Back
Science

AI Model ARTIMES Outperforms Human Physicians in Evaluating Pleural Mesothelioma Treatment Response

Generated on: Last updated: 1 sources
View source

ARTIMES: AI Outperforms Physicians in Measuring Mesothelioma Treatment Response

"We can now assess tumor response much more accurately... This allows a patient to stop treatment earlier and, if possible, switch to a different treatment."
— Sjaak Burgers, Pulmonologist, Netherlands Cancer Institute

Researchers and physicians at the Netherlands Cancer Institute (NKI) have developed an artificial intelligence model called ARTIMES that evaluates treatment response in pleural mesothelioma by measuring tumor volume from CT scans. According to a study published in The Lancet Oncology, the AI model outperforms physicians and the current standard RECIST criteria in accuracy.

Key Breakthroughs

  • Massive Training Data: The AI model was trained on over 11,000 CT scans from more than 2,000 patients across 121 hospitals worldwide.
  • Volume vs. Diameter: ARTIMES measures entire tumor volume, whereas RECIST relies on diameter measurements—which are problematic for pleural mesothelioma due to its irregular growth along the lung wall.
  • Superior Accuracy: In tests using data from eight clinical trials, AI measurements were more accurate than RECIST criteria in assessing treatment efficacy and predicting survival.
  • Human Oversight: The model's outputs are always reviewed by a physician before clinical decisions are made.
  • Regulatory Status: Currently, ARTIMES is approved for use only at the Netherlands Cancer Institute under EU regulations; the team is pursuing broader approval.

Statements from the Research Team

"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."
— Kevin Groot Lipman, Technical Physician and Lead Author

Sjaak Burgers added: "We can now assess tumor response much more accurately... This allows a patient to stop treatment earlier and, if possible, switch to a different treatment."

Background

RECIST criteria have been the international standard for evaluating solid tumor response but have been criticized for inaccuracy in predicting patient outcomes, especially for pleural mesothelioma.

The NKI is also testing AI models for lung cancer and brain metastases. Notably, the ARTIMES model is publicly available online for researchers worldwide.

Implications for the Future

The model is expected to improve clinical trial reliability and efficiency by providing more accurate measurements of treatment response, potentially accelerating the development of new therapies.