# Automated cervix biometry, volumetry and normative models for 3D motion-corrected T2-weighted 0.55-3T fetal MRI during 2nd and 3rd trimesters

## TL;DR

This paper introduces an automated method for analyzing fetal cervix MRI scans, enabling faster and more consistent measurements during pregnancy.

## Contribution

The first deep learning pipeline for automated multi-layer segmentation and biometry of the pregnant cervix in 3D T2-weighted MRI.

## Key findings

- The automated pipeline showed good performance compared to manual measurements in 20 datasets.
- Inlet diameter and length correlated strongest with gestational age, reflecting cervical remodeling.
- 3D population-averaged MRI atlases of the cervix were generated and made publicly available.

## Abstract

Fetal MRI provides superior tissue contrast and true 3D spatial information however there is only a limited number of MRI studies investigating cervix during pregnancy1–3. Furthermore, there are no clearly formalised protocols or automated methods for MRI cervical measurements. This work introduces the first deep learning pipeline for automated multi-layer segmentation and biometry for 3D T2w reconstructed images of the pregnant cervix. Evaluation on 20 datasets from 0.55T and 3T acquisitions showed good performance in comparison to manual measurements. This solution could potentially minimise the need for manual editing, significantly reduce analysis time and address inter- and intra-observer bias. Next, we used the pipeline to process 270 normal term cases from 16 to 40 weeks gestational age (GA) range. The inlet diameter and length showed the strongest correlation with GA which is in agreement with the gradual remodeling and softening of the cervix prior to birth. We also generated 3D population-averaged MRI atlases of the cervix that are publicly available online.