As the first wave of AI-based solutions is reaching maturity, image-guided cancer treatment is moving toward the molecular level, observes Peter de With.
With the advent of deep learning and AI techniques in healthcare, the past decade has shown many developments in AI-based support of cancer treatment. Nowadays, the majority of this treatment depends on a form of imaging, such as the video-based endoscopic analysis of the esophagus, stomach and colon using endoscopy. Breast and prostate cancer are visualized with X-ray imaging and ultrasound.
For these cases, the imaging data have become of sufficiently high quality to enable detection and/or characterization of the disease. This setting is attractive for the deployment and application of computer-aided design-based support for the physician, where the CAD system helps in real-time detection of cancer and/or the classification during patient inspection.
In the VCA group at Eindhoven University of Technology’s EE Department, we’ve specialized in esophageal and colon cancer. In collaboration with the Amsterdam University Medical Centers and the Catharina Hospital Eindhoven, our research has matured into an efficient AI solution for embedded integration into endoscopy equipment and a small yet programmable add-on box. Since the benefit to physicians is undisputable, the research has piqued the interest of the medical equipment industry.
While this industry is working on enhancing the existing equipment with AI support, the research community is shifting focus to new barriers. An interesting path to consider is the growing success of new medical drugs for cancer disease control and containment. Such pharmacological treatment will benefit from even more precise imaging at the cell level.
For example, immunotherapy doesn’t attack the cancer itself but enhances the resistance of healthy cells and makes them stronger. Such therapy works best if a drug is delivered to the right place in the body, preferably without surgical intervention. Last spring, Holland Health approved a study called Foresight, spanning several sub-projects in this direction. At the TUE, our interest is obviously in improving the imaging technology here.
Drug molecules are usually much larger than other compounds in the human body. Such compounds can be imaged by high-quality MRI or even specialized forms of ultrasound. Hence, the breakthrough in ultra-high-resolution imaging of modern digital pathology scanners approaching the level of cells has to be further explored and expanded to arrive at imaging of large complex molecules and finding their locations in the human body.
This research is of world-class level (several participants have published in Nature and the like) but is severely hampered by Dutch bureaucracy. The research partners have to write extensive reports, make websites, keep track of budgets, have technical evaluation meetings and so on, without any form of reimbursement for their time. Not to mention the fact that after two (!) budget cuts ordered by The Hague, funding was more than halved from 20 million to 9 million euros.
To solve this motion of no confidence toward our Dutch top cancer researchers, I suggest a new way of working, which has already proved itself elsewhere. If Dutch top researchers reach world-class status without financial mistakes, they’re assigned a title like “chartered/trusted” scientist, similar to the chartered engineers in the United Kingdom. They would become a person of trust, so that they’re treated with the respect they deserve and can stand in for the people and projects they supervise without being subjected to the full extent of Dutch bureaucracy.
This new ‘knife’ will cut beneficially in two ways. First, it’s known that if you give trust to people, the good ones (ie the large majority) work harder, and with low bureaucracy, the efficiency and output will become higher. Secondly, faster progress in research will benefit the growing number of cancer patients all over the world.
There will be ample opportunity for spinoff companies and drug creation. Let’s see whether we can sustain or extend the strong position of the Netherlands for cancer treatment and related drugs in the near future.
