Related Experiment Video
Updated: Nov 7, 2025
10:26
Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
Published on: May 19, 2023
2.2K
Evaluating a Fully Automated Pulmonary Nodule Detection Approach and Its Impact on Radiologist Performance
Kai Liu1, Qiong Li1, Jiechao Ma1
1Department of Radiology, Changzheng Hospital, Second Military Medical University, 415 Fengyang Rd, Shanghai, China 20003 (K.L., Q.L., W.T., Y.W., L.F., Y.X., S.L.); and Infervision Advanced Institute, Beijing, China (J.M., Z.Z., M.S., Y.D., C.X., R.Z.).
Radiology. Artificial Intelligence
|May 3, 2021
Summary
Deep learning (DL) models demonstrate higher sensitivity in detecting pulmonary nodules than manual review. Assisting radiologists with DL tools improved their performance and reduced reading time for lung nodule identification.
Area of Science:
- Radiology
- Artificial Intelligence
- Medical Imaging
Background:
- Pulmonary nodules require accurate detection for effective management.
- Radiologist performance in nodule detection can be influenced by various factors.
Purpose of the Study:
- To compare the sensitivity of a deep learning (DL) model against radiologists in detecting lung nodules.
- To evaluate if DL assistance enhances radiologist performance and efficiency.
Main Methods:
- Retrospective analysis of 12,754 chest CT scans for DL model development and testing.
- Comparison of DL model and radiologist detection sensitivity using free-response receiver operating characteristic curves.
- Assessment of radiologist performance with and without DL assistance.
Main Results:
- The DL model exhibited higher overall sensitivity for pulmonary nodule detection compared to manual review.
- Radiologist performance, particularly for junior radiologists, showed dependency on patient age.
- DL assistance significantly improved radiologist detection performance and reduced reading time.
Conclusions:
- Deep learning models show significant potential to improve pulmonary nodule identification.
- DL-assisted workflows can enhance nodule management and radiologist efficiency.
