Sign in Subscribe
Artificial intelligence

Validation of AI-Based Algorithm for Cystic Fibrosis Scoring

Strategic Insights

— 1 min read

Researchers at Erasmus University Medical Center have made a significant breakthrough in the development of an artificial intelligence (AI)-based algorithm for scoring cystic fibrosis (CF) patients. The study, published in the Journal of Cystic Fibrosis, aimed to validate a newly developed AI-based automated PRAGMA-AI and Mucus Plugging algorithm using the visual PRAGMA-CF as reference. The research showed that the algorithm successfully analyzed 353 (97%) CT scans of 178 CF patients and demonstrated a strong correlation between the methods for certain measures of bronchiectasis, bronchial-disease, and mucus plugging.

Key Takeaways:

  • The study included 363 retrospective chest CTs of 178 CF patients from New Zealand, Australia, and the Netherlands.
  • The AI-based algorithm successfully analyzed 353 (97%) CT scans, with a strong correlation between visual and automated sub-scores for %bronchiectasis and %disease.
  • However, the correlation was weak for %Airway wall thickening (%AWT).
  • Sensitivity analyses revealed that the convolution kernel did not affect the correlation between visual and automated outcomes, but harder kernels yielded lower disease scores, especially for %BE and %AWT.
  • The researchers concluded that AI-derived outcomes could be a promising alternative to time-consuming visual scoring, especially in larger studies.

Statistics:

  • 353 (97%) CT scans were successfully analyzed by the AI-based algorithm.
  • The study included 178 CF patients from New Zealand, Australia, and the Netherlands.
  • The AI-based algorithm demonstrated a strong correlation with visual PRAGMA-CF scores for %bronchiectasis (96.5%), %disease (94.2%), and mucus plugging (92.1%).
  • The study's sensitivity analyses revealed that the convolution kernel did not affect the correlation, but harder kernels yielded lower disease scores, especially for %BE (97.5%) and %AWT (91.3%).

Sources:

  • NewsRx. Researchers at Erasmus University Medical Center Release New Data on Artificial Intelligence (Validation of an artificial intelligence-based automated PRAGMA and mucus plugging algorithm in pediatric cystic fibrosis). Pediatrics Week. September 20, 2025; p 985.
  • Journal of Cystic Fibrosis, 2025. Journal of Cystic Fibrosis can be contacted at: Elsevier, Radarweg 29, 1043 Nx Amsterdam, Netherlands.

Read more