Towards automated fetal brain biometry reporting for 3-dimensional T2-weighted 0.55-3T magnetic resonance imaging at 20-40 weeks gestational age range

BackgroundThe detailed assessment of fetal brain maturation and development involves morphological evaluation, gyration analysis, and reliable biometric measurements. Manual measurements on conventional 2-D magnetic resonance imaging (MRI) are affected by fetal motion and there is no clear consensus regarding definitions for brain biometric parameters and anatomical landmark placements, making consistent reference plane and slice selection challenging. Automated biometry with 3-D slice-to-volume reconstruction (SVR) has the potential to improve the reliability of derived measurements, allowing precise quantification of fetal brain development. Previous published works have primarily focused on the technical feasibility of automated fetal brain biometry methods for T2-weighted (T2W) MRI. However, none have proposed solutions for automating the reporting of biometry results, which could enhance clinical utility and support real-time integration into routine clinical workflows. Furthermore, there is no consensus on a universal fetal biometry protocol for 3D fetal MRI. AimTo formalise biometry protocol for 3-D SVR T2W fetal brain MRI and to develop and validate a fully automated biometry reporting pipeline. Materials and methodsAutomated extraction of 13 routinely reported linear fetal biometry measurements using deep learning localisation of anatomical landmarks in 3-D reconstructed T2W brain images and presentation of the results in .html report with centile calculation. The automated biometry method was tested on 90 retrospective cases and the fully automated, end-to-end biometry reporting pipeline was prospectively evaluated on 111 cases across a wide range of gestational ages, field strengths and scanning parameters. We also generated normal centile ranges for 19 - 40 weaks GA range from 406 normal control datasets. ResultsThe retrospective evaluation showed excellent to good localisation for the majority of cases, with the maximum absolute difference between automated vs. manual measurement within 1-3mm range. For the prospective evaluation, more than 98% of all landmark placement were graded as acceptable for interpretation and measurements. The processing time of the pipeline was less than five minutes per case, with the measurements and centiles available at the time of reporting. Inspection of the automated landmark placement and computed biometrics took 1-3 minutes per case. The generated normative growth charts demonstrate high correlation with the trends in the previously reported works. ConclusionOur approach is the first to develop a pipeline which integrates automated fetal brain biometric measurements, centile calculations and normative growth charts into the clinical workflow, allowing fully automated biometry reporting for T2-weighted motion-corrected SVR MRI.