Automatic Approach for Lung Segmentation with Juxta-Pleural Nodules from Thoracic CT Based on Contour Tracing and...

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Pulmonary Imaging Analysis

Background:

• Accurate lung segmentation is crucial for diagnosing pulmonary diseases.
• Juxta-pleural nodules present a significant segmentation challenge in medical imaging.

Purpose of the Study:

• To develop a fully automatic framework for lung segmentation.
• To specifically address the challenge of segmenting juxta-pleural nodules.

Main Methods:

• The framework employs a three-phase approach: skin boundary detection, rough lung contour segmentation, and pulmonary parenchyma refinement.
• Key techniques include image aligning, morphology operations, connective region analysis, diagonal-based border tracing, maximum cost path algorithm, arc-based border smoothing, and concave-based border correction.

Main Results:

• The system was evaluated on 45 chest scan volumes.
• Achieved promising results with metrics like Volume Overlap Error (VOE) of 3.5057 ± 1.3719% and Average Surface Distance (ASD) of 0.7917 ± 0.2741 mm.
• Demonstrated an average processing time of 2 seconds per image.

Conclusions:

• The proposed automatic lung segmentation framework is effective, particularly for cases involving juxta-pleural nodules.
• The method offers a balance of accuracy and computational efficiency, making it a valuable tool for clinical applications.