A $4 million research project led by the University of Auckland is developing a new AI tool to predict who might develop dementia, aiming to pick up early warning signs before symptoms set in.

Dr Catherine Morgan and Professor Lynette Tippett, alongside experts from Singapore and Otago, are creating a tool that could help doctors more accurately identify people at high risk of dementia using memory tests, blood results, genetics, MRI scans, and more.

Backed by the Ministry of Business, Innovation and Employment, the project combines years of local health data, including findings from New Zealand’s Dementia Prevention Research Clinics, to train the AI system.

The team’s goal?

Over 85% accuracy in predicting dementia risk.

Why does this matter to locals?

With 70,000 Kiwis currently living with dementia—and that number expected to double by 2050—this research could help Hibiscus Coast families get earlier answers, plan for the future, and potentially slow the disease’s progression.

“It’s about finding those red flags early,” said Morgan.

“Then we might be able to delay or even prevent it.”

The tool will produce an explainable risk score, showing doctors exactly which factors contributed to the result.

That could lead to earlier referrals, specialist advice, and more proactive care.

The project, which also includes Siemens Healthcare and Te Whatu Ora clinicians, will run for three years and finish with a pilot trial of the new tool.

Know something local worth sharing?

Send it to [email protected] — we’ll help spread the word.