Related Experiment Video
Updated: Dec 30, 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
2.0K
An unsupervised semi-automated pulmonary nodule segmentation method based on enhanced region growing
He Ren1,2, Lingxiao Zhou1, Gang Liu1
1Shanghai Public Health Clinical Center & Institutes of Biomedical Sciences, School of Basic Medical Sciences, School of Data Science, Fudan University, Shanghai 200032, China.
Quantitative Imaging in Medicine and Surgery
|January 21, 2020
Summary
A new semi-automated pulmonary nodule segmentation algorithm (ReGANS) offers faster and more precise lung nodule identification in CT images. This computer-aided diagnosis tool improves efficiency and accuracy for radiologists, aiding clinical decisions.
Area of Science:
- Medical Imaging
- Computer-Aided Diagnosis
- Artificial Intelligence in Medicine
Background:
- Computer technology is increasingly vital for clinical diagnosis, particularly in medical imaging.
- Accurate segmentation of lung nodules enhances physician diagnostic efficiency.
Purpose of the Study:
- To develop and evaluate a novel semi-automated pulmonary nodule segmentation algorithm, ReGANS.
- To improve the speed and precision of lung nodule segmentation in computed tomography (CT) images.
Main Methods:
- Developed ReGANS, a region growing-based algorithm with automatic threshold calculation, lesion pre-projection, and optimized region growing.
- Algorithm segments lung nodules from CT images based on an initial manual point.
Main Results:
- ReGANS segmented a pulmonary nodule in an average of 0.83 seconds.
- Achieved high accuracy metrics (PRI=0.93, GCE=0.06, VoI=0.3) compared to radiologist segmentations.
- Demonstrated robustness across multiple datasets with a 15% error rate in nodule coverage evaluation.
Conclusions:
- ReGANS offers superior speed and precision in lung nodule segmentation compared to existing algorithms.
- The algorithm holds significant clinical value for medical imaging diagnosis and pre-data preparation.
- Provides a faster, more convenient method for AI-driven medical image analysis.
