Related Experiment Video
Updated: Sep 26, 2025
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
313
BreastScreening-AI: Evaluating medical intelligent agents for human-AI interactions
Francisco Maria Calisto1, Carlos Santiago1, Nuno Nunes2
1Institute for Systems and Robotics, Avenida Rovisco Pais 1, 1049-001 Lisbon, Portugal.
Artificial Intelligence in Medicine
|April 17, 2022
Summary
BreastScreening-AI, a novel deep learning tool, significantly improved multimodal breast image classification. The Clinician-AI scenario reduced false positives by 27% and false negatives by 4%, enhancing diagnostic accuracy.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Healthcare
- Clinical Decision Support
Background:
- Multimodal breast image classification is crucial for early cancer detection.
- Integrating Artificial Intelligence (AI) into clinical workflows presents challenges and opportunities.
- Evaluating clinician interaction and AI impact is vital for successful adoption.
Purpose of the Study:
- To develop and evaluate BreastScreening-AI for multimodal breast image classification.
- To compare diagnostic performance between a Clinician-Only and a Clinician-AI scenario.
- To assess clinician acceptance, AI impact, and potential benefits in a real clinical setting.
Main Methods:
- Development of BreastScreening-AI using a deep learning method.
- Implementation in two scenarios: Clinician-Only and Clinician-AI.
- Extensive evaluation with 45 clinicians across nine institutions, including patient selection and qualitative/quantitative analysis.
Main Results:
- The Clinician-AI scenario demonstrated superior diagnostic performance.
- A 27% decrease in False-Positives and a 4% decrease in False-Negatives were observed with AI assistance.
- 91% of clinicians reported positive impacts on expectations and satisfaction, with a 3-minute reduction in diagnosis time per patient.
Conclusions:
- The integration of AI in breast cancer screening significantly enhances diagnostic accuracy and efficiency.
- BreastScreening-AI positively influences clinician perception and satisfaction.
- AI-assisted systems show potential to mitigate clinical errors and improve patient outcomes.
