Project Details
Multi-scale algorithms and simulation methodologies for the long-term prognosis of endovascular interventions in cerebral aneurysms
Applicants
Professor Dr. Jan Stefan Kirschke; Professor Dr.-Ing. Alexander Popp; Professorin Dr. Barbara Wohlmuth
Subject Area
Applied Mechanics, Statics and Dynamics
Mathematics
Medical Informatics and Medical Bioinformatics
Mathematics
Medical Informatics and Medical Bioinformatics
Term
since 2021
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 465242983
We aim at a simulation-based tool for neuro-interventionalists to evaluate the long-term outcome of endovascularly treated cerebral aneurysms. Immediately after insertion of the endovascular device, a thrombus formation begins due to flow impairment and the interaction with the coil material. Due to secondary coil compaction, i.e., through a pulsatile flow or intraaneurysmatic thrombolysis, and by inflammatory thinning and expansion of the aneurysm wall, recurrence or secondary rupture of the aneurysm can also occur in an interval of weeks to years after aneurysm treatment. Based on our first phase results, dealing with the „in-time“ simulation of the treatment process by a mechanical coil model, blood flow simulations within the aneurysm, porous media based surrogate models and fluid-structure interaction, we now shift our goal towards the aforementioned long-term perspective of the patient and target the prediction of recurrence. We will focus on a modular fluid-structure-solid coupling allowing for a bidirectionally mixed-dimensional framework. To fully exploit the potential of different discretization paradigms, we provide not only methodologically coupled models of different dimension but also the coupling of modern open-source software packages. Main challenges are hereby the complexity of highly non-linear bio-mechanical models, different dimensions from 0D Windkessel to 1D and 2D structure up to 3D solid models and the coupling of different data-structures in software concepts. In-silico Raymond--Roy occlusion quality criteria, developed in phase one, will be calibrated and validated by comparison to experimental data provided by our cooperation partners. Regarding endovascular devices, we will enhance our complexity reduced mechanical models by bio-active components, e.g., HydroCoils or coated flow diverters and include structure-structure interaction. Starting from a reduced wall model, bio-mechanical models incorporating inflammatory processes that change the wall's material properties will be included. Our lattice Boltzmann/finite element coupled fluid-structure interaction implementation naturally extends to a fluid-structure-solid approach taking into account the outside tissue to capture the edema formation process and the thrombus within. To quantify and predict potential coil compactification, we will develop a phenomenological shrinkage model for the thrombus, coupled to wall and coil mechanics. To enhance clinical applicability and to be of day-to-day use for neuro-interventionalists, we will advance machine learning techniques as prediction tools and provide a software tutorial for virtual training on coiling. In summary of both project phases, we set out a modelling and algorithmic virtual treatment pipeline from acquisition of patient-specific data, to the support of the coiling choice within a personalized treatment plan and to a risk analysis for recurrence and secondary rupture on a longer post-treatment time scale.
DFG Programme
Priority Programmes