Coronary artery disease (CAD) is the leading cause of mortality in the world, resulting in over 9 million deaths annually. Coronary Computed Tomography Angiography (CCTA) is an imaging method routinely used for CAD diagnosis, as it provides non-invasive access to the coronary anatomy and enables stenosis assessment.
In current clinical practice, stenosis assessment is typically performed manually by the radiologist. As this is a laborious and error-prone task, there is a significant clinical need to automate it. To date, relevant AI-based algorithms are developed primarily for conventional CTs, whereas the use of scans from advanced CT technologies such as spectral and photon counting acquisitions remains largely under-explored. Since these types of scans include significantly more information than conventional ones, this project aims to investigate how this novel information can be leveraged to develop robust and accurate AI-based methods for automatic detection and grading of coronary artery stenosis across all types of CCTAs.
This project is part of the Horizon Europe-funded COMBINE-CT consortium.