Artificial intelligence is not going to replace doctors any time soon. However, it can become a very useful tool for diagnostics and choosing the right treatment. A team of engineers, led by the University of Waterloo, have developed a new powerful technology combines human wisdom and artificial intelligence for faster, better cancer diagnosis.
AI systems could compare images of tissues with huge databases with confirmed cancer cases. Image credit: mikemacmarketing via Wikimedia (CC BY 2.0)
Cancers have something in common – the outcome is always better if the disease is caught in its early stages. Diagnosis is sometimes difficult, however, even though new techniques are being developed all the time. Artificial intelligence (AI) could be one of these solutions that would streamline the entire process.
For example, this new software is able to compare images of suspected cancer cases with previously diagnosed cases in a database. And it does so with incredible accuracy – in tests using archives of about 30,000 digitized slides from almost 11,000 patients this software showed up to 100 % accuracy, despite having to deal with 32 forms of cancer in 25 organs and body parts. It literally works with virtually everything from melanoma to prostate cancer.
Simply put, this software looks through the database, created from confirmed cancer cases, and finds images that look the most similar. In other words, AI taps into human wisdom, gathered over the years to find the most similar cases that could lead to similar diagnosis. It is something that no humans could do. This AI technology almost absorbs knowledge of thousands of doctors, who dealt with thousands of cancer cases. This technology could significantly speed up cancer diagnosis and accuracy. Of course, some refinement is still needed before it can be available for clinical use.
This AI technology is easily combined with today’s communications. It would not have to be used in every hospital to benefit a large number of people. For example, doctors in less developed countries could take images and send them to bigger medicine centres in other parts of the world. Hamid Tizhoosh, one of the scientists behind this project, said: “This technology could be a blessing in places where there simply aren’t enough specialists. One could just send an image attached to an email and get a report back”. It will be months or even years till this technology is properly developed and tested, but soon enough it will be put to service to save people’s lives.
And that is just one AI system, developed at one university. Scientists around the world are working on this, achieving some impressive results. Hopefully, it won’t be too long till AI tools will be available to most hospitals for quick and reliable diagnosis.
Source: University of Waterloo