How AI Uses Music to Help Dementia and Parkinson’s Care

Rave Colony — Wed, 28 Jan 2026 08:00:00 +0000

Artificial intelligence (AI) is now at the center of a growing shift in neurological research.

Scientists and neuroscientists are using AI to reframe music as more than entertainment.

In this context, music is becoming a clinical tool designed to support neurological care.

They combine data, timing, and personalization.

The focus is on understanding each patient’s needs in real time.

From there, the system delivers the right music at the moment it can help most.

At the University of Melbourne, a new approach is taking shape.

Professor Felicity Baker is leading the Matchplus.ai project, bringing together neuroscience, music, and technology.

Recently, the initiative gained major momentum after securing $1.3 million in funding from Google.

With this support, her team is developing wearable sensors to predict agitation in people living with dementia.

The AI system works quietly in the background, tracking subtle changes in the body.

Over time, it learns what those signals mean.

As a result, it can spot early warning signs five to fifteen minutes before distress or disruptive behavior starts to surface.

Once triggered, it automatically plays carefully selected, personalized music to help calm the patient and prevent escalation.

Baker’s project stood out among more than 800 applicants competing for Google’s philanthropic funding.

Based on the team’s explanation, the process starts with wearable devices.

These tools collect constant streams of data to train predictive algorithms.

Over time, the system begins to learn. It recognizes patterns that often come before trouble starts.

From wandering and falls to sudden physical outbursts, the technology can spot these risks in nursing home settings before they happen.

As a result, this approach could significantly reduce dependence on psychotropic medications in aged care facilities.

Still, Baker points to a bigger challenge ahead.

The real work is teaching AI how to pick the right songs.

It also needs to know how to arrange them in the right sequence and adjust to different neurological conditions.

Meanwhile, similar innovation is unfolding across the Atlantic.

Across the Atlantic, a parallel effort is taking shape in Canada.

At the University of Montreal, neuroscientist Simone Dalla Bella helped co-found BeatHealth.

The startup focuses on developing music-based therapy specifically for people living with Parkinson’s disease.

One of its key tools, the BeatMove app, adjusts music tempo in real time to support movement.

The app gradually speeds up to encourage faster walking and then slows down as users become fatigued.

Right now, the idea is being tested in the real world.

In France, a clinical trial is underway to see whether patients can improve their walking patterns.

The study focuses on how they use the app during outdoor activities, including walks in public parks.

Dalla Bella has described the experience as moving alongside an invisible partner.

At times, the music takes the lead, gently pushing the listener forward when extra motivation is needed.

When the body begins to slow down, the sound eases back and follows, adapting to the moment.

Taken together, these projects point to a shift in how music is used in neurological care.

AI turns listening into an active, responsive experience rather than a passive one.

Through personalized soundscapes, the technology supports both patient safety and physical movement, hinting at a new model for therapy.

University of Melbourne • Rave Colony