Quantification of lung function on CT images based on pulmonary radiomic filtering

Area of Science:

• Radiology
• Medical Imaging
• Computational Pathology

Background:

• Pulmonary ventilation assessment is crucial for diagnosing lung diseases.
• Current methods like PET/SPECT imaging provide functional information but can be limited.
• Radiomics offers a novel approach to extract quantitative features from medical images.

Purpose of the Study:

• To develop and validate a radiomics filtering technique for characterizing regional pulmonary ventilation using lung CT.
• To assess the potential of radiomic features as biomarkers for lung ventilation defects.

Main Methods:

• Lung CT images from 46 patients were segmented.
• A 3D sliding window kernel was applied to extract 53 radiomic features, creating feature maps.
• Radiomic feature maps were correlated with functional imaging data (Galligas PET or DTPA-SPECT) using Spearman correlation.

Main Results:

• Two radiomic features, GLRLM-based Run-Length Non-Uniformity and GLCOM-based Sum Average, showed high correlation with functional imaging.
• Median Spearman correlation coefficients (ρ) of 0.46 and 0.45 were observed for these features, respectively.
• These findings were consistent across patients and imaging modalities.

Conclusions:

• Radiomics analysis of lung CT can identify heterogeneous lung parenchyma associated with diminished ventilation.
• The developed radiomics technique shows potential as a complementary tool for lung ventilation quantification.
• This study provides a foundation for future research into radiomics for lung disease assessment.