Caution - Investigational Device
The fetal MRI Post-Processing platform is only for research purposes and is not yet CE marked or FDA cleared.

Fetal MRI Post-Processing
Enhances fetal MR Imaging with Slice-to-Volume Reconstruction
Transform motion-corrupted 2D MR image stacks into motion-compensated 3D volumes using the advanced slice-to-volume reconstruction pipeline developed by King’s College London.
Looking to improve your fetal MR imaging? Discover our new Post-Processing platform!
Our Platform
Fetal MRI plays a crucial role in clinical settings, offering valuable insights for diagnosing and supporting fetal health. While post-processing of MR images is well-established across all patient groups, there remains a significant gap when it comes to fetal imaging – no dedicated software solution exists.
To address this unmet need, we have developed a fetal MRI post-processing platform. Our vision is to make post-processing of fetal MR images as seamless and accessible as it is for pediatric and adult imaging. The platform already integrates a slice-to-volume reconstruction algorithm, developed by King’s College London (auto-proc-SVRTK package), to provide a single volumetric motion-corrected dataset.
It is readily available for the brain, thorax, and body for download and can be easily integrated into your preferred software for further analysis or annotation. Furthermore, we have incorporated the automated brain and lung volume segmentation algorithm developed by Uus et al..
Fetal Motion Correction

Brain

Thorax

Body
Turning 2D motion-corrupted images in a
comprehensive motion-corrected 3D volume
through slice-to-volume reconstruction.
Volume Segmentation

Brain

Thorax
Automated segmentation and multi-regional
parcellation with no manual work necessary.

Region of Interest: Brain
Challenge: Image acquisition in specific planes can be complex and may require repeated scans for accurate results. Extensive training of MR personnel is necessary.
Solution: A comprehensive brain volume offers re-orientation and accessibility in any imaging plane, significantly improving diagnostic capabilities. Additionally, automated brain segmentation and parcellation provide enhanced insights and efficiency.
.png)

Region of Interest: Thorax
Challenge: Manual segmentation of the lungs in motion-corrupted 2D slices is both time-intensive and prone to errors due to motion artifacts, often resulting in incorrect volume assessments.
Solution: A comprehensive 3D volume reconstruction facilitates tracking of anatomical structures. This is further enhanced by automated multi-regional lung segmentation and accurate volume calculation, streamlining the diagnostic process.
.png)
Related Publications
2012
Medical Image Analysis
M. Kuklisova-Murgasovaet al.
Uus et al.
IEEE Transactions on Medical Imaging
2020
Uus et al.
PIPPI MICCAI Workshop
2024