Related Experiment Video
Updated: Jul 13, 2025
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
1.5K
Computer-Aided Pulmonary Fibrosis Detection Leveraging an Advanced Artificial Intelligence Triage and Notification
Kavitha C Selvan1, Angad Kalra2, Joshua Reicher2,3
1Section of Pulmonary and Critical Care Medicine, Department of Internal Medicine, University of Chicago Medicine, Chicago, IL, USA.
Journal of Clinical Medicine Research
|October 12, 2023
Summary
ScreenDx-LungFibrosis™, an AI tool, accurately detects pulmonary fibrosis (PF) in chest CT scans. This software shows potential to significantly improve patient outcomes for interstitial lung disease by enabling earlier recognition and referral.
Area of Science:
- Pulmonary Medicine
- Radiology
- Artificial Intelligence in Healthcare
Background:
- Early recognition and referral of pulmonary fibrosis (PF) are crucial for improving patient outcomes in interstitial lung disease.
- An artificial intelligence (AI) triage and notification software, ScreenDx-LungFibrosis™, was developed to enhance PF detection.
Purpose of the Study:
- To evaluate the performance metrics and processing time of the ScreenDx-LungFibrosis™ AI software for detecting pulmonary fibrosis.
- To determine if the software meets predefined diagnostic performance criteria for sensitivity, specificity, and processing speed.
Main Methods:
- ScreenDx-LungFibrosis™ was applied to chest computed tomography (CT) scans from a multisource dataset.
- The software's output (presence or absence of PF) was compared against clinical diagnoses.
- Diagnostic performance was assessed, with primary endpoints set at >80% sensitivity and specificity, and <4.5 min processing time.
Main Results:
- The study included 3,018 patients, with 22.9% diagnosed with PF.
- ScreenDx-LungFibrosis™ achieved a sensitivity of 91.3% and a specificity of 95.1% for PF detection.
- The mean processing time per scan was rapid, averaging 27.6 seconds.
Conclusions:
- ScreenDx-LungFibrosis™ demonstrates accurate and reliable identification of pulmonary fibrosis.
- The software's rapid processing time suggests its potential to significantly improve PF outcomes when integrated into routine chest CT analysis.
