The “Multimodal Data Science” team, made up of engineers, digital science researchers and doctors, aims to develop tools for diagnosis and therapeutic guidance from multiple data acquired as part of the treatment of patients with cardiac electrical disorders.
The research focuses on :
- the combined analysis of imaging (CT-scan, MRI, PET) and electrophysiology (invasive, non-invasive mapping or image-based simulation) data to improve knowledge about the structural substrate behind arrhythmias.
- the development of AI-based image processing tools for the segmentation of the cardiac structures and quantification of robust markers associated with the risk of arrhythmia.
- the development of digital modelling methods customised by image to simulate virtual electrophysiological tests in computing times compatible with clinical practice.
- the development of registration and visualisation interface tools for optimal instrument navigation in the patient's virtual heart during procedures.
- the clinical validation of these various tools for diagnosis, prognosis or real-time guided therapy in patients with cardiac electrical disorders.
This research has been made possible due to the development of a dedicated software environment, MUSIC software, consisting of image processing tools, interfaces for the combined analysis of imaging data and recorded or simulated electrophysiological signals, and import and export functions to 3D mapping platforms used for patient care.
It also requires a dedicated computing infrastructure to be set up to achieve performances compatible with clinical use.
This research also benefits from the MUSIC network, a network of 30 international expert centres (USA, Europe and Australia) which use our technology in the clinic, and so are involved in its validation and the creation of a unique imaging database about arrhythmia.