Become a Client

Artificial Intelligence: Cancer’s Biggest Threat

Dyllan Furness

Dyllan explores technology and the human condition for Tech Emergence. His interests include but are not limited to whiskey, kimchi, and Catahoulas.

Artificial Intelligence: Cancer's Biggest Threat

Cancer cells have found new foes in AI and the Nordic countries. In just the passed month, both Sweden’s Uppsala University (UU) and the Technical University of Denmark (DTU) have announced systems that accelerate cancer treatment research and diagnoses. By joining genetics with computer science researchers at the DTU have developed a self-learning computer program that can successfully locate the disease 85 percent of the time. Meanwhile UU’s “new smart research robot” assists doctors in finding optimal cancer treatment combinations for their patients.

In most cases, doctors can currently diagnose cancers and simultaneously locate where in the patient’s body the disease exists. So, for example, a doctor can typically detect cancer cells and determine whether it’s cancer of the bran, liver, or lungs. However, in about 5 percent of cancer diagnoses, doctors detect cancer cells but cannot identify from where the cells derive. In these cases, patients are referred elsewhere – to any number of specialists – and subjected to numerous diagnostic tests before the exact location of the cancer can be discerned; a process that delays treatment and induces stress.

Worse still, in some instances, even after consulting specialists the patient’s cancer can’t be located. According to DTU’s announcements, these “patient[s] will be treated with a cocktail of chemotherapy instead of a more appropriately targeted treatment,” which may or may not even attack the proper cancer cells. But with 85 percent accuracy, DTU’s new program looks like it will help diminish this uncertainty and subsequent trauma.

Alliteratively named TumorTracer, the system works by analyzing the DNA of a biopsy from a metastasis (when cancer cells spread from one organ to another). Advanced algorithms then compare this cancer’s DNA to the DNA of other cancers to determine potential areas of localization. As self-learning software, the system uses its successes and failures to refine its analyses, meaning the more it tests, the better it gets. Associate Professor Aron Eklund who led the study says, At the moment, it takes researchers two days to obtain a biopsy result, but we expect this time to be reduced as it becomes possible to do the sequencing increasingly faster.”

Meanwhile in Uppsala, Dr Mats Gustafsson and his team of researchers have built a smart robot that cuts down on the leg work required to identify drug combinations with the highest efficacy in cancer treatment. Usually, doctors pair together individually affective drugs to fight a particular disease. Though these drugs are proven to work by themselves, the best combinations are ultimately unknown until tested.

UU’s robot system plans and conducts experiments with a myriad of drugs. It then uses these results to draw its own conclusions. The system is notable in it’s ability to test dozens of combinations at once, with a goal to eventually handle hundreds of combinations simultaneously. The team’s research was published in Scientific Reports.


Flow charts and graph depicting the process of their smart robots experimentation, analysis, and conclusion.

“We have built a robot system that plans and conducts experiments with many substances, and draws its own conclusions from the results. The idea is to gradually refine combinations of substances so that they kill cancer cells without harming healthy cells”, says Dr Claes Andersson, another of the study’s leading researchers, in a press release. To be sure, UU’s system is not the first of its kind. However, according to Dr Gustafsson, it is the first to take into account both efficacy in killing cancer cells and mitigation of side effects.

Neither the researchers at DTU nor at UU are content with the systems they’ve developed. In Denmark, Eklund and his team hope to eventually engineer TumorTracer to locate the source of free cancer cells from a blood sample, information that would help monitor patients at risk of developing cancer. At UU, Gustafsson, Andersson, et all want to fully automate their smart robot, which currently still requires a few manual inputs to begin it’s analysis. They also hope to help patients with returning cancer cases – as these patients often become drug resistant and require new, unique combinations.

With the aid of AI, the Nordic countries are making valiant efforts to fight cancer. As researchers develop these systems further – and the systems likewise further develop themselves – AI might indeed prove to be cancer’s biggest threat.

Photo credit: M. Kashif et al./Scientific Reports

Stay Ahead of the AI Curve

Discover the critical AI trends and applications that separate winners from losers in the future of business.

Sign up for the ‘AI Advantage’ newsletter:

Stay Ahead of the Machine Learning Curve

At Emerj, we have the largest audience of AI-focused business readers online - join other industry leaders and receive our latest AI research, trends analysis, and interviews sent to your inbox weekly.

Thanks for subscribing to the Emerj "AI Advantage" newsletter, check your email inbox for confirmation.