Semi-Automated Lung Segmentation Based on Region-Growing Methods in Interstitial Lung Disease

Area of Science:

• Medical Imaging
• Radiology
• Pulmonary Medicine

Background:

• Computed tomography (CT) is essential for diagnosing pulmonary disorders.
• Quantitative CT analysis, crucial for conditions like fibrosis, necessitates accurate lung segmentation.
• Manual segmentation is laborious and subjective; automated methods can be unreliable.

Purpose of the Study:

• To develop and evaluate a semi-automated lung segmentation method for CT images.
• To address the limitations of manual and fully automated segmentation techniques.
• To improve the efficiency and accuracy of lung segmentation in computer-aided diagnosis (CAD) systems.

Main Methods:

• Implementation of a region-growing algorithm for lung segmentation.
• Development of a semi-automated approach balancing automation and user interaction.
• Testing and validation of the segmentation method on a software platform.

Main Results:

• The proposed semi-automated method effectively delineates lung boundaries in CT scans.
• The region-growing approach balances automation with necessary user control.
• The technique minimizes both computational complexity and manual effort required for segmentation.

Conclusions:

• The developed semi-automated lung segmentation technique provides accurate delineation of lung boundaries.
• This method offers an efficient alternative to manual segmentation for quantitative CT analysis.
• The approach is suitable for CAD systems in diagnosing pulmonary diseases like COPD, pneumonia, and lung cancer.