Three-dimensional selective-scale texture analysis of computed tomography pulmonary angiograms

Area of Science:

• Radiology
• Medical Imaging
• Pulmonary Medicine

Background:

• Pulmonary disorders often involve altered lung vasculature and ventilation.
• Accurate assessment of these changes is crucial for diagnosis and management.

Purpose of the Study:

• To develop and assess 3-dimensional (3D) selective-scale texture analysis of computed tomography pulmonary angiography (CTPA).
• To identify texture correlates of ventilated and vascular lung for quantitative assessment.
• To aid in the visual and quantitative evaluation of pulmonary disorders with altered vasculature.

Main Methods:

• CTPA data from 8 patients (normals, pulmonary embolism, emphysema) were analyzed.
• Automated lung parenchyma segmentation and volume generation were performed.
• Ventilated and vascular lung components were isolated for texture analysis.
• 3D Laplacian of Gaussian filtering was applied to highlight fine and coarse textures.
• Texture quantification included mean gray-level intensity, entropy, and uniformity.
• Volume rendering and image fusion were used for visualization.

Main Results:

• Fine texture in ventilated lung showed a postural gradient and differed in emphysematous lungs.
• Coarse texture in vascular lung correlated with the degree of vascularity/perfusion.
• 3D texture analysis highlighted matched and mismatched defects in pulmonary disorders.

Conclusions:

• 3D filtered texture analysis is feasible for assessing ventilated and vascular lung.
• This technique shows potential for diagnosing pulmonary embolism (PE) in complex cases.
• Texture analysis may provide valuable quantitative insights into pulmonary disorders.