Chronic Obstructive Pulmonary Disease (COPD) is a major cause of morbidity and mortality. Despite declines in smoking, mortality from COPD continues to increase and is now the 3rd leading cause of death in the US. The chronic airflow limitation of COPD is caused by a mixture of small airway disease and parenchymal destruction (emphysema). Recent studies have suggested a central role of small airway destruction in the pathogenesis of COPD. This evidence has sparked the interest in in-vivo assessment of small airway disease overall at the early onset of the disease. Early identification of small airway disease could lead to better patient diagnosis, early therapeutic intervention and provide more sensitive markers to elucidate the pathogenesis of the disease and its biomolecular basis that could inform much-needed drug discovery. Computed Tomography (CT) is an imaging modality that has proven to be effective in the quantification of parenchymal destruction. However, the imaging resolution required to obtain direct measures from small airways is beyond the limits of CT scans. A recent technique called parametric response mapping (PRM) proposes to distinguish gas trapping due to small airway disease from emphysema by matching inspiratory and expiratory CT scans and applying density thresholds to distinguish functional small airway disease (FSAD) and emphysema. Despite its success, the PRM shows some limitations that are precluding the accuracy, robustness and interpretation of its results in early disease: The CT density values highly depend on acquisition parameters (dosage, reconstruction kernel, changes in body size) that introduce subject- and scanner-dependent confounders. Although clinical trials use well defined acquisition protocols and phantom-calibrated acquisitions, the biases and noise patterns still are subject-dependent. In particular, many studies using PRM employ inspiratory and expiratory images that are obtained at different dose levels. This project will take full advantage of our most recent developments in image-driven statistical characterization of tissues to reduce the harmful effects of the main factors affecting PRM. The harmonization of CT scans in a statistical framework will enable robust PRM metrics in cross-sectional and longitudinal studies. The statistical characterization will also lead to define adaptive thresholds to detect the emphysema and FSAD minimizing type I and II error trade-off. We will validate the robustness of harmonized PRM metrics in multiparametric acquisitions and study its clinical relevance by studying associations with lung function. Our preliminary data shows that we can obtain harmonized images that minimize the scanner and subject- dependent confounders. Our tissue characterization in CT images also has proved its suitability to provide a statistical framework to define robust adaptive thresholds. Together, the research proposed in the aims of this award will take full advantage of the comprehensive dataset available through the COPDGene study.