Artificial intelligence is changing the way doctors make complex medical decisions. Capricorn, an AI tool developed by Google in collaboration with the Princess Máxima Center, searches millions of scientific publications in minutes and links them to anonymized patient data. This gives pediatric oncologists faster and more accurate insight into possible treatment options.
What if artificial intelligence could help doctors find better treatment options for children who are failing standard therapies? That question is at the heart of Capricorn, an AI tool that helps oncologists map relevant literature on potential treatments. The initiator is Dr. Uri Ilan, a non-practicing pediatric oncologist and physician-researcher at the Princess Máxima Center. He also works at Google Research and is one of the driving forces behind digital innovations at the Princess Máxima Center. His interest in AI stems directly from clinical practice. "The human body is simply too complex for humans to fully understand," says Ilan. "That's where AI can make a difference."
Capricorn
Capricorn helps doctors with complex patient cases, especially when standard treatments fail. The AI tool searches tens of millions of scientific publications in PubMed in minutes and links them to anonymized patient data, such as genetic characteristics and previous treatment responses. "You want to know if there has ever been a patient similar to yours, what treatment was chosen, and whether it worked." Based on this, Capricorn quickly and systematically maps similar cases, helping doctors make treatment decisions.
Capricorn was deliberately designed as a transparent tool: doctors can see exactly why certain publications are selected. The tool uses a points system in which studies are weighted based on factors such as clinical evidence, relevance to the disease, and applicability to children. "A treatment tested in a clinical trial is given more weight than a lab experiment," explains Ilan. Privacy is a prerequisite: patient information is automatically anonymized before data entry.
While a comprehensive literature review can easily take doctors one to three days, Capricorn delivers an overview in just a few minutes. The benefits go beyond saving time, Ilan emphasizes. "If you search yourself, you're mainly looking at what you already know," he says. "But many genetic abnormalities are overlooked, simply because you don't have the time." Capricorn can process all this information simultaneously and also include lesser-known mutations. In practice, this means that doctors come to multidisciplinary meetings with a more complete and up-to-date overview.
Sharing knowledge
For Ilan, the power of Capricorn lies not only in the technology itself, but especially in what it enables for doctors outside specialized centers. "AI allows us to share our expertise," he says. "The knowledge someone has accumulated over years can now be available to every doctor in seconds. This can make all the difference, especially for hospitals with fewer resources." Capricorn wasn't developed for the Princess Máxima Center itself, but rather to make specialist knowledge more widely accessible. "We have a responsibility to share that technology."
From supporting to deciding
According to Ilan, we're only just scratching the surface of what AI can do for healthcare. The biggest gains, for now, lie in bringing together enormous amounts of data into something doctors can actually work with. "Biology is too complex to understand in its individual pieces," he says. "AI can translate that complexity into concrete, usable information." In this phase, the doctor retains ultimate responsibility: AI makes recommendations, humans make the final decision. "This collaboration helps to visibly improve the quality of medical decision-making."
Still, he expects this division of roles won't last. "Statistically, AI makes better choices than humans. Doctors often bring their own experiences and biases, and that can influence a decision. AI looks purely at the data and can therefore make systematically better choices." He acknowledges that this encounters resistance: "Compare it to self-driving cars: statistically, they cause far fewer accidents. But when things do go wrong, people find it harder to accept than when a human makes a mistake. Yet, this is the direction we're heading. It's not a matter of whether AI will make decisions, but when."
Culture
For Ilan, embracing AI is primarily a matter of culture and consciously managing risks. At the Princess Máxima Center, he encourages a mindset where innovation and technology are seen as welcome tools, not threats. "When a doctor treats a patient, they always assess the risks: what could happen, what are the side effects, what can you expect? That's how we should work with AI: use it, but the risks must be clear and understood," he says. Ultimately, according to Ilan, it's about doctors learning to collaborate with AI and utilize its potential: "We have to learn to work with AI, not run away from it because it might replace us."
Want to learn more? Dr. Uri Ilan will speak at the Medical Technology event on January 28th about the development of Capricorn and the role of artificial intelligence in healthcare.
Nederlands