Computed Tomography Spatial and Temporal Trajectories to Quantify Progression Patterns in Chronic Obstructive Pulmonary Disease

Abstract

<p>Parametric Response Mapping (PRM) is a computed tomography image analysis technique that quantifies emphysema (PRM Emph) and small airway disease (PRM fSAD) in Chronic Obstructive Pulmonary Disease. In this study, we quantified the longitudinal changes in PRM fSAD and PRM Emph using 3-D clustering and voxel-wise classification approaches and investigated the association with lung function decline. The slope ”D” of the linear regression of the number of clusters vs. cluster size for baseline (BL) and follow-up (FU) PRM fSAD and PRM Emph images were quantified. Increased D indicates more large clusters; small D indicates many small clusters. PRM fSADvoxels at FU were classified as “growth” if they were spatially connected to voxels at BL, and “newly-formed” if not connected to PRM fSADvoxels at BL. The same process was repeated for emphysema voxels. For PRM Emph, there was a significant increase in D at FU for At Risk (p=0.001) and moderate-severe (p=0.02) participants. For PRM fSAD, there was a significant increase in D at FU for At Risk, mild, and moderate-severe (p<0.0001). In COPD participants, the FU-BL change in lung function FEV m1asure was associated with the FU-BL change in D PRM Emph (SEM=0.13, p=0.034). The FU-BL change in FEV /FVC 1as associated with the FU-BL change in D PRM fSAD (SEM= -5.84, p=0.0003). Overall, newly-formed PRM Emph and growth PRM Fsadvoxels were associated with FEV 1ecline (p= 0.038 and p=0.031, respectively), and with FEV /FV1 decline (p<0.0001 and p=<0.0001, respectively). In conclusion, increase in the number of large emphysema and small airway disease clusters over time. Emphysema small clusters and large clusters of small airway disease were associated with lung function decline. A growth in small airway disease and formation of new emphysema regions over time had the greatest impact on lung function decline.</p>