Virtus Health has announced in partnership with Harrison-AI and Vitrolife that it will commence randomised controlled trials of its artificial intelligence (AI) technology, called Ivy, by the end of the year.
Speaking at The Future of Health event in Sydney this week, Virtus Health group CEO Sue Channon explained the tests will be used to further validate the use of AI when it comes to in-vitro fertilisation (IVF).
She explained how for the last 12 months, embryologists have been using Ivy as a supporting tool to increase the potential success of pregnancy through IVF.
"At this stage Ivy is still a supporting tool, we're not letting Ivy make the decision on its own," she said, explaining how one patient got pregnant during the cycle that Ivy was used after five unsuccessful IVF cycles.
"We are seeing an improvement of pregnancy outcome as a result of Ivy."
See also: How AI and next-generation genomic sequencing is helping cancer patients (TechRepublic)
But now, the clinic needs more data, after starting with some 10,000 data points.
"In order to achieve this [validation], we are gathering a much larger database of around 300,000 bits of data to improve the performance of Ivy," Channon said.
Eventually, Channon said it would lead to the development of Ivy 2, which will be integrated into EmbryoScope, an embryo time-lapse imaging technology developed by Vitrolife, to further improve the workflow of embryologists.
EmbryoScope currently features a high-resolution camera and computer software that automatically takes images of an embryo every seven minutes to carefully monitor the development of an embryo.
Data gathered from 10,280 embryo videos produced by EmbryScope, and data from various IVF clinics, including partners from the UK, Ireland, and Denmark, were used initially to train Ivy.
"Only a day after training Ivy, it exceeded human ability to predict which embryo was better than the other one, and by the fourth day it was able to achieve 94% predictive power in selecting a better embryo out of the set this was given," Harrison-AI co-founder Dimitry Tran said.
Traditionally, embryologists manually assessed each embryo based on an industry standard grading system in conjunction with imagery produced by the EmbryoScope to select the best embryo to transfer.
Read also: The role of data science and technology in mental health treatment (TechRepublic)
Channon however, said this methodology would result in too many variables.
"Ivy has taken away this very inconsistency in interpretation. Ivy is able to tell us which embryo is likely to result in a foetal heart meaning we can get patients pregnant sooner rather than later," she said.
From a workflow perspective, Channon said with EmbyroScope and Ivy, embryologists no longer need to manually monitor each embryo.
"This frees us the embryologist in monitoring the embryo in a planned and structured time frame. The embryo remains undisturbed and improves the development environment for the embryo, which is really important," she said.
Leveraging Google Cloud to identify associations between microorganisms and genes to predict different disease states.
Like anything, frameworks and boundaries need to be set -- and artificial intelligence should be no different.
CSIRO promotes ethical use of AI in Australia's future guidelines
For Australia to realise the benefits of artificial intelligence, CSIRO said it's important for citizens to have trust in how AI is being designed, developed, and used by business and government.
Singaporean medtech businesses are linking up with industry peers in Sichuan to enhance diagnostics and patient care, while local companies from other sectors look for opportunities in the Chinese province's transport and logistics sectors.