Objective: Lung deformation beyond of physiological capacity is associated with cell death and inflammation. Lung strain has been estimated as a global strain, but uneven strain distribution may lead to regional stress concentrations and lung damage. Local lung inflammation can be estimated using PET imaging of [18F]fluoro-2-deoxy-D-glucose. We hypothesized that local lung deformation correlates well with local inflammation. The aim of this study was to assess local tidal deformations by using a new mathematical model of finite-elements to analyze CT images, and to correlate them with local inflammation in a porcine experimental model of early acute respiratory distress syndrome.
Design: Retrospective images analysis, laboratory investigation.
Setting: University animal research laboratory.
Subjects: Seven piglets submitted to experimental ventilator-induced lung injury and five healthy ventilated controls.
Intervention: Lung injury was induced by repeated lung lavages and 210 minutes of injurious mechanical ventilation using low positive end-expiratory pressure and high inspiratory pressures. All animals were subsequently studied with dynamic PET imaging of [18F]fluoro-2-deoxy-D-glucose. CT scans were acquired at end expiration and end inspiration. Then maps of deformation were constructed and regional deformation was estimated. We divided the lung parenchyma in 10 horizontal ROIs, and correlations of local volumetric strain and [18F]fluoro-2-deoxy-D-glucose uptake were analyzed in each ROI.
Measurements and Main Results: The deformation maps showed a heterogeneous distribution with a greater concentration in the intermediate gravitational regions. We found a strong correlation between local strain and inflammation (R2 > 0.5) for the whole lung, when we eliminate the 3/10 dorsal ROIs R2 increased until>0.8.
Conclusion: the present findings suggest that the greater local stretches were mainly concentrated in the intermediate gravitational zones of injured heterogeneous lungs. Additionally, local lung deformations correlated well with local inflammation in this experimental model of VILI. And the new proposed image-based estimation of regional volumetric strain based on finite element interpolations has the potential to give new insights of local pathogenic mechanisms of VILI and how best design protective-ventilations strategies.