Controlling unruptured intracranial aneurysms using Fluid-Structure interaction and Deep Reinforcement Learning – siCURE

Recent advances in the development of Deep Reinforcement Learning (DRL) algorithms have led to the advent of deep neural networks, powerful tools capable of leveraging the ever-increasing volume of numerical and experimental data generated for research and engineering purposes into novel insight and actionable information. DRL has consistently, albeit slowly, spread to fluid mechanics, and has led to a handful of seminal, high-potential publications touting high potential for flow control and optimization purposes.

While the approach is expected to pave the way for future progress in the optimal control of computational fluid dynamics (CFD) systems, it has never been applied to hemodynamics simulations. This project is an attempt at pushing forward the development of the method while expanding its scope of real-life applications via targeted control of unruptured intracranial aneurysms.

Developing new capabilities to predict the risk of intracranial aneurysm rupture and to improve treatment outcomes in the follow-up of endovascular repair is indeed of tremendous medical and societal interest, both to support decision-making and assessment of treatment options by medical doctors, and to improve the life quality and expectancy of patients.
The proposal aims at designing and characterizing novel flow-deviator stent devices from a state-of-the art computational framework combining advanced numerical methods for high-fidelity fluid-structure interaction modeling (to accurately describe the coupling between the blood flow, the vessel tissue, and the stent) and deep reinforcement learning algorithms (to optimize the functional parameters in the implanted state), which we envision as a first step towards enabling patient-specific treatment. This has never been done before in this context and should open both new theoretical and numerical opportunities.

SiCURE brings novel computational and optimization frameworks to the next level capable of studying selected flow diverter treatment in order to reduce the risk of hemorrhage in cerebral aneurysms, of supporting the decisions of treatment options by medical doctors and finally of providing guidance in the development of new implant design. Such capabilities can save millions of lives worldwide, improve the life quality of patients, eliminate lifelong side-effects due to sub-optimal treatment planning and delivery, and reduce the tremendous economic burden linked to poor patient outcome.

The project is highly multidisciplinary, and the methods proposed and developed as a part of this research can be quickly adapted to a wide range of engineering and bio-medical applications. The open-source provision of the obtained results for research and dissemination purposes will accelerate the spreading of DRL and its application to tangible applications, while maintaining a high level of reproducibility.