and a $156,000 grant from Aged Care Research and Industry Innovation Australia, MATCH has already demonstrated significant reductions in agitation in pilot programs.

The impact

MATCH dementia clinical trials have shown a 20–30 per cent reduction in sedative use, demonstrating that agitation can be effectively managed without over-reliance on pharmaceuticals. With fewer distress-related incidents, hospitalisations decline, and people with dementia can remain in familiar, supportive environments for longer. The emotional impact is equally profound. Families witness loved ones reconnecting with memories once thought lost, as a familiar song sparks a flicker of recognition, a smile or even a conversation. Caregivers also feel more confident and supported, using MATCH to intervene before distress escalates, easing care and enhancing caregiver resilience.

“Human connection is one of our most fundamental needs, yet dementia often disrupts it. MATCH is designed to restore those moments of connection, for the person with dementia and for the family who loves them,” Professor Baker said.