Edmonton-based health company Nanospeed Diagnostics is using AI to develop faster, cheaper diagnostic tests using just a single drop of blood

Rethinking blood testing

In medicine, a lot of time is spent waiting. Doctors' offices even have large rooms dedicated solely to doing just that. And few things are more stress-inducing to wait for than test results. That's a lesson that Dr. Rajan Gupta learned many times over during his career working in a hospital. His patients would sometimes wait days or weeks for the results of a blood test.

"You go to the doctor, you have the prescription for the test, then go to the lab. Then we have the blood drawn through the vein, which is very painful and then wait for the results," he said.“We saw this as a major problem that all patients go through during blood tests, and that gave us the idea that we need to do something… very different from the routine technologies. "

That idea led Rajan and his wife, Dr. Seema Gupta, to form NanoSpeed Diagnostics in 2009. The Edmonton-based startup works to take some of the pain out of blood tests. First, they developed a rapid test that can measure a patient's Vitamin D levels, using a drop of blood and within minutes.

Vitamin D is the key

Part of the reason NanoSpeed focused heavily on Vitamin D is due to how widespread the problem is. The body uses the vitamin to help it absorb calcium, used in bone growth. Vitamin D deficiency can be painful, leading to weak bones and fatigue. There's also evidence linking low Vitamin D to Type 1 diabetes and an increased risk of some cancers. Since we get most of our Vitamin D from sunlight, up to 90% of people in North America can be deficient during the winter months, Rajan says. Even warmer regions, like Africa and China, can see Vitamin D deficiency rates of around 50%.

Now, the pair are developing an AI-assisted method that might be the next generation of blood tests: faster, less expensive and able to do multiple assessments on a single drop of blood.

"It seems like the timing is there for artificial intelligence technology. It's becoming very important in medicine," Seema says.

She explains that current Vitamin D tests -- both traditional ones and NanoSpeed's current rapid tests - are based on "wet blood chemistry." The patient's sample is taken and mixed with certain compounds, which react with elements in the blood and reveal information on what it contains. The downside is that reaction changes the sample, meaning it is often useless for further testing. If you have more tests, you need to draw more blood.

A different approach

To meet this challenge, NanoSpeed is testing an AI-based approach. The blood would be analysed by a spectrometer and microscope, which reads how light interacts with the sample and uses that information to determine exactly what elements are in the blood. The result is a snapshot of the blood: everything that is in it, and how much is present. That info is then fed into specialized software trained to find indicators of low Vitamin D (or other conditions).

Nanospeed's Dr. Seema Gupta tests samples in their Edmonton-based labAside from being faster and less expensive, the new method would have a more forgiving time limit: wet chemistry tests generally have to be done quickly after taking the blood, while the AI-based approach could work on blood droplets that have been stored for years on microscope slides. All that would make blood tests easier and more accessible to patients.

"With the artificial intelligence-based test, it will be much, much easier for people to make informed decisions about their health. That's the important thing," she says. By making testing less expensive and less time-consuming, she says vitamin deficiencies can be caught sooner when they are easier to treat and cause less pain.

Another advantage of using machine learning is that the approach may be easily transferred to other types of testing. While Nanospeed is currently training its software to measure Vitamin D, different algorithms can be taught to look for other compounds in the blood. The company already has tests for other vitamins and minerals that could use the same AI analysis process they're developing for their Vitamin D test. They also hope to use the method to measure the amount of THC in a blood sample, which could lead to tests for impaired driving.

This project was part of the Western Economic Diversification Canada Regional Innovation Ecosystems (RIE) program. The initiative brought together nine organizations from non-profit, business and academia to establish viable uses for artificial intelligence and machine learning in health and data analytics.