Background Pulmonary fibrosis is of critical importance in childhood interstitial lung disease (chILD), yet fibrosis prevalence, impact on the clinical progression, and survival have not been systematically evaluated. Methods Data were extracted from the chILD-EU register, a European prospective multicenter cohort study with centralized peer-review on patient inclusion and systematic scoring of computed tomography (CT) scans and lung biopsies. Pulmonary fibrosis was determined based on predefined criteria (fibrosis register) or criteria used in clinical trials (fibrosis trial). We calculated fibrosis rates of chILD entities, evaluated fibrosis criteria and assessed longitudinal pulmonary function testing and survival rates of children with or without fibrosis. Results 1,071 children diagnosed with chILD were included in the final analysis. The childhood prevalence of fibrosis was for 20.5% (220/1071) according to a single time point, register definition and 11.6% (62/534) according to the dual time point, trial definition. At the age when the children were able to perform pulmonary function tests, those with fibrosis had 15-20% worse predicted forced vital capacity (ppFVC), were older and diagnosed later. Throughout childhood, the disease trajectories assessed as decline in ppFVC and survival did not differ between children with or without pulmonary fibrosis or between the two fibrosis definitions. Overall, survival until the age of 20 years was about 70%. Conclusions This study assesses the prevalence, pulmonary function progression and survival of pulmonary fibrosis in chILD. The application of standardized criteria for pulmonary fibrosis enables identification of affected children among patients and may support early selection for anti-fibrotic therapies.
The Hidden Power of Normalization: Exponential Capacity Control in Deep Neural Networks
arXiv:2511.00958v1 Announce Type: cross Abstract: Normalization methods are fundamental components of modern deep neural networks (DNNs). Empirically, they are known to stabilize optimization dynamics and


