Background
Congenital fetal body anomalies are among the most challenging conditions to diagnose and monitor during pregnancy, often associated with complex syndromes and neonatal complications. Fetal MRI provides detailed anatomical information but currently depends on manual interpretation, which can be subjective and time-consuming. This PhD project focuses on developing automated methods to detect and characterise fetal body anomalies from 3D MRI scans, improving diagnostic consistency and supporting clinical decision-making.
The research will build on existing deep-learning pipelines developed at King’s College London and St Thomas’ Hospital, which already enable motion-corrected reconstruction and organ-level segmentation in fetal MRI. By extending these capabilities toward automated anomaly identification and quantification, the project aims to create clinically usable tools that bridge research innovation with real-world perinatal care.
Novelty & Importance
This project introduces a clinically focused AI framework that combines image segmentation, anomaly detection, and simulation-based modelling to objectively analyse fetal body development. Unlike previous work centred on brain imaging, this research targets the whole fetal body, capturing a broader range of structural abnormalities across gestational ages.
The system will be designed for clinical translation, generating interpretable metrics and structured outputs that can be integrated into radiological workflows. By reducing diagnostic variability and providing quantitative support for multidisciplinary teams, the work has the potential to transform prenatal diagnosis and counselling.
Aims & Objectives
• Develop deep learning models for automatic segmentation and volumetric quantification of fetal body MRI.
• Implement algorithms for detection and characterisation of anatomical anomalies.
• Integrate predictive modelling to simulate growth and structural deviation.
• Validate performance on large clinical datasets, ensuring translational readiness and clinical usability.
References
Uus et al. (2024) Automated body organ segmentation, volumetry and population-averaged atlas for 3D motion-corrected T2-weighted fetal body MR. Scientific Reports. https://doi.org/10.1038/s41598-024-57087-x
Davidson et al. Fetal body MRI and its application to fetal and neonatal treatment: an illustrative review. Lancet Child & Adolescent Health (2021). https://doi.org/10.1016/S2352-4642(20)30313-8
Uus et al. (2025) Scanner-based real-time three-dimensional brain + body slice-to-volume reconstruction for T2-weighted 0.55-T low-field fetal magnetic resonance imaging. Pediatric Radiology. Pediatric Radiology https://doi.org/10.1007/s00247-025-06165-x