Echo IQ trial shows EchoSolv screening platform improves diagnosis of severe aortic stenosis
Echo IQ’s clinical trial found women were 66% less likely to have been accurately diagnosed than men, but EchoSolv was effective in identifying disease without gender bias.
A flagship clinical trial of Echo IQ’s (ASX: EIQ) EchoSolv artificial intelligence (AI) medical software technology has identified 72% more patients with severe aortic stenosis than human diagnosis alone.
The trial was designed to test the effectiveness of the screening platform in identifying patients at risk of dying from the condition, which is reported to be one of the most common forms of heart valve disease in older people and is associated with high risks to life if not diagnosed and treated quickly.
It was fully funded by US medical technology company Edwards Lifesciences and was conducted in conjunction with the National Echo Database of Australia (NEDA) at St Vincent’s Hospital sites in Sydney and Melbourne.
It complements an earlier study completed at Harvard Medical School’s Beth Israel Deaconess Medical Center in Boston which found EchoSolv to be effective in identifying patients with guideline-defined aortic stenosis as well as those exhibiting risk-of-death indicators.
Patient echocardiograms
The latest trial applied EchoSolv to 9,189 patient echocardiograms from both St Vincent’s sites.
The echocardiograms showed 218 individuals had been diagnosed by humans as per routine clinical care with guideline-defined severe aortic stenosis.
However when EchoSolv independently reviewed the same population, it identified 376 individuals with the same level of disease.
This represented an increase of 72% in disease detection.
Gender difference
The study also revealed a difference in the way women had been identified and earmarked for treatment, with women 66% less likely to have been accurately diagnosed than men using human-only assessment.
In contrast, EchoSolv did not discriminate in its identification of patients by gender.
Other results found that even when identified as having severe aortic stenosis, women were 50% less likely to receive intervention than men, which reinforced the need for unbiased identification of disease.
The software also showed 94% of the study population to be at low risk.
Echo IQ said being able to categorise patients in this way has the potential to support healthcare facilities to better allocate resources and prioritise patients for additional review or treatment consideration.
Enhancing human diagnosis
St Vincent’s director of heart lung program and cardiology Professor Michael Feneley said the clinical trial demonstrated the effectiveness of novel technologies such as EchoSolv in enhancing human diagnosis.
“With the general ageing of the population leading to an increasing prevalence of aortic stenosis, it is encouraging to see how artificial intelligence could be used to improve the identification of disease and increase the opportunities to treat patients in a timely manner,” he said.
Improved detection
Echo IQ chief medical officer Professor David Playford said EchoSolv was the first decision-support platform in the world to show improved detection of severe aortic stenosis compared with current clinical practice.
“We expect the automatic highlighting of patients with significant aortic stenosis risk using EchoSolv will assist doctors in decision-making for aortic valve intervention and follow-up in a consistent, systematic and efficient manner,” he said.
“Our goal is to support improved diagnosis free of unconscious bias and irrespective of age, gender, background or socioeconomic status … these findings are a significant step towards our goal of assisting doctors in finding the right patients, every time, for the right intervention for heart valve disease.”