From the desk of CAP-AI data scientist Andrew Bard:

Project title: Prediction of aortic changes from surveillance magnetic resonance imaging in congenital ascending aortic aneurysm

The CAP-AI programme enables a wide range of projects, and facilitates collaboration across Barts Health clinicians, QMUL academics and industry in a unique and free-flowing manner. Andrew Bard is a data scientist and CAP-AI fellow, with a diverse portfolio of projects.

Pericardial fat, a buildup of fatty tissue in and around the heart, has been implicated in a wide range of cardiovascular problems. Most recently, people have hypothesised that its abundance may predict COVID-19 outcomes. However, more research is needed to test this hypothesis, and to examine more precisely what these fat deposits can tell us. They have previously been examined using CT scans, which unfortunately requires exposure to harmful X-rays. It would be safer to quantify them using MRI, particularly as cardiac MRI is routine in a number of different clinical scenarios.

Working with Dr Zahra Raisi-Estabragh, a clinical PhD fellow, and Professor Steffen Petersen, of QMUL, Andrew has developed a machine learning algorithm to perform this analysis. This algorithm can not only draw contours around these fat deposits on cardiac MRI, but also estimates how accurate these contours are. When data from a large number of people is analysed, quality control becomes particularly important as it is not practical for researchers to check every contour.

This project has been supported by Motilent, a medtech SME spun out of University College London. Motilent’s flagship software product, GIQuant, quantifies gut motility. Another project, in which Andrew works with Professor Steve Greenwald of QMUL, seeks to examine whether this same software can be repurposed to give meaningful information about the properties of blood vessels. For example, it is currently being used to examine the biomechanical properties of the aorta, and ultimately to assess whether such information can predict how aortic aneurysms change over time.

Motilent is dedicated to producing the best technologies for researchers to use to further our understanding of the digestive system in health and disease. Whether it’s a small retrospective study or a large multicentre trial in partnership with a CRO, Motilent can provide you with the technology and the support to help you produce high quality research.