Research Associate AI and fetal MRI
Fetal MRI, with its ability to visualize and quantify structure and function of all fetal organs as well as the surrounding womb, allows unique and fascinating insights into early human development. However, it poses significant engineering challenges:
Fetal motion, maternal breathing and subclinical contractions are normal during fetal MRI scans, but hamper the image quality especially for high resolution scans. On the other hand, these events contain important information, crucial to understand eg how a fetus a high risk will react to labor contractions. These are typically identified after the respective scans leading to repeated scans and missed information.
AI techniques detecting fetal motion, breathing and other dynamic events such as contractions in real-time during the scan allow to redesign fetal MRI by immediately adapting the acquisition to the ongoing fetal life.
Automatic detection of the fetal state, his/her location, the gestational age and any areas of concern (such as reduced amniotic fluid) followed by the adjustments of the scan to the ongoing, detected fetal life will contribute to changing the paradigm of fetal imaging and thus advance our knowledge and capacities in this challenging area.
This post is part of a UKRI fellowship which aims to develop an entirely self-driving MRI scan from the engineering challenges described to clinical translation. The post will use the MR infrastructure provided by the School of Biomedical Engineering and Imaging Sciences and will be based at St Thomas’ Hospital. This project and team is embedded in the Centre for Medical Engineering – an outstanding, interdisciplinary and enriching research environment.
Collaboration with external academic and industrial partners including Imperial College, NYU, Siemens Healthineers and north Hamburg will provide ample opportunities.
Successful candidates will have knowledge in computational AI techniques and be interested in interdisciplinary work.
Objectives: The main objective of this post is to develop a novel real-time guidance for fetal MRI scans.
The aims of this post are (i) to use AI techniques in real-time to detect the fetal sleep and motion state during the MRI examination and (ii) to use this information in real-time to guide the MRI scan.
The post holder will be primarily responsible for adapting and developing real-time AI techniques to obtain spatially-localized information from the entire uterus together with the project partners. One or multiple research stays with collaboration partners are therefore planned.