Spatial Dependence of CT Emphysema in Chronic Obstructive Pulmonary Disease Quantified by Using Join-Count Statistics

Area of Science:

• Pulmonary Medicine
• Radiology
• Quantitative Imaging

Background:

• Current CT emphysema metrics in COPD assess extent or clustering but not voxel spatial arrangement.
• There is a need for advanced CT measurements to better characterize emphysema heterogeneity.

Purpose of the Study:

• To develop and validate a novel CT measurement, the normalized join count (NJC), for quantifying emphysema voxel compactness.
• To assess if NJC can differentiate COPD disease severity and correlate with lung function and visual emphysema scores.

Main Methods:

• Secondary analysis of the Canadian Cohort Obstructive Lung Disease study (n=1294) with CT scans and lung function data.
• Development of NJC to measure emphysema voxel compactness; comparison with %LAA-950insp, LAC, and visual scores.
• Multivariable ridge regression used to assess associations between CT measurements, lung function (DLCO/VA), and radiologist scores.

Main Results:

• NJC, %LAA-950insp, and LAC were higher in moderate-severe COPD vs. never-smokers/at-risk groups (P < .05).
• NJC uniquely differentiated mild from moderate-severe COPD (P < .05).
• NJC showed the strongest association with diffusion capacity (DLCO/VA) (P = .002) and visual emphysema score (P < .001) in multivariable analysis.

Conclusions:

• The normalized join count (NJC) effectively quantifies emphysema voxel compactness.
• NJC demonstrates potential as a valuable tool for assessing COPD severity and its impact on lung function.
• Emphysema assessment in COPD may benefit from considering the spatial arrangement of low-attenuating voxels beyond just their extent or cluster size.