! POSITION FILLED !

Download the internship proposal

Type of internship

Master’s 2 Internship or Final Year of Engineering School

Internship date or duration

6-month internship

Context

Bicuspid aortic valve (BAV) is the most common congenital heart defect, affecting approximately 0.5% of the general population. Its most common complication is aortic dilation, known as aortopathy. The use of 4D flow MRI has revealed disorganized blood flow in the ascending aorta, affecting the aortic wall and potentially inducing further dilation, which can progress to aneurysm. The potential of this imaging modality for monitoring bicuspid patients is only partially explored in the literature. Indeed, when 4D flow is combined with artificial intelligence, it enables 3D+t segmentation of the aorta and, consequently, provides anatomical, functional, and hemodynamic parameters (volumes, 3D and regional deformation, pressure gradients) that have been scarcely explored to date. Accordingly, the goal of this project is to extract these innovative parameters in bicuspid patients who underwent MRI at baseline (T0) and at 2 years, to assess the progression of their disease and better understand the underlying mechanisms. To achieve this objective, the candidate will have access to data from the BAO MRI protocol conducted at Hospital European Georges Pompidou in Paris, as well as to the clinical expertise of cardiovascular radiology department.

Objective

To extract innovative aortic imaging parameters by combining artificial intelligence and 4D flow MRI to study the progression of aortic disease in patients with bicuspid aortic valve.

Tasks

This internship will include several steps:
1. Literature review on bicuspid aortic valve imaging
2. Preparation of 4D flow images and quality control using well-established criteria
3. Use of the AI software developed by a Ph.D. student in the team to segment the aorta from these images
4. Quality control of segmentations
5. Extraction of anatomical, functional, and hemodynamic parameters
6. Performing tasks 3, 4 and 5 blindly on data acquired at T0 and T24 (months)
7. Statistical analysis of the evolution of extracted imaging indices and the relationship between anatomical, functional, and hemodynamic components.

Skills

1. Reading and synthesizing scientific documents in English
2. Image processing
3. Basic statistical analyses
4. Programming in Python/Matlab

Compensation

Internship grant

Contact

Nadjia Kachenoura
nadjia.kachenoura[at]inserm.fr