Related Experiment Video
Updated: Jul 16, 2026
10:26
A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
Published on: May 19, 2023
Automatic detection and segmentation of ground glass opacity nodules
Jinghao Zhou1, Sukmoon Chang, Dimitris N Metaxas
1CBIM, Rutgers, The State University of New Jersey, NJ, USA. jhzhou@eden.rutgers.edu
Medical Image Computing and Computer-Assisted Intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
|March 16, 2007
Summary
A new method automatically detects and segments Ground Glass Opacity (GGO) in CT scans. This aids early lung cancer diagnosis by overcoming manual detection challenges.
Area of Science:
- Radiology
- Medical Imaging
- Computer-Aided Diagnosis
Background:
- Ground Glass Opacity (GGO) on CT scans can indicate early lung cancer.
- Manual GGO detection is challenging due to indistinct boundaries and observer variability.
- Accurate GGO detection and segmentation are crucial for improved lung cancer prognosis.
Purpose of the Study:
- To develop an automated method for detecting and segmenting Ground Glass Opacity (GGO) in chest CT images.
- To address the limitations of manual GGO detection and segmentation.
- To provide a tool for more accurate and reproducible GGO analysis.
Main Methods:
- A novel method for automatic GGO detection and segmentation from chest CT images.
- GGO detection using a boosted k-NN classifier with nonparametric density estimates.
- GGO segmentation via analysis of texture likelihood maps in detected regions.
Main Results:
- The proposed method successfully detected all 10 GGO nodules in clinical chest CT volumes.
- Only one false positive nodule was identified.
- Statistical validation confirmed the classifier's efficacy for GGO detection.
- Promising results were achieved for automatic GGO segmentation.
Conclusions:
- The developed method offers a powerful tool for automatic GGO detection.
- It provides accurate and reproducible segmentation of GGO regions.
- This approach can significantly aid in the early diagnosis and management of lung cancer.
