Related Experiment Video
Updated: Mar 24, 2026
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.9K
Automated Lung Segmentation from HRCT Scans with Diffuse Parenchymal Lung Diseases
Ammi Reddy Pulagam1, Giri Babu Kande2, Venkata Krishna Rao Ede3
1Vasireddy Venkatadri Institute of Technology, Nambur, Guntur, AP, India. pulagamammireddy@gmail.com.
Journal of Digital Imaging
|March 11, 2016
Summary
This study introduces an automated lung segmentation algorithm for high-resolution computed tomography (HRCT) scans. The novel method accurately segments lungs, even with dense abnormalities, showing promise for computer-aided diagnosis systems.
Area of Science:
- Medical Imaging
- Computer-Aided Diagnosis
- Radiology
Background:
- Accurate lung segmentation in high-resolution computed tomography (HRCT) is crucial but challenging, especially with dense abnormalities.
- Existing methods often struggle with diseased lung tissue, limiting their clinical applicability.
- Automated segmentation is essential for efficient computer-aided diagnosis (CAD) systems for diffuse lung diseases.
Purpose of the Study:
- To develop and evaluate a novel, fully automated algorithm for lung segmentation in HRCT images.
- To assess the algorithm's performance in the presence of various lung abnormalities.
- To determine the suitability of the algorithm as a preliminary step in CAD systems for diffuse lung diseases.
Main Methods:
- A novel automated lung segmentation algorithm was developed using modified convex hull and mathematical morphology techniques.
- The algorithm was tested on sixty randomly selected lung HRCT scans exhibiting diverse abnormalities.
- Performance was quantitatively evaluated using metrics such as Dice Similarity Coefficient (DSC) and shape differentiation metrics (dmean, drms).
Main Results:
- The proposed algorithm achieved high segmentation accuracy, with a Dice Similarity Coefficient of 98.62%.
- Shape differentiation metrics were favorable (dmean = 1.39 mm, drms = 2.76 mm), indicating precise boundary delineation.
- The algorithm demonstrated robustness in segmenting lungs affected by various disease patterns, though minor limitations were noted at lung apices and bases.
Conclusions:
- The developed automated lung segmentation algorithm is effective and accurate, even for HRCT images with dense abnormalities.
- The algorithm's high performance makes it a strong candidate for the initial stage of computer-aided diagnosis systems for diffuse lung diseases.
- Further refinement may address limitations observed at the lung apices and bases to enhance overall clinical utility.
