Related Experiment Video
Updated: Jul 15, 2025
07:15
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
6.8K
External validation of deep learning-based automated detection algorithm for chest radiograph: practical issues in
Da Eul Lee1, Kum Ju Chae1,2, Gong Yong Jin1
1Department of Radiology, Research Institute of Clinical Medicine of Jeonbuk National University-Biomedical Research Institute of Jeonbuk National University Hospital, Jeonju, Republic of Korea.
Acta Radiologica (Stockholm, Sweden : 1987)
|September 26, 2023
Summary
Deep learning-based automated detection (DLAD) shows high diagnostic performance for thoracic diseases in outpatient clinics. DLAD significantly improved interpretation accuracy and efficiency, especially for pulmonologists.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Medicine
- Thoracic Radiology
Background:
- Limited data exists on the diagnostic performance of deep learning-based automated detection (DLAD) for thoracic diseases in real-world outpatient settings.
- Validating DLAD in clinical practice is crucial for its adoption.
Purpose of the Study:
- To validate the diagnostic performance of DLAD for thoracic diseases in an outpatient clinic setting.
- To analyze the impact of DLAD on the interpretation time of chest radiographs.
Main Methods:
- A retrospective single-center study analyzed 205 chest radiographs with DLAD and paired chest CT scans.
- Radiologists and pulmonologists performed observer performance tests with and without DLAD assistance.
- Interpretation time was measured for each observer group.
Main Results:
- DLAD achieved a high area under the receiver operating characteristic curve (AUC) of 0.920.
- DLAD significantly improved AUC for pulmonologists (0.756 to 0.853) and radiologists (0.782 to 0.854).
- Pulmonologists experienced over a 50% reduction in interpretation time when using DLAD.
Conclusions:
- DLAD demonstrates strong diagnostic performance for interpreting chest radiographs in outpatient clinics.
- DLAD is particularly beneficial for pulmonologists, enhancing both accuracy and efficiency.
- The findings support the integration of DLAD into routine clinical workflows for thoracic disease diagnosis.
