Assessing Placental Structure and Function by Unified Fluid Mechanical Modelling and in-vivo MRI

Lead Research Organisation: University College London
Department Name: Computer Science

Abstract

Our plan is to reduce poor pregnancy outcomes by creating a revolutionary way to assess and analyse the pregnancy environment. Specifically, we will establish specialised MRI (which is non-invasive and emits no ionising radiation) techniques focusing on women with high-risk pregnancies, which will enable personalised care and individualised treatments for mothers and babies.

The placenta (or afterbirth) is vital for a successful pregnancy, and most major pregnancy complications are associated with placental problems. Ultrasound is currently the main way to monitor pregnancies, however it has limitations: the whole placenta can't be seen at once, findings vary depending on who is performing it, and it cannot tell us how the placenta is functioning, only what it looks like. Ultrasound images can also be poor due to the baby's position and factors such as the mother's weight. We will use key features of MRI to enable a more detailed assessment of placental health - information on how the placenta is working and a more detailed picture of its structure. Although we can see clear changes in placental MRI scans for a number of important pregnancy complications, such as fetal growth restriction and pre-eclampsia, the specific changes in the placenta that affect the MRI signal are yet to be determined, not least because the relationship between the placenta's structure and how well it functions is complex. Techniques for detecting subtle placental changes would enable early and accurate diagnosis of pregnancy complications. In this project, we will develop such a technique by combining MRI with computational modelling to give valuable insight into how well the placenta is working.

First of all, we will build highly detailed computational models of whole placentas, incorporating fine structural details such as tissue and blood vessel dimensions. Next we will develop tools that can predict blood flow and oxygen levels in these computational placentas. We then use simulations to predict what MRI scans would look like for our computational placentas. Finally, using machine learning techniques, we will instruct an algorithm to learn from thousands of simulations on computational placentas, in order to deliver an imaging tool that, given a placental MRI scan as input, can show the structure, blood flow, and oxygen levels of the placenta.

By using this MRI-based tool to investigate pregnancies with complications, we will test its ability to identify cases where the placenta is not working properly. We will do this by comparing our tool with a post-pregnancy assessment of the delivered placenta by a specialised pathologist (many placental conditions, such as abnormal structure or poor blood supply, can currently only be diagnosed in this way). For multiple pregnant participants, we will use our MRI-based tool during the pregnancy to calculate values relating to placental structure and blood flow, then undertake an after delivery assessment of the placenta. This will allow us to test if our MRI-based tool can tell if the placenta is not working during pregnancy without having to wait for an after delivery assessment by a specialist pathologist. We will first explore ways of determining if problems with the placenta are caused by issues with the supply of maternal or fetal blood. This would be extremely important in clinical practice and in the future may allow focused treatment of the mother to reduce poor outcomes.

Showing differences between maternal and fetal blood supply problems is just one example of a diagnosis that cannot be during the pregnancy at present, and we use it here to demonstrate the potential of our model-driven placenta imaging tools in a directly impactful application. Further potential conditions that could be identified with the developed tools include damage to fetal villous trees (this is the detailed structure of the placenta), and levels of oxygen supply.

Publications

10 25 50
 
Description Work in this grant has developed novel MRI techniques for probing the health of the placenta. Recent work (Cromb et al preprint: 10.21203/rs.3.rs-3873412/v1) verifies that these techniques can reveal how placental development is affected by various pathologies or conditions such as congenital heart disease.
Exploitation Route Early screening for malformation of placenta
Sectors Healthcare

URL https://pubmed.ncbi.nlm.nih.gov/38343847/
 
Description CMICHACKS Project co-leaders, UCL 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Postgraduate students
Results and Impact Coordinated with other co-leaders from the Centre for Medical Image Computing to design Hackathon project focusing on knowledge transfer between imaging and modelling domains of lung and placenta
Year(s) Of Engagement Activity 2022
 
Description Invited public lecture, Institute of Physics and Engineering in Medicine, Hiding in Plain Sight: The Unseen Impact of Engineering in Healthcare 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact Invited talk for flagship event through IPEM, reaching ~150 people from diverse backgrounds - school children, healthcare professionals, engineering.
Year(s) Of Engagement Activity 2023
 
Description Selected talk at the BioMedEng22 Conference, UCL, London, UK 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Postgraduate students
Results and Impact Talk titled "A flexible generative algorithm for growing placental vascular trees", which served to present novel work developed so far as part of the research grant. This talk was included in a session with 40-60 attendees, and sparked questions and discussion afterwards.
Year(s) Of Engagement Activity 2022