New AI-powered test allows for quicker, more accurate cancer diagnoses

Research from the University of Waterloo uses AI for meaningful change.

University of Waterloo

Artificial intelligence is perfect for making everyday tasks simpler for individuals, but it’s also being used to fight for some semblance of a greater good. New research from the University of Waterloo reveals that AI could contribute to more accurate cancer diagnoses with quicker results.

The new AI-powered test matches digital tissue samples to a huge database of past cancerous samples. It’s a prime example of how innovation in artificial intelligence can be used for meaningful progress.

All about the archive — In order to detect cancerous cells, the test utilizes the largest publicly available archive of tissue samples in the world — more than 30,000 digitized slides from nearly 11,000 patients — courtesy of the National Cancer Institute.

The AI system searches the database visually to find those slides most similar to a new patient's sample. Those archive images, which otherwise would have been sitting gathering dust, are being repurposed to change the way we diagnose cancer.

Using the archive of images has allowed the test’s AI to achieve up to 100 percent accuracy in detecting 32 forms of cancer in 25 organs and body parts. The tool will continue to be refined in future research.

AI for all — Artificial intelligence is often viewed as a highly conceptual research area with minimal practical applications for the average person’s everyday life. Research like this goes a long way for proving the potential for real-life changes based on AI.

We can only hope funding continues for similar projects in the future. Now if we could just figure out a way to use AI to fight the spread of COVID-19, we’d be set.