Sjaak Burgers, a pulmonologist at the Netherlands Cancer Institute in Amsterdam, no longer squints at CT scans trying to guess where one irregular mesothelioma tumor ends and healthy tissue begins. Instead, he lets ARTIMES—an AI model developed at his own hospital—map the entire tumor volume with pixel-level precision, transforming a once-frustrating diagnostic process into a clear, data-driven roadmap for treatment. Pleural mesothelioma, a rare and aggressive cancer caused by asbestos exposure, spreads like smoke along the lung lining, making traditional measurement methods nearly useless. For years, doctors relied on RECIST, an international standard that measures tumor diameter, but this approach fails when tumors grow in thin, irregular layers. The result? Uncertainty for patients and flawed data in clinical trials. Now, ARTIMES is changing the game.
Developed by a team of AI experts, radiologists, and pulmonologists at the NKI, ARTIMES analyzes over 11,000 CT scans from more than 2,000 patients across 121 hospitals worldwide. It doesn’t just estimate—it meticulously segments every slice of tumor tissue, calculating total volume with a consistency and accuracy no human can match. "We are the first in the world to demonstrate that AI outperforms humans in this area, and that physicians can actually base their decisions on it," says Kevin Groot Lipman, technical physician and lead author of the study published in The Lancet Oncology. The model doesn’t replace doctors; every result is reviewed by a physician. But it empowers them to detect treatment failure earlier, switch therapies faster, and spare patients from unnecessary side effects and costs.
Already in clinical use at the NKI, ARTIMES is bridging the gap between cutting-edge AI and real-world medicine. The team didn’t stop at measurement—they built clinical guidelines to help doctors interpret AI-generated data and make timely treatment decisions. In tests across eight clinical trials, ARTIMES proved far more accurate than RECIST, promising to make future trials smaller, faster, and more reliable. This could accelerate the development of new therapies for a disease with few options. And the impact won’t stop at mesothelioma. The NKI is now adapting the technology for lung cancer and brain metastases, opening doors for AI-driven oncology across tumor types.
Though currently limited to use within the Netherlands under EU in-house exemption rules, the model is publicly available online, inviting researchers worldwide to adopt and validate it. The team is pushing for broader regulatory approval, hoping to see ARTIMES used globally. "I expect this model to come as a shock to physicians and researchers outside the mesothelioma field," Groot Lipman predicts. For patients, it’s not just about better data—it’s about hope, clarity, and the chance to fight smarter.