‘Digital twin’ research features at SI launch of AI network

NEWS - eHealthNews.nz editor Rebecca McBeth 

A project to create a ‘digital twin’ of a diabetes patient was amongst research featured at the South Island launch of the AI in Health Research Network.

The national network aims to improve artificial intelligence (AI) research in healthcare and enable testing of AI tools using New Zealand data so they are relevant for the local population.

Arindam Basu from Canterbury University spoke at the event about his research which uses AI to create digital versions of patients. 

These "digital twins" will be used to guide real-world behaviour changes for people with type 2 diabetes, which affects four to five percent of New Zealanders.

"We are creating a digital copy of a person on a server," Basu told the network event. 

"We will then make changes to this copy to help the actual person adjust their habits and get back on track."

The system will use data from everyday devices like smartwatches to give simple personalised advice, with the ultimate goal being to effectively reverse their diabetes.

Basu said his team has already developed an algorithm that can analyse individual patient data and create tailored messages and training materials and they are planning to apply for a randomised controlled trial in a community setting.

At the launch on September 4, network member Rosie Dobson from Auckland University’s school of population health, said it is important not to rush implementation of AI tools without clinical trials and validation using New Zealand data.

She told attendees that it is concerning to see how rapidly AI is being adopted in some cases without proper clinical testing.

Dobson, who also sits on the Health New Zealand | Te Whatu Ora National AI and Algorithm Expert Advisory Group, said that it would typically take 17 years for new medical developments to move from idea into clinical practice, but that AI is being treated quite differently.

"We are seeing things go from idea to clinical practice in months and we really want to change that," Dobson said.

The Research Network will help by bringing together experts from across New Zealand to ensure rigorous research with real-world applications.

"It is really important that AI research is done here in Aotearoa and AI products are validated and tested on our data to ensure that they are appropriate and safe for our people," she said.

Recent government funding rounds have targeted AI and health research, such as the Catalyst Fund run by the Ministry of Business, Innovation and Employment (MBIE) and Health Research Council AI in Health funding of $4.6 million.

MBIE is co-funding, with Singapore’s National Research Foundation, three projects which apply AI to improve health outcomes for older adults.

Hamish Jamieson, a geriatrician with Health New Zealand, spoke at the launch about his research which is looking to transform the approach of interRAI assessments for older people.

"We are doing hundreds of thousands of these assessments, and each one takes about two hours. That is a lot of manual work generating information from these systems," he told the Christchurch audience.

The research will focus on three key areas where AI could make a significant impact; extracting information from medical notes; developing predictive analytics for health outcomes; and creating personalised care plans for patients and their whānau.

The team plans to analyse a massive dataset, including 1.3 million existing assessments, to help create predictive models for health issues in older adults, such as fracture risk, cognitive decline, and depression.

"We are aiming to use existing health data to build these new predictive models. It is about making the most of what we already have,” Jamieson explained.

"We have deliberately structured our research team to include both health professionals and AI experts. We are not just doing research, we are aiming to create real change in how we care for our kaumātua."

Laurie McLay from the University of Canterbury spoke about her research into creating digital support systems for neurodivergent people, with a focus on autistic children and their whānau, who face long wait times to get the help they need.

Her work focuses on collaboratively designing, building, and testing digital tools that improve the wellbeing of neurodivergent people and their support networks.

"We are aiming to work alongside autistic people, Māori, and Pacific advisory groups to develop a digital programme that will make support more accessible," McLay explained.

The AI in Health Research Network is holding a workshop in Christchurch this November 27, as part of Digital Health Week 2025, which will feature researchers who have been funded in these rounds.

Read more about the workshop and register to attend today.

To comment on or discuss this news story, go to the eHealthNews category on the HiNZ eHealth Forum

You’ve read this article for free, but good journalism takes time and resource to produce. Please consider supporting eHealthNews by becoming a member of HiNZ, for just $17 a month

Read more AI & Analytics news

Return to eHealthNews.nz home page