Halifax Company Using A.I. To Better Screen Breast Cancer Patients
HALIFAX – A Halifax medical company, using the latest technological advances, is helping women all over the world in the battle against breast cancer. Using artificial intelligence, Densitas has processed two million mammograms worldwide, and their systems are being used in more medical centres as time goes by.
Densitas, which set up in Halifax in 2013, uses its A.I. to help breast cancer screening in three main areas. The first is assessing breast density as it appears on a mammogram. This is important because mammograms often fail to detect cancer in dense breasts. Another algorithm ensures better image quality of the mammogram itself. And the other technology determines the likelihood that a woman has breast cancer based on the mammogram results.
Artificial intelligence, in lay terms, uses a set of data to predict an outcome, without any biases that are inherent in human beings. According to Densitas CEO Mo Abdolell, even highly trained radiologists can often make mistakes when looking at a mammogram.
“One of the big advantages of artificial intelligence is that it allows you to develop these algorithms that can standardize and make a uniform ruler, in essence,” said Abdolell.
“The reality is, when you present a mammogram to several radiologists, you’ll get several, slightly differing perspectives. Sometimes there will be a lot of overlap in terms of their final conclusions, but there will often be disagreements. That’s due to the subjectivity involved in human perception.”
A computer system, unlike people, does not become exhausted after long work hours. Unlike humans, A.I. can keep processing mammogram data as long as it is needed.
“Everyone knows after 20 minutes radiologists start to tire; there’s visual fatigue after reviewing images and mammograms,” said Abdolell. “But A.I. doesn’t get tired; there’s no fatigue.”
It took Densitas four years to develop their first algorithm, but now their systems are being used all over the world, including Germany, the UK, Nova Scotia, Ontario, Chicago, New York, and Atlanta.
In order to get started from the ground up, Densitas needed a place that already had a wealth of data on breast cancer screening. That is one of the reasons the company chose Halifax as its headquarters. Nova Scotia has done a great job at preserving such data.
“It’s unusual to have that data all in one spot,” said Abdolell.
“When we talk about artificial intelligence, you got to think of one thing: data is the currency of A.I.”
Densitas’ systems are not meant to replace radiologists, but rather help them do their work better. One thing the algorithms can’t do, yet, actually detects cancer. So radiologists are still needed for diagnostics.
“What radiologists are really good at is detecting cancers. So, what you want to do is alleviate them from some of these other tasks,” said Abdolell.
“We want to make sure we’ve done as well as we can to present the best quality mammograms to the radiologists so they can do their job in diagnostics.”
There’s no doubt artificial intelligence will play a key role in health care in the future. Abdolell believes such technology will help alleviate stress on our overburdened healthcare system.
“A.I. really has the potential to, especially at a time like this, in the middle of Covid, where resources are scarce… to really automate and standardize and improve performance.”