Related Experiment Video
Updated: Apr 28, 2026
07:53
Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer
Published on: October 13, 2023
2.3K
Artificial Intelligence in Pulmonary Endoscopy: Current Evidence, Limitations, and Future Directions
Sara Lopes1, Miguel Mascarenhas2,3,4, João Fonseca2,3,4
1Thoracic Surgery, Portuguese Institute of Oncology of Porto, 4200-072 Porto, Portugal.
Journal of Imaging
|April 27, 2026
Summary
Artificial intelligence (AI) shows promise in pulmonary endoscopy for lesion detection and navigation. However, further validation and integration are needed for widespread clinical adoption of these advanced AI tools.
Area of Science:
- Pulmonary Medicine
- Medical Imaging
- Artificial Intelligence
Background:
- Artificial intelligence (AI) is increasingly integrated into pulmonary endoscopy, enhancing diagnostic capabilities and procedural support.
- Advancements in AI, computer vision, and robotics offer potential for automated lesion detection and navigation in the lungs.
- Current evidence for AI in pulmonary endoscopy is varied, with ongoing challenges in translating research into practice.
Purpose of the Study:
- To review the current applications and advancements of AI in various pulmonary endoscopy techniques.
- To examine the limitations, regulatory aspects, and barriers to implementing AI in clinical practice.
- To assess the impact of AI on training and workflow optimization in interventional pulmonology.
Main Methods:
- Comprehensive review of AI applications in white-light bronchoscopy, image-enhanced bronchoscopy, endobronchial ultrasound (EBUS), and virtual/robotic bronchoscopies.
- Analysis of AI's role in workflow optimization, training simulations, and decision-support tools.
- Examination of methodological limitations, regulatory hurdles, and implementation barriers for AI in endoscopy.
Main Results:
- Deep learning models demonstrate efficacy in detecting mucosal abnormalities and characterizing lymph nodes during EBUS-TBNA.
- AI improves lesion localization and reduces operator variability in pulmonary procedures.
- AI-assisted simulation platforms are transforming training paradigms, though most studies lack external validation and clear model explainability.
Conclusions:
- AI holds significant potential to enhance lesion detection, navigation, and training within pulmonary endoscopy.
- Widespread adoption requires robust prospective validation, standardized datasets, and transparent reporting of AI models.
- Multidisciplinary collaboration and careful integration into clinical workflows are crucial for successful AI implementation.
