Project 4 - BirthView: A Digital Twin of Childbirth

Applying for Summer 2025

School/Institute: School of Computing Sciences

Introduction: BirthView is a computer-based simulation of human childbirth that runs in real-time on the GPU. Due to its capability of exhibiting the rotational fetal head movements, that are commonly observed during normal labour, it can be considered as a digital twin of childbirth (Lapeer, 2019). BirthView uses mesh models of a virtual baby and maternal pelvis that have been derived from publicly available image data that have been subsequently scaled to average size. A person-specific adaptation of BirthView would render it capable of predicting adverse childbirth outcomes before the actual event. This would allow clinicians to avoid emergency Caesarean Sections (CSs) and opt for the safer planned (or elective) CSs instead.

Objectives: The main objective of this Research Internship project is to create a proof-of-concept (PoC) person-specific simulator adapted from the BirthView simulator. To adapt the “average-sized” BirthView simulator to person-specific dimensions we can use (fetal) MRI scans of a near-term pregnant person. These MRI scans will be provided by collaborators from the University of Sheffield. The scans will be processed using medical image processing techniques including image segmentation and registration. Segmentation will extract the relevant fetal and maternal anatomy. The segmented images will then be reconstructed in 3D and then registered with the BirthView simulator to create a person-specific simulator – or digital twin – that matches the real pregnant person.

Specific tasks:

• Apply image processing techniques to derive fetal and maternal anatomy models from fetal MRI in 3D.

• Morph (and register) BirthView’s baby and maternal mesh models so they match the models derived from the fetal MRI scans.

Skills gained and essential prior knowledge: The underpinning physics and mathematics subjects of the simulator include continuum mechanics, contact mechanics, tensor calculus and the Finite Element Method (FEM). The underpinning computer science subjects include computer graphics and games, virtual reality (VR), software engineering, parallel programming on the GPU with OpenCL and C/C++ and (medical) image processing. It is not expected that the candidate matches the physics/maths skills but an interest in these subjects is expected. Though good programming and software engineering skills are essential as well as a solid understanding of computer graphics algorithms and procedures.

Reference: Rudy Lapeer, Zelimkhan Gerikhanov, Said‑Magomed Sadulaev, Vilius Audinis, Roger Rowland, Kenda Crozier and Edward Morris. 2019. A computer‑based simulation of childbirth using the partial Dirichlet–Neumann contact method with total Lagrangian explicit dynamics on the GPU. Biomechanics and Modeling in Mechanobiology, https://doi.org/10.1007/s10237-018-01109-x , 18:681–700.

The Lapeer et al, paper above can be accessed via the link where a video of the BirthView simulator in action can be viewed under the “Electronic supplementary material” section.