Comparison and calibration of a real-time virtual stenting algorithm using Finite Element Analysis and Genetic Algorithms

High variability in physician estimations of flow-diverting stent deployment versus PreSize Neurovascular software simulation: a comparison study

BACKGROUND

Physician variablity in preoperative planning of endovascular implant deployment and associated inaccuracies have not been documented. This study aimed to quantify the variability in accuracy of physician flow diverter (FD) planning and directly compares it with PreSize Neurovascular (Oxford Heartbeat Ltd) software simulations.

METHODS

Eight experienced neurointerventionalists (NIs), blinded to procedural details, were provided with preoperative 3D rotational angiography (3D-RA) volumetric data along with images annotated with the distal landing location of a deployed Surpass Evolve (Stryker Neurovascular) FD from 51 patient cases. NIs were asked to perform a planning routine reflecting their normal practice and estimate the stent's proximal landing using volumetric data and the labeled dimensions of the FD used. Equivalent deployed length estimation was performed using PreSize software. NI- and software-estimated lengths were compared with postprocedural observed deployed stent length (control) using Bland-Altman plots. NI assessment agreement was assessed with the intraclass correlation coefficient (ICC).

RESULTS

The mean accuracy of NI-estimated deployed FD length was 81% (±15%) versus PreSize's accuracy of 95% (±4%), demonstrating significantly higher accuracy for the software (p<0.001). The mean absolute error between estimated and control lengths was 4 mm (±3.5 mm, range 0.03-30.2 mm) for NIs and 1 mm (±0.9 mm, range 0.01-3.9 mm) for PreSize. No discernable trends in accuracy among NIs or across vasculature and aneurysm morphology (size, vessel diameter, tortuousity) were found.

CONCLUSIONS

The study quantified experienced physicians' significant variablity in predicting an FD deployment with current planning approaches. In comparison, PreSize-simulated FD deployment was consistently more accurate and reliable, demonstrating its potential to improve standard of practice.

A CT-based predictive model for stent-induced vessel damage: application to type B aortic dissection

OBJECTIVES

The distal stent-induced new entry (distal SINE) is a life-threatening device-related complication after thoracic endovascular aortic repair (TEVAR). However, risk factors for distal SINE are not fully determined, and prediction models are lacking. This study aimed to establish a predictive model for distal SINE based on the preoperative dataset.

METHODS

Two hundred and six patients with Stanford type B aortic dissection (TBAD) that experienced TEVAR were involved in this study. Among them, thirty patients developed distal SINE. Pre-TEVAR morphological parameters were measured based on the CT-reconstructed configurations. Virtual post-TEVAR morphological and mechanical parameters were computed via the virtual stenting algorithm (VSA). Two predictive models (PM-1 and PM-2) were developed and presented as nomograms to help risk evaluation of distal SINE. The performance of the proposed predictive models was evaluated and internal validation was conducted.

RESULTS

Machine-selected variables for PM-1 included key pre-TEVAR parameters, and those for PM-2 included key virtual post-TEVAR parameters. Both models showed good calibration in both development and validation subsamples, while PM-2 outperformed PM-1. The discrimination of PM-2 was better than PM-1 in the development subsample, with an optimism-corrected area under the curve (AUC) of 0.95 and 0.77, respectively. Application of PM-2 in the validation subsample presented good discrimination with an AUC of 0.9727. The decision curve demonstrated that PM-2 was clinically useful.

CONCLUSION

This study proposed a predictive model for distal SINE incorporating the CT-based VSA. This predictive model could efficiently predict the risk of distal SINE and thus might contribute to personalized intervention planning.

CLINICAL RELEVANCE STATEMENT

This study established a predictive model to evaluate the risk of distal SINE based on the pre-stenting CT dataset and planned device information. With an accurate VSA tool, the predictive model could help to improve the safety of the endovascular repair procedure.

KEY POINTS

• Clinically useful prediction models for distal stent-induced new entry are still lacking, and the safety of the stent implantation is hard to guarantee. • Our proposed predictive tool based on a virtual stenting algorithm supports different stenting planning rehearsals and real-time risk evaluation, guiding clinicians to optimize the presurgical plan when necessary. • The established prediction model provides accurate risk evaluation for vessel damage, improving the safety of the intervention procedure.

Machine learning and reduced order modelling for the simulation of braided stent deployment

Endoluminal reconstruction using flow diverters represents a novel paradigm for the minimally invasive treatment of intracranial aneurysms. The configuration assumed by these very dense braided stents once deployed within the parent vessel is not easily predictable and medical volumetric images alone may be insufficient to plan the treatment satisfactorily. Therefore, here we propose a fast and accurate machine learning and reduced order modelling framework, based on finite element simulations, to assist practitioners in the planning and interventional stages. It consists of a first classification step to determine a priori whether a simulation will be successful (good conformity between stent and vessel) or not from a clinical perspective, followed by a regression step that provides an approximated solution of the deployed stent configuration. The latter is achieved using a non-intrusive reduced order modelling scheme that combines the proper orthogonal decomposition algorithm and Gaussian process regression. The workflow was validated on an idealized intracranial artery with a saccular aneurysm and the effect of six geometrical and surgical parameters on the outcome of stent deployment was studied. We trained six machine learning models on a dataset of varying size and obtained classifiers with up to 95% accuracy in predicting the deployment outcome. The support vector machine model outperformed the others when considering a small dataset of 50 training cases, with an accuracy of 93% and a specificity of 97%. On the other hand, real-time predictions of the stent deployed configuration were achieved with an average validation error between predicted and high-fidelity results never greater than the spatial resolution of 3D rotational angiography, the imaging technique with the best spatial resolution (0.15 mm). Such accurate predictions can be reached even with a small database of 47 simulations: by increasing the training simulations to 147, the average prediction error is reduced to 0.07 mm. These results are promising as they demonstrate the ability of these techniques to achieve simulations within a few milliseconds while retaining the mechanical realism and predictability of the stent deployed configuration.

Evaluation and Verification of Fast Computational Simulations of Stent-Graft Deployment in Endovascular Aneurysmal Repair

Fenestrated Endovascular Aortic Repair, also known as FEVAR, is a minimally invasive procedure that allows surgeons to repair the aorta while still preserving blood flow to kidneys and other critical organs. Given the high complexity of FEVAR, there is a pressing need to develop numerical tools that can assist practitioners at the preoperative planning stage and during the intervention. The aim of the present study is to introduce and to assess an assistance solution named Fast Method for Virtual Stent-graft Deployment for computer assisted FEVAR. This solution, which relies on virtual reality, is based on a single intraoperative X-ray image. It is a hybrid method that includes the use of intraoperative images and a simplified mechanical model based on corotational beam elements. The method was verified on a phantom and validated on three clinical cases, including a case with fenestrations. More specifically, we quantified the errors induced by the different simplifications of the mechanical model, related to fabric simulation and aortic wall mechanical properties. Overall, all errors for both stent and fenestration positioning were less than 5 mm, making this method compatible with clinical expectations. More specifically, the errors related to fenestration positioning were less than 3 mm. Although requiring further validation with a higher number of test cases, our method could achieve an accuracy compatible with clinical specifications within limited calculation time, which is promising for future implementation in a clinical context.

Evaluation of Two Fast Virtual Stenting Algorithms for Intracranial Aneurysm Flow Diversion

BACKGROUND

Endovascular treatment of intracranial aneurysms (IAs) by flow diverter (FD) stents depends on flow modification. Patient-specific modeling of FD deployment and computational fluid dynamics (CFD) could enable a priori endovascular strategy optimization. We developed a fast, simplistic, expansion-free balls-weeping algorithm to model FDs in patientspecific aneurysm geometry. However, since such strong simplification could result in less accurate simulations, we also developed a fast virtual stenting workflow (VSW) that explicitly models stent expansion using pseudo-physical forces.

METHODS

To test which of these two fast algorithms more accurately simulates real FDs, we applied them to virtually treat three representative patient-specific IAs. We deployed Pipeline Embolization Device into 3 patient-specific silicone aneurysm phantoms and simulated the treatments using both balls-weeping and VSW algorithms in computational aneurysm models. We then compared the virtually deployed FD stents against experimental results in terms of geometry and post-treatment flow fields. For stent geometry, we evaluated gross configurations and porosity. For post-treatment aneurysmal flow, we compared CFD results against experimental measurements by particle image velocimetry.

RESULTS

We found that VSW created more realistic FD deployments than balls-weeping in terms of stent geometry, porosity and pore density. In particular, balls-weeping produced unrealistic FD bulging at the aneurysm neck, and this artifact drastically increased with neck size. Both FD deployment methods resulted in similar flow patterns, but the VSW had less error in flow velocity and inflow rate.

CONCLUSION

In conclusion, modeling stent expansion is critical for preventing unrealistic bulging effects and thus should be considered in virtual FD deployment algorithms. Also endowed with its high computational efficiency and superior accuracy, the VSW algorithm is a better candidate for implementation into a bedside clinical tool for FD deployment simulation.

Incomplete stent expansion in flow-diversion treatment affects aneurysmal haemodynamics: a quantitative comparison of treatments affected by different severities of malapposition occurring in different segments of the parent artery

Incomplete stent expansion (IncSE) is occasionally seen in flow-diversion (FD) treatment of intracranial aneurysms; however, its haemodynamic consequences remain inconclusive. Through a parametric study, we quantify the aneurysmal haemodynamics subject to different severities of IncSE occurring in different portions of the stent. Two patient cases with IncSE confirmed in vivo were studied. To investigate a wider variety of IncSE scenarios, we modelled IncSE at two severity levels respectively located in the proximal, central, or distal segment of a stent, yielding a total of 14 treatment scenarios (including the ideal deployment). We examined stent wire configurations in 14 scenarios and resolved aneurysm haemodynamics through computational fluid dynamics (CFD). A considerable degradation of aneurysm flow-reduction performance was observed when central or distal IncSE occurred, with the maximal elevations of the inflow rate (IR) and energy loss (EL) being 10% and 15%. The underlying mechanism might be the increased resistance for flow to remain within the FD stent, which forces more blood to leak into the aneurysm sac. Counter-intuitively, a slight reduction of aneurysm inflow was associated with proximal IncSE, with the maximal further reduction of the IR and EL being 5% and 8%. This may be due to the disruption of the predominant parent-artery flow by the collapsed wires, which decreased the strength and altered the direction of aneurysmal inflow. The effects of IncSE vary greatly with the location of occurrence, revealing the importance of performing individualised, patient-specific risk assessment before treatment.

Association Between Aneurysmal Haemodynamics and Device Microstructural Characteristics After Flow-Diversion Treatments With Dual Stents of Different Sizes: A Numerical Study

Objectives

Treating intracranial aneurysms with flow-diverting stents sometimes requires deployment of a second device. Herein we quantify the sizing effects of devices in dual-stent treatments upon the final stent microstructure and the post-treatment aneurysmal haemodynamics.

Methods

Fifteen sidewall ICA aneurysm geometries were included. Using a virtual stenting technique, we implanted either one or two stents for each aneurysm treatment considered, with each stent specified as one of two different sizes, yielding a total of two single-stent and fouir dual-stent treatment scenarios for each aneurysm. Three stent microstructural parameters and nine aneurysmal haemodynamic parameters were quantified and systematically compared across the 90 treatment scenarios.

Results

Deployment of a second stent further reduced the aneurysmal inflow rate (IR) and energy loss (EL) by, respectively, 14 ± 11% (p = 0.001) and 9 ± 12% (p = 0.056), relative to the untreated condition. Sizing effects of the earlier-deployed stent led to largest differences of 6.9% for the final IR reduction and 11.1% for the EL, whereas sizing effects from the later-deployed stent were minor (≤2.1%). The change in stent pore size was the only microstructural parameter demonstrating a strong correlation with the reduction in the post-treatment aneurysmal haemodynamics, in terms of the IR (r = 0.50, p < 0.001) and pressure drop (r = 0.63, p < 0.001).

Conclusion

Size of the earlier-deployed stent has substantial effects on the final haemodynamic outcomes after dual-stent treatment. The average pore size of stent wires at the aneurysm orifice shows promise as a potential index for predicting the efficacy of flow-diversion treatments.

Multi-stage learning for segmentation of aortic dissections using a prior aortic anatomy simplification

Aortic dissection (AD) is a life-threatening cardiovascular disease with a high mortality rate. The accurate and generalized 3-D reconstruction of AD from CT-angiography can effectively assist clinical procedures and surgery plans, however, is clinically unavaliable due to the lacking of efficient tools. In this study, we presented a novel multi-stage segmentation framework for type B AD to extract true lumen (TL), false lumen (FL) and all branches (BR) as different classes. Two cascaded neural networks were used to segment the aortic trunk and branches and to separate the dual lumen, respectively. An aortic straightening method was designed based on the prior vascular anatomy of AD, simplifying the curved aortic shape before the second network. The straightening-based method achieved the mean Dice scores of 0.96, 0.95 and 0.89 for TL, FL, and BR on a multi-center dataset involving 120 patients, outperforming the end-to-end multi-class methods and the multi-stage methods without straightening on the dual-lumen segmentation, even using different network architectures. Both the global volumetric features of the aorta and the local characteristics of the primary tear could be better identified and quantified based on the straightening. Comparing to previous deep learning methods dealing with AD segmentations, the proposed framework presented advantages in segmentation accuracy.

Deployment of flow diverter devices: prediction of foreshortening and validation of the simulation in 18 clinical cases

PURPOSE

Flow diverter (FD) devices show severe shortening during deployment in dependency of the vessel geometry. Valid information regarding the geometry of the targeted vessel is therefore mandatory for correct device selection, and to avoid complications. But the geometry of diseased tortuous intracranial vessels cannot be measured accurately with standard methods. The goal of this study is to prove the accuracy of a novel virtual stenting method in prediction of the behavior of a FD in an individual vessel geometry.

METHODS

We applied a virtual stenting method on angiographic 3D imaging data of the specific vasculature of patients, who underwent FD treatment. The planning tool analyzes the local vessel morphology and deploys the FD virtually. We measured in 18 cases the difference between simulated FD length and real FD length after treatment in a landmark-based registration of pre-/post-interventional 3D angiographic datasets.

RESULTS

The mean value of length deviation of the virtual FD was 2.2 mm (SD ± 1.9 mm) equaling 9.5% (SD ± 8.2%). Underestimated cases present lower deviations compared with overestimated FDs. Flow diverter cases with a nominal device length of 20 mm had the highest prediction accuracy.

CONCLUSION

The results suggest that the virtual stenting method used in this study is capable of predicting FD length with a clinically sufficient accuracy in advance and could therefore be a helpful tool in intervention planning. Imaging data of high quality are mandatory, while processing and manipulation of the FD during the intervention may impact the accuracy.

Mechanical Identification of Materials and Structures with Optical Methods and Metaheuristic Optimization

This study presents a hybrid framework for mechanical identification of materials and structures. The inverse problem is solved by combining experimental measurements performed by optical methods and non-linear optimization using metaheuristic algorithms. In particular, we develop three advanced formulations of Simulated Annealing (SA), Harmony Search (HS) and Big Bang-Big Crunch (BBBC) including enhanced approximate line search and computationally cheap gradient evaluation strategies. The rationale behind the new algorithms-denoted as Hybrid Fast Simulated Annealing (HFSA), Hybrid Fast Harmony Search (HFHS) and Hybrid Fast Big Bang-Big Crunch (HFBBBC)-is to generate high quality trial designs lying on a properly selected set of descent directions. Besides hybridizing SA/HS/BBBC metaheuristic search engines with gradient information and approximate line search, HS and BBBC are also hybridized with an enhanced 1-D probabilistic search derived from SA. The results obtained in three inverse problems regarding composite and transversely isotropic hyperelastic materials/structures with up to 17 unknown properties clearly demonstrate the validity of the proposed approach, which allows to significantly reduce the number of structural analyses with respect to previous SA/HS/BBBC formulations and improves robustness of metaheuristic search engines.

Virtual stenting with simplex mesh and mechanical contact analysis for real-time planning of thoracic endovascular aortic repair

In aortic endovascular repair, the prediction of stented vessel remodeling informs treatment plans and risk evaluation; however, there are no highly accurate and efficient methods to quantitatively simulate stented vessels. This study developed a fast virtual stenting algorithm to simulate stent-induced aortic remodeling to assist in real-time thoracic endovascular aortic repair planning.

Methods: The virtual stenting algorithm was established based on simplex deformable mesh and mechanical contact analysis. The key parameters of the mechanical contact analysis were derived from mechanical tests on aortic tissue (n=40) and commonly used stent-grafts (n=6). Genetic algorithm was applied to select weighting parameters. Testing and validation of the algorithm were performed using pre- and post-treatment computed tomography angiography datasets of type-B aortic dissection cases (n=66).

Results: The algorithm was efficient in simulating stent-induced aortic deformation (mean computing time on a single processor: 13.78±2.80s) and accurate at the morphological (curvature difference: 1.57±0.57%; cross-sectional area difference: 4.11±0.85%) and hemodynamic (similarity of wall shear stress-derived parameters: 90.16-90.94%) levels. Stent-induced wall deformation was higher (p<0.05) in distal stent-induced new entry cases than in successfully treated cases, and this deformation did not differ significantly among the different stent groups. Additionally, the high stent-induced wall deformation regions and the new-entry sites overlapped, indicating the usefulness of wall deformation to evaluate the risks of device-induced complications.

Conclusion: The novel algorithm provided fast real-time and accurate predictions of stent-graft deployment with luminal deformation tracking, thereby potentially informing individualized stenting planning and improving endovascular aortic repair outcomes. Large, multicenter studies are warranted to extend the algorithm validation and determine stress-induced wall deformation cutoff values for the risk stratification of particular complications.

Applying computer simulation to the design of flow-diversion treatment for intracranial aneurysms

Although flow-diversion (FD) treatment has been proven to be able to induce intracranial aneurysm (IA) occlusion, clinical follow-ups reported that a number of patients may still suffer from delayed IA rupture or incomplete aneurysm occlusion post-treatment. Complete aneurysm occlusion is believed to be associated with favourable haemodynamic alteration post-treatment, which may be greatly affected by the selection of device size and quantity, as well as the FD deployment procedure. However, clinicians have to choose and deploy the FD relying on their experience, since no post-stenting haemodynamic information is generally available to them prior to a specific treatment. In this study, using a virtual FD deployment technique and computational fluid dynamics method, we demonstrate and compare the haemodynamic changes after virtual FD treatments using a variety of prospective treating strategies.

Haemodynamic effects of stent diameter and compaction ratio on flow-diversion treatment of intracranial aneurysms: A numerical study of a successful and an unsuccessful case

BACKGROUND

Compacting a flow-diverting (FD) stent is an emerging technique to create a denser configuration of wires across the aneurysm ostium. However, quantitative analyses of post-stenting haemodynamics affected by the compaction level of different stent sizes remain inconclusive.

OBJECTIVE

To compare the aneurysmal haemodynamic alterations after virtual FD treatments with different device diameters at different compaction ratios.

METHODS

We virtually implanted three sizes of FD stent, with each size deployed at four compaction ratios, into two patient aneurysms previously treated with the Silk+FD-one successful case and the other unsuccessful. Wire configurations of the FD in the 24 treatment scenarios were examined, and aneurysmal haemodynamic alterations were resolved by computational fluid dynamics (CFD) simulations. We investigated the aneurysmal flow patterns, aneurysmal average velocity (AAV), mass flowrate (MF), and energy loss (EL) in each scenario.

RESULTS

Compactions of the stent in the successful case resulted in a greater metal coverage rate than that achieved in the unsuccessful one. A 25% increment in compaction ratio further decreased the AAV (12%), MF (11%), and EL (9%) in both cases (average values). The averaged maximum differences attributable to device size were 10% (AAV), 8% (MF), and 9% (EL).

CONCLUSIONS

Both stent size and compaction level could markedly affect the FD treatment outcomes. It is therefore important to individualise the treatment plan by selecting the optimal stent size and deployment procedure. CFD simulation can be used to investigate the treatment outcomes, thereby assisting doctors in choosing a favourable treatment plan.

Towards Predicting Patient-Specific Flow-Diverter Treatment Outcomes for Bifurcation Aneurysms: From Implantation Rehearsal to Virtual Angiograms

Despite accounting for the majority of all cerebral aneurysm cases, bifurcation aneurysms present many challenges to standard endovascular treatment techniques. This study examines the treatment of bifurcation aneurysms endovascularly with flow-diverting stents and presents an integrative computational modeling suite allowing for rehearsing all aspects of the treatment. Six bifurcation aneurysms are virtually treated with 70% porosity flow-diverters. Substantial reduction (>50%) in aneurysm inflow due to device deployment is predicted in addition to reductions in peak and average aneurysm wall shear stress to values considered physiologically normal. The subsequent impact of flow-diverter deployment on daughter vessels that are jailed by the device is investigated further, with a number of simulations conducted with increased outlet pressure conditions at jailed vessels. Increased outlet pressures at jailed daughter vessels are found to have little effect on device-induced aneurysm inflow reduction, but large variation (13–86%) is seen in the resulting reduction in daughter vessel flow rate. Finally, we propose a potentially powerful approach for validation of such models, by introducing an angiographic contrast model, with contrast transport modeled both before and after virtual treatment. Virtual angiograms and contrast residence curves are created, which offer unique clinical relevance and the potential for future in vivo verification of simulated results.