Related Experiment Video
Updated: Jan 10, 2026
08:05
Detection and Isolation of Cancer in Prostate Biopsies Using Stimulated Raman Histology and Artificial Intelligence
Published on: June 10, 2025
1.1K
Adding artificial intelligence case malignancy scoring to reduce screen-reading workload in breast screening program:
Andrea Nitrosi1, Paolo Giorgi Rossi2, Laura Verzellesi3,4
1Medical Physics Unit - Oncology and Innovative Technologies Department, Azienda USL - IRCCS di Reggio Emilia, Reggio Emilia, Italy.
La Radiologia Medica
|November 26, 2025
Summary
Integrating AI case malignancy score (AI-CMS) into mammography screening reduces radiologist workload by up to 36.1% while maintaining cancer detection rates. This AI-supported strategy also lowers recall and false positive rates, proving economically sustainable.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Healthcare
- Radiology
Background:
- Mammography screening is crucial for early breast cancer detection.
- Radiologist workload can impact efficiency and diagnostic accuracy.
- AI algorithms can provide confidence scores for malignancy assessment.
Purpose of the Study:
- To evaluate a strategy integrating AI case malignancy score (AI-CMS) into mammography screening.
- To assess the impact of AI-CMS on radiologist workload and screening performance.
- To determine the economic sustainability of AI-CMS integration.
Main Methods:
- Retrospective analysis of 89,176 screening mammograms.
- Simulation of a strategy using AI-CMS to guide radiologist recall decisions.
- Evaluation of AI-CMS thresholds (5-25%) on recall rate, cancer detection, and workload.
Main Results:
- AI-CMS integration reduced human workload by 13.4% to 36.1%.
- Recall rate decreased to 4.0-4.3%, and false positive rate decreased from 3.9% to 3.5-3.8%.
- Positive predictive value increased to 12.6-13.3%, and no cancers were missed at 5% threshold.
Conclusions:
- AI-CMS support can significantly decrease radiologist workload in mammography screening.
- The strategy shows potential for reducing unnecessary recalls without compromising cancer detection.
- The AI-CMS integration approach is economically viable and sustainable.
