Fully automatic quantification of pulmonary fat attenuation volume by CT: an exploratory pilot study

Area of Science:

• Pulmonary medicine and radiology
• Medical imaging analysis
• Biomarker discovery

Background:

• Non-malignant chronic diseases pose significant public health challenges.
• Altered lung lipid metabolism and deposition are linked to fibrotic interstitial lung disease (fILD) and chronic obstructive pulmonary disease (COPD).
• Computed tomography (CT) is a key imaging modality for lung disease assessment.

Purpose of the Study:

• To investigate alterations in lung lipid metabolism using CT-based analysis of pulmonary fat attenuation volume (CTpfav).
• To evaluate CTpfav and the derived pulmonary fat index (PFI) as potential imaging biomarkers in fILD and COPD.
• To introduce a fully automated method for CTpfav quantification.

Main Methods:

• Observational retrospective single-center study of 716 chest CT scans (279 control, 283 COPD, 154 fILD).
• Fully automated quantification of CTpfav using lung segmentation and Hounsfield unit (HU) thresholding.
• Calculation of PFI by normalizing CTpfav to CT lung volume; statistical analysis via Kruskal-Wallis test.

Main Results:

• Patients with fILD showed significantly increased CTpfav and PFI compared to controls (p < 0.001).
• Patients with COPD exhibited significantly decreased CTpfav and PFI compared to controls (p < 0.001).
• The study utilized an open-source, automated algorithm for CTpfav quantification.

Conclusions:

• CTpfav and PFI show potential as imaging biomarkers for altered lung lipid metabolism in fILD and COPD.
• Automated CTpfav quantification may establish it as a novel imaging biomarker for chronic lung diseases.
• Further research is warranted to validate these findings and explore clinical relevance in lung disease management.