BPD-Neo: An MRI Dataset for Lung-Trachea Segmentation with Clinical Data for Neonatal Bronchopulmonary Dysplasia

  • 2025-07-17 15:06:30
  • Rachit Saluja, Arzu Kovanlikaya, Candace Chien, Lauren Kathryn Blatt, Jeffrey M. Perlman, Stefan Worgall, Mert R. Sabuncu, Jonathan P. Dyke
  • 0

Abstract

Bronchopulmonary dysplasia (BPD) is a common complication among pretermneonates, with portable X-ray imaging serving as the standard diagnosticmodality in neonatal intensive care units (NICUs). However, lung magneticresonance imaging (MRI) offers a non-invasive alternative that avoids sedationand radiation while providing detailed insights into the underlying mechanismsof BPD. Leveraging high-resolution 3D MRI data, advanced image processing andsemantic segmentation algorithms can be developed to assist clinicians inidentifying the etiology of BPD. In this dataset, we present MRI scans pairedwith corresponding semantic segmentations of the lungs and trachea for 40neonates, the majority of whom are diagnosed with BPD. The imaging data consistof free-breathing 3D stack-of-stars radial gradient echo acquisitions, known asthe StarVIBE series. Additionally, we provide comprehensive clinical data andbaseline segmentation models, validated against clinical assessments, tosupport further research and development in neonatal lung imaging.

 

Quick Read (beta)

loading the full paper ...