Abstract Background and aims In-silico models, often by means of the finite element (FE) method, are increasingly used to provide insight into biomechanical processes difficult to assess in-vivo. Patient-specific simulations are currently not integrated into the clinical workflow of acute ischemic stroke (AIS) due to time constraints. However, they can support the development of predictive models to aid clinical decision-making. For such applications to be reliable and clinically meaningful, computational results must be credible, making model validation a fundamental requirement. Methods After validation achieved by successfully simulating eight in-vitro tests, four stent-retriever-only, two aspiration-only and two combined procedures, six AIS patient-specific cases, two per procedure, were simulated. The patient-specific validation workflow was: (i) identification of clot location, length and device position, (ii) segmentation of vessels, (iii) in-silico replica of the pre-procedural anatomy comprising the occluding clot, (iv) in-silico replica of the entire procedure (v) comparison of in-silico and technical outcomes in terms of positive or negative recanalization of the occluded vessel, and (vi) computation of the largest clot fragment dimensions generated during retrieval to assess the potential for secondary distal occlusions. Results In all cases, the thrombectomy simulation outcomes agreed with the technical outcomes. The figure shows a representative example of stent-retriever-only case with negligible clot fragmentation, consistent with an eTICI score of 3. Conclusions Although computational time currently limits clinical applicability, the FE-based thrombectomy simulations proved credibly reproduced clinical procedures for three techniques. and will be extended to a virtual population (around thousand of cases) to create a tool to support clinical decision-making. Conflict of interest This project has received funding from the European Union’s Horizon Europe research and innovation programme under the grant agreement No 101136438. VF: nothing to disclose. LB: nothing to disclose. TPK: nothing to disclose. PJVD: nothing to disclose. CBLM: shareholder of Nicolab (minority interest). GL: nothing to disclose. FM: nothing to disclose. FJHG: nothing to disclose. HM: cofounder and shareholder of Nicolab, TrianecT and inSteps. JFRM: nothing to disclose. Figure 1 - belongs to Results
Fregona et al. (Fri,) studied this question.