Computed Tomography Registration-Derived Regional Ventilation Indices Compared to Global Lung Function Parameters in Patients With COPD

CT registration-derived indices provide data on regional lung functional changes in COPD. However, because unlike spirometry which involves dynamic maximal breathing maneuvers, CT-based functional parameters are assessed between two static breath-holds, it is not clear how regional and global lung function parameters relate to each other. We assessed the relationship between CT-density change (dHU), specific volume change (dsV), and regional lung tissue deformation (J) with global spirometric and plethysmographic parameters, gas exchange, exercise capacity, dyspnoea, and disease stage in a prospective cohort study in 102 COPD patients. There were positive correlations of dHU, dsV, and J with spirometric variables, DLCO and gas exchange, 6-min walking distance, and negative correlations with plethysmographic lung volumes and indices of trapping and lung distension as well as GOLD stage. Stepwise regression identified FEV1/FVC (standardized β = 0.429, p < 0.0001), RV/TLC (β = −0.37, p < 0.0001), and BMI (β = 0.27, p=<0.001) as the strongest predictors of CT intensity-based metrics dHU, with similar findings for dsV, while FEV1/FVC (β = 0.32, p=<0.001) and RV/TLC (β = −0.48, p=<0.0001) were identified as those for J. These data suggest that regional lung function is related to two major pathophysiological processes involved in global lung function deterioration in COPD: chronic airflow obstruction and gas trapping, with an additional contribution of nutritional status, which in turn determines respiratory muscle strength. Our data confirm previous findings in the literature, suggesting the potential of CT image-based regional lung function metrics as the biomarkers of disease severity and provide mechanistic insight into the interpretation of regional lung function indices in patients with COPD.

3D Slicer (http://www.slicer.org). This consists in a region growing algorithm (Adams and Bischof, 1994) based on user-provided seeds, followed by a morphological closing and smoothing of the obtained binary masks. As a prerequisite for the utilized registration algorithm, the masked volumes were then processed to have isotropic voxels and identical dimensions. The spacing between the isotropic voxels was 0.625 mm and the in-plane matrix size was kept constant as 512×512 for all the images. After an initial rigid alignment of the two volumes, a deformable image registration was performed by the deedsBCV algorithm [https://github.com/mattiaspaul/deedsBCV] (Heinrich et al., 2013a). Deformable image registration consists in warping one image (moving image, ) to morphologically match the other image (fixed image, ). The algorithm uses a dense displacement sampling strategy (deeds) to cope with large displacements of small structures. Self-similarity context (SSC) descriptors (Heinrich et al., 2013b) used to calculate the similarity measure (data-attachment term) ensure robustness against image-intensity changes, and pair-wise regularization term inferred on a minimum-spanning-tree model controls the smoothness of the estimated transformation Φ. Here the inspiration image was warped (moving image) to match the expiratory image (fixed image). The resulting deformation field is a matrix containing the displacement vectors that align the corresponding voxels between the fixed and moving images. A median filter with a radius of 2 voxels was applied to the images to remove the effect of acquisition noise.
The following outcome measures were computed: 1) the inspiratory -expiratory x-ray density change in Hounsfield units (dHU) between the fixed (expiration) and warped (inspiration) images, →^, →^, calculated as: where ^ stands for the filtered moving (inspiratory) image warped by the estimated transformation Φ, and ^ for the filtered fixed (expiratory) image. Voxels within the interval -1100 to -500 HU were included in the analysis, thus excluding dense structures such as blood vessels.
2) The specific volume change between fixed and warped images, as defined by Simon et al. (Simon, 2000): Where dsV is specific volume change; dV is the local volume change upon inspiration; Vexp is the local gas volume at end-expiration; dHU the attenuation change (Ding et al., 2012).
3) The determinant of the Jacobian matrix was calculated from the non-linear transformation function: Φ (Reinhardt et al., 2008). This parameter referred to as the Jacobian (J), expresses the local relative volume change between inspiration and expiration (Fleming, 2012). The Jacobian is independent of the image intensity values and therefore of dHU. A Jacobian value greater than 1 means local expansion whereas a J<1 indicates local contraction and J=1 indicates no volume change.
Scattering in dHU, dsV and J was expressed as the quantile variation coefficient (QVC) defined as interquartile range/median rather than coefficient of variation, given the non-normal distribution of these parameters. Supplementary Tables   Supplemental Table 1. Correlation between the kurtosis of regional lung function parameter distribution and demographic, global lung function, gas exchange exercise capacity, dyspnoea and GOLD classification.

Supplementary Figures
Supplemental Figure 1. Dorsal and ventral views of the 3D rendering of the Jacobian determinant in two GOLD stage 1 and two GOLD stage 4 patients. Note the low values of the Jacobian in the dependent dorsal regions in the GOLD 1 patients, while the dorsal -ventral regional differences were reduced in the GOLD 4 patients.