Abu Dhabi Building AI Algorithms that Are Transforming Healthcare Tech
January 7, 2022
56         0

by Stephen Kanyi

I’ve seen it in virtually all futuristic movie shows and movies, an AI doctor (usually in the form of a human hologram) offers medical consultation and/or diagnosis. UAE’s Mohamed bin Zayed University of Artificial Intelligence wants to make this a reality.

Healthcare is difficult and complex. Even with the technology we possess and the knowledge man has accrued over the centuries, our current medical professionals routinely fall short of individual medical needs. Late diagnosis and even misdiagnosis are still a common occurrence in hospitals all over the world, even in developed nations. This can cost lives.

A research team from the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) is however working to change this. Led by Dr Mohammad Yaqub, an assistant professor at the school, the team aims to use machine learning to make “diagnostics more immediate, more accurate, and more widely available.”

Talking to Wired Dr Mohammad described a story that inspired him to pursue this task:

A friend of mine—we studied together at university—developed multiple sclerosis. He had the symptoms, he went to many doctors, but it took two years for him to be diagnosed with it. It’s like House, where you see these doctors talking, trying to solve the problem. They can’t figure out what’s going on. So now we are working on a set of algorithms to come up with diagnoses in real-time, which can tell you whether MS is there, and if so will highlight the region in the brain where the damage is. I don’t want these kinds of delays.”

He goes on to narrate another story of a friend from his alma mater, the University of Oxford who also contributed to his career path.

He was captain of our football team, very strong, very fit, but he died,” Yaqub says. “I found out later he had undiagnosed congenital heart disease. There are many, many children born every day with this condition. I was already working on AI in the medical field at the time, but this encouraged me to think that what I am doing is worthwhile, something that can save lives.

It is this story and countless others that inspired Yaqub to develop ScaNav, a fetal anomaly assessment system based on AI that is already exclusively licensed by industry giant GE Healthcare who use in its Voluson SWIFT ultrasound machines.

“What happens is, after 20 weeks of gestation, the woman goes for an anomaly scan, which checks for things like heart conditions, spina bifida, and brain diseases. The doctor is supposed to look at the scan, tick some boxes, and say whether everything is fine.”

This is how it is supposed to work, but it is far from perfect.

Globally, 2 to 3 million babies are born with congenital defects each year—around a million of these with a heart condition. And the shocking thing is, even in some developed countries, in approximately half of these cases the doctors aren’t aware of a condition until after the birth. This is a problem, because the earlier you can tell if there is something wrong, the more effectively you can intervene,” he explains.

Yaqub’s machine-learning algorithms help by verifying the normality in the fetus and also alerting doctors if the standards are not met.

“It was trained using a million images, covering everything from the brain to toes, which were analysed by a team of doctors,” he says. “Then we built algorithms that mimic how doctors analyse these images. Imagine, bringing all of these brains together into one objective decision-making system.”

He predicts that around 10 million women will benefit from his system each year. He however says that this is only the beginning.

My team and I are working to develop AI algorithms for these devices, which can hopefully be put to use anywhere, including a refugee camp or war zone.

subscribe for YouMedia Newsletter
0 comments

Leave a Reply

Your email address will not be published. Required fields are marked *

newsletter
subscribe for YouMedia Newsletter
LET'S HANG OUT ON SOCIAL