Longitudinal Dark-Field Mouse CT Dataset

The LDM-CT (Longitudinal Dark-Field Mouse CT) dataset provides the first publicly available in-vivo longitudinal X-ray dark-field computed tomography data of pulmonary disease progression in mice. The dataset contains multi-time-point attenuation (\u00b5) and dark-field (\u00b5d) volumes acquired over 12 weeks in healthy, inflammatory-injury, and COPD mouse models, together with corresponding 3D lung segmentation masks. All scans were obtained using a laboratory-based Talbot\u2013Lau dark-field CT system under a dose-optimized and repeatable protocol designed for longitudinal preclinical studies.The longitudinal design enables time-resolved assessment of microscopic structural alterations in the lung parenchyma, offering substantially higher sensitivity than conventional attenuation-based CT. This dataset supports a wide range of biomedical imaging research topics, including quantitative pulmonary imaging, disease progression modeling, biomarker discovery, dark-field reconstruction, motion-robust algorithms, and longitudinal deep-learning\u2013based segmentation.All data are anonymized and provided in open format, along with trained segmentation model weights, to ensure transparency and reproducibility. The LDM-CT dataset aims to facilitate future preclinical and translational research in pulmonary imaging, and to serve as a benchmark resource for the development of advanced longitudinal imaging and computational analysis methods.