French
English
Nathan Painchaud completed his PhD under a joint supervision agreement between INSA Lyon and the Université de Sherbrooke (UdS, Canada), under the supervision of Professor Olivier Bernard (INSA Lyon), Professor Nicolas Duchateau (Université Lyon 1), and Professor Pierre-Marc Jodoin (Department of Computer Science, Faculty of Science, Université de Sherbrooke).
He is the recipient of the Best PhD Thesis Award awarded in the framework of a Franco–Quebec joint supervision agreement for his dissertation entitled Deep Manifold Learning for Improved Characterization of Arterial Hypertension in Echocardiographic Imaging.
Arterial hypertension affects 1.5 billion people worldwide. The assessment of this complex cardiovascular disease is often challenging due to the wide variability of symptoms and the lack of fast and accurate analysis tools. As the field of artificial intelligence continues to rapidly evolve, numerous models have been developed to assist clinicians in interpreting cardiac ultrasound images. However, these techniques can yield inconsistent results in many cases.
This is where Nathan Painchaud’s work makes a significant contribution. His PhD research proposes solutions to errors commonly made by neural networks by introducing constraints based on expected image characteristics—an approach that is unprecedented in the field. His work has also enabled the study of hypertension-related manifestations with unprecedented precision, revealing potential early warning signs of the disease.