The Pro Medicus Limited (ASX: PME) share price is dropping lower on Tuesday despite the release of a positive announcement this morning.
At the time of writing, the leading health imaging company's shares are down 1% to $42.39.
What did Pro Medicus announce?
This morning the company announced that it has received clearance by the U.S. Food and Drug Administration (FDA) for its Breast Density artificial intelligence (AI) algorithm.
This means the company now has clearance in the US, Europe, and Australia markets, which paves the way for management to begin marketing the algorithm across all three jurisdictions.
What does this algorithm do?
According to the release, the Breast Density algorithm is intended for use with compatible full field digital and digital breast tomosynthesis systems.
It assesses breast density from a mammography study and provides an ACR BI-RADS Atlas 5th Edition breast density category to aid radiologists in the assessment of breast tissue composition.
In December 2019, CEO Dr Sam Hupert spoke about the algorithm.
He explained: "Working with the breast imaging team at Yale, which is one of the most highly regarded in the US, we were able, using Visage Research Server, to anonymise over 30,000 cases. This data was then fed into our machine learning model which was used to produce the Breast Density algorithm."
"As an extra validation, we had a team of five breast imaging specialists at Yale each curate over 500 cases using our curation tools and then tested the algorithm against these curated cases. The results were so promising we decided to commercialise the product and have applied for FDA approval," he added.
Dr Hupert also pointed out that the product differs from its competition due to its software being GPU based.
He commented: "Because Visage is GPU based, we are able to run our algorithm on our client's existing hardware which takes around a second to provide the analysis, so it is pretty much on demand. This is unlike others that either require a fair amount of additional onsite hardware to run their algorithm or they have to do it in the cloud [..] This can be costly and because the files are so large, time consuming."
What now?
Dr Hupert believes today's approval could pave the way for similar developments in the future.
"We developed this algorithm using our own AI Accelerator platform which enabled us to significantly speed up every stage of the process from concept to FDA approval. This not only provides us with a fast-track mechanism to develop our own algorithms in future, it paves the way for further collaborations with the growing number of our research oriented clients," he concluded.
