AI stroke imaging study aims to improve treatment times

NEWS - eHealthNews.nz editor Rebecca McBeth 

New Zealand researchers are launching a study to assess the impact of artificial intelligence (AI) stroke imaging technology, aiming to reduce treatment delays and improve intervention rates for stroke patients. 

The project, funded by the Health Research Council, will evaluate whether AI based imaging tools can speed up the stroke assessment workflow and provide better access to treatment. 

Consultant neurologist at Wellington Hospital Anna Ranta is leading the research and says that delays in stroke treatment occur most often when patients present at smaller hospitals. 

"Currently, if a patient comes into smaller hospital, an ED doctor needs to decide, is it possibly a stroke, is it not a stroke? Then they need to make imaging decisions, then the imaging gets interpreted, then somebody is called from a tertiary centre to discuss the films to make a treatment decision," Ranta says. 

The project will trial imaging technology to provide a rapid initial assessment to trigger specialist calls earlier. Similar tools are in wide clinical use, including in New Zealand, but most smaller hospitals are still missing out. 

“Instead of waiting for, often overseas, radiologists who might take an hour to do the initial interpretation, and then calling the tertiary specialist to make a treatment decision, this will give the doctor in ED the confidence to call a specialist much faster," she explains. 

Ranta, who is also medical director of the Stroke Aotearoa, says treatment delays affect a patient’s eligibility for intervention, as brain tissue becomes less salvageable over time. 

"The longer you go, the less brain recovery potential brain there is, so some people do not just have less effective treatment, they actually become ineligible for treatment," she says. 

In this case, AI relates to the use of machine learning during the development of the computational algorithm used for image display and decision support. The tool provides an initial preliminary read that triggers an expert referral, but all final interpretations and treatment decisions will be made by NZ clinical experts.  

“These tools do not incorporate new data to continuously learn and no New Zealand patient data will be used for this purpose as part of this project or planned future application,” she explains.  

A similar tool introduced by the NHS has been shown to reduce the time between someone presenting at a hospital with a stroke and receiving treatment by more than 60 minutes – potentially tripling the likelihood of recovery without a disability. 

Ranta says the potential cost savings for New Zealand’s health system are significant and estimated to be on the order of $20-40 million per annum when considering both short-term healthcare and long-term societal costs. 

The research will compare door-to-treatment times and intervention rates using existing data collection systems, including the national stroke register where all patients receiving reperfusion treatment are recorded. 

The study aims to achieve regional network-level rollout of the imaging tool for stroke, allowing neurologists and stroke physicians on call to use a single tool across multiple hospitals. 

"The idea of having clinicians use the same tool across an entire network or even the country is also really exciting because we will all talk all the same language and communication will be improved further speeding up treatment decisions," Ranta says. 

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