Purpose or Case Report: Fetal MRI is increasingly used in congenital diaphragmatic hernia (CDH) cases for enhanced anatomical imaging and prognostic assessment. Radiomics allows extraction of complex quantitative data that may enhance prediction capabilities beyond standard lung volume measurements. This investigation sought to determine which radiomic characteristics distinguish normal fetal lungs from those affected by CDH.
Methods & Materials: We analyzed retrospective fetal MRI data from fetuses diagnosed with left-sided type C or D CDH between 20- and 26-weeks gestational age (GA). The control cohort included fetuses who underwent MRI for unrelated reasons and had imaging-confirmed normal lung development. Using ITK-Snap software, we manually outlined both normal lungs bilaterally and the right contralateral lung in CDH cases on coronal T2-weighted single-shot sequences. PyRadiomics was employed to extract radiomic characteristics. Statistical comparison of radiomic measurements used p <0.05 as the threshold for significance.
Results: The analysis included 26 fetal MRI studies: 11 (42%) with CDH and 15 (58%) without CDH. Average gestational age at imaging was comparable between cohorts—24.0 weeks (95% CI 22.6–25.4) in the CDH group versus 26.5 weeks (95% CI 23.9–29.2) in the control group (p=0.114). A total of 107 radiomic characteristics were analyzed across seven categories: 18 First Order Statistics features, 14 Shape features, 24 Gray Level Co-occurrence Matrix (GLCM) features, 14 Gray Level Dependence Matrix (GLDM) features, 16 Gray Level Run Length Matrix (GLRLM) features, 16 Gray Level Size Zone Matrix (GLSZM) features, and 5 Neighboring Gray Tone Difference Matrix (NGTDM) features. Between normal right and left lungs, just one feature demonstrated statistical significance (p=0.025). When comparing normal right lungs to contralateral right lungs in CDH cases, 46 radiomic features exhibited statistically significant differences.
Conclusions: Our comparison of radiomic characteristics between normal lungs and contralateral CDH-affected lungs revealed 46 significantly different variables. Future work will examine associations between these radiomic features and clinical outcomes to enhance prenatal evaluation and prognostic accuracy for CDH patients.