By Prof Philip Poortmans, ECCO President 2018-2019
There was nothing artificial about the intelligence of the late Professor Stephen Hawking. He lifted our horizons on all manner of topics, not only mathematics and cosmology. When considering the future of computer technology, he warned: “Computers will overtake humans with AI within the next 100 years... When that happens, we need to make sure the computers have goals aligned with ours.”
Artificial intelligence in cancer care is now well past the realm of science fiction. It is here already. Some examples include:
- The use of AI to evaluate whether an X-ray is normal, allowing radiologists to focus their time more effectively on the analysis of abnormal images;
- Machine learning techniques to improve the identification of DNA mutations within cancers and even to forecast future genetic changes; and,
- A surge of start-up companies focused on using AI and machine learning to accelerate new drug discovery and optimal use of technology in surgery and radiation oncology.
However, as might be imagined, and with Professor Hawking’s message much in mind, the introduction of artificial intelligence into cancer care is not without controversy nor unresolved issues.
For example, a report by STAT (www.STATnews.com) claimed socalled ‘supercomputers’ have been making invalid conclusions about cancer treatment, raising concerns about whether healthcare professionals will be able to detect and prevent such new forms of potential error in treatment decisions. Equally, there is a need to manage effectively the shift in healthcare professional roles that must surely follow the introduction of artificial intelligence in the conduct of tasks currently conducted by humans.
As Co-Chair of the ECCO 2019 European Cancer Summit, I am convinced that the cancer community needs to come together urgently to focus on this topic.
We must further shape our collective response to this irresistible force for change, and in so doing, shape the future well. We need to maximise the benefits of AI for cancer care, while simultaneously taking action to prevent unintended harmful impacts.
Or to steal some phrasing from that Oxford mathematician whom I mentioned, we need to better align the goals of computers with those of the cancer community.
Page last modified: