VIROLAAI project

Atrial Fibrillation (AF) is the most common cardiac arrhythmia diagnosed in clinical practice. Problems related with AF increase together with the increment of the world population and in life expectancy. The 90% of the thrombus related with this arrhythmia is created in the left atrial appendage (LAA). One way to reduce this thrombus risk is to give drugs (e.g. anticoagulants) to AF patients. Unfortunately, many of them have problems related with bleeding risk. In those cases a left atrial appendage occluder (LAAO) is implanted to prevent the passage of blood flow in this structure and then reduce the risk of thrombus generation. Morphological information is used to select the best device for a given patient and which trajectory clinicians should take to bring the selected device to the optimal position for a good implantation. These clinical decisions are nowadays highly dependent on the experience of the clinician since there is not a computational platform to provide objective recommendations available. The main goal of the VIROLAAI project is to combine and adapt some of the most advanced computational technologies, such as machine learning, biophysical models and virtual/augmented reality interfaces, to support clinical decisions on LAAO interventions. The feasibility of 4D-flow magnetic resonance imaging to complement standard echocardiography acquisitions to characterize LA/LAA blood flow patterns will be studied. Subsequently, machine-learning techniques will be applied on morphological and haemodynamics indices to cluster patients in different groups and relate to the risk of thrombus formation. Finally, VIROLAAI will develop virtual/augmented and interactive interfaces for a multi-user joint exploration of complex multi- modal data in a clinical environment, targeting educational, patient-doctor communication and treatment planning use cases.

Job description

We propose a research assistant position directly associated to the project to investigate, develop, validate and translate computational tools to analyse and simulate blood flow patterns in the left atria. The work of will be fully complementary to other developments in VIROLAAI. The main goal will be to process haemodynamics imaging data, specially 4D-flow MRI, for the validation of simulations derived from classical Computational Fluid Dynamics (CFD) and data- driven (e.g. deep learning) methods. The research will be divided into the following stages: 1) processing 4D-flow MRI data to derive haemodynamics indices of the LA/LAA in patients with atrial fibrillation; 2) comparison of DL-based blood flow simulations with classical CFD ones and with patient-specific 4D-flow MRI data. If realistic enough, DL-based simulations will be integrated with the virtual reality VIDAA interfaces for a fast user interaction and visualization of in silico simulations. This PhD will be in close collaboration with technicians from General Electric Medical Systems as well as clinicians from Hospital Clínic de Barcelona and Hospital Vall d’Hebron for the acquisition of 4D-flow MRI data and the processing of this data.

Workplace

The scientific supervisor of this work is Oscar Camara, Associate Professor at the Department of Information and Communication Technologies of the Universitat Pompeu Fabra (DTIC-UPF, https://www.upf.edu/web/etic), and leader of the PhySense research group. The DTIC at the UPF is the first Spanish ICT department that has been awarded with the María de Maeztu grant (excellence in science and innovation accreditation, 2016-2019) on data-driven knowledge extraction (https://www.upf.edu/web/mdm-dtic), and the Spanish university department with the largest number of ERC grants (15, including 6 Advanced ERC Grants). PhySense was recognized as an Emerging Research Group by the Government of Catalonia in 2014 and it is currently composed of 19 members. In September 2016, the group was one of the founding members of the BCN-MedTech unit (https://bcn-medtech.upf.edu/), a Research Unit at UPF that holds more than 200 I+D projects, 95 external collaborations, 1000 high-impact publications, 19 patents and 61 PhD Thesis. It recently (2018) obtained the TECNIO certification from the Catalan Government, given to research centres with proven record of technology transfer. This work will require a constant presence at Hospital Clínic de Barcelona to closely work with technicians of General Electric Medical Systems and cardiologists with expertise on LAAO interventions.

Profile of the candidate

We are looking for a highly motivated junior researcher with an undergraduate degree and a research-oriented MSc in Biomedical Engineering, Physics, Computational Science, or related disciplines, and a research-oriented MSc. Experience in computational modelling, especially in computational fluid dynamics simulations and deep-learning methods would be of importance. Candidates must have excellent teamwork and communication skills and be enthusiastic about collaborating with a diverse range of partners. Interest in clinical translation is essential since meetings with clinicians will regularly take place. Female applicants are explicitly encouraged to apply and will be treated preferentially whenever they are equally qualified as other male candidates.

Deadline and contact information

Applicants should send a curriculum vitae and a motivation letter describing their research interests to Oscar Camara <[email protected]>. Deadline: open until filled.