Virtual implantation of Derivo Embolization Devices using PreSize Neurovascular software: Accuracy evaluation and size selection comparison study.

BACKGROUND: Evaluating the deployed length of flow diverting stents (FDs) to select the optimal device size remains a challenging, yet crucial, task in aneurysm treatment. This study reports on the accuracy of PreSize Neurovascular (Oxford Heartbeat Ltd), a visualization and simulation software for FD intervention planning, in predicting FD deployed length, and on its impact on device size selection. METHODS: Imaging data from consecutive patients treated with Derivo Embolization Device (Acandis GmbH) were collected from University Medical Center Hamburg-Eppendorf and retrospectively analyzed. Accuracy evaluation: prediction accuracy was calculated by comparing deployed FD lengths measured from imaging data and simulated by PreSize. Size selection comparison: two Interventional Neuroradiologists (INR1, INR2), blinded to the devices deployed, used PreSize to select the optimal device size (diameter and length). Their choices were compared against the deployed devices selected by conventional planning. RESULTS: Among 98 implanted devices, PreSize predicted deployed FD length with a mean accuracy of 94.54% (95% confidence interval [93.72%, 95.35%]). Among 98 aneurysm cases, PreSize-informed device lengths were significantly shorter (Wilcoxon signed-rank test, INR1: W = 394, P < .001, INR2: W = 305, P < .001) by 4.13 and 4.18 mm on average, and up to 20 and 25 mm, for INR1 and INR2, respectively, than the conventionally selected FDs. In 32% of cases, PreSize-informed devices resulted in fewer vessel bends covered by the FD while achieving sufficient aneurysm coverage. CONCLUSIONS: PreSize retrospectively predicted deployed FD lengths with high accuracy. Moreover, INRs in this study were more inclined to select shorter stent length in the simulation than they would have done conventionally.

Towards Reduced Order Models via Robust Proper Orthogonal Decomposition to capture personalised aortic haemodynamics.

Data driven, reduced order modelling has shown promise in tackling the challenges associated with computational and experimental haemodynamic models. In this work, we focus on the use of Reduced Order Models (ROMs) to reconstruct velocity fields in a patient-specific dissected aorta, with the objective being to compare the ROMs obtained from Robust Proper Orthogonal Decomposition (RPOD) to those obtained from the traditional Proper Orthogonal Decomposition (POD). POD and RPOD are applied to in vitro, haemodynamic data acquired by Particle Image Velocimetry and compare the decomposed flows to those derived from Computational Fluid Dynamics (CFD) data for the same geometry and flow conditions. In this work, PIV and CFD results act as surrogates for clinical haemodynamic data e.g. MR, helping to demonstrate the potential use of ROMS in real clinical scenarios. The flow is reconstructed using different numbers of POD modes and the flow features obtained throughout the cardiac cycle are compared to the original Full Order Models (FOMs). Robust Principal Component Analysis (RPCA), the first step of RPOD, has been found to enhance the quality of PIV data, allowing POD to capture most of the kinetic energy of the flow in just two modes similar to the numerical data that are free from measurement noise. The reconstruction errors differ along the cardiac cycle with diastolic flows requiring more modes for accurate reconstruction. In general, modes 1-10 are found sufficient to represent the flow field. The results demonstrate that the coherent structures that characterise this aortic dissection flow are described by the first few POD modes suggesting that it is possible to represent the macroscale behaviour of aortic flow in a low-dimensional space; thus significantly simplifying the problem, and allowing for more computationally efficient flow simulations or machine learning based flow predictions that can pave the way for translation of such models to the clinic.

Experimental evaluation of the patient-specific haemodynamics of an aortic dissection model using particle image velocimetry.

Aortic Dissection (AD) is a complex pathology that affects the aorta. Diagnosis, management and treatment remain a challenge as it is a highly patient-specific pathology and there is still a limited understanding of the fluid-mechanics phenomena underlying clinical outcomes. Although in vitro models can allow the accurate study of AD flow fields in physical phantoms, they are currently scarce and almost exclusively rely on over simplifying assumptions. In this work, we present the first experimental study of a patient-specific case of AD. An anatomically correct phantom was produced and combined with a state-of-the-art in vitro platform, informed by clinical data, employed to accurately reproduce personalised conditions. The complex AD haemodynamics reproduced by the platform was characterised by flow rate and pressure acquisitions as well as Particle Image Velocimetry (PIV) derived velocity fields. Clinically relevant haemodynamic indices, that can be correlated with AD prognosis - such as velocity, shear rate, turbulent kinetic energy distributions - were extracted in two regions of interest in the aortic domain. The acquired data highlighted the complex nature of the flow (e.g. recirculation regions, low shear rate in the false lumen) and was in very good agreement with the available clinical data and the CFD results of a study conducted alongside, demonstrating the accuracy of the findings. These results demonstrate that the described platform constitutes a powerful, unique tool to reproduce in vitro personalised haemodynamic conditions, which can be used to support the evaluation of surgical procedures, medical devices testing and to validate state-of-the-art numerical models.

A Computational Framework for Pre-Interventional Planning of Peripheral Arteriovenous Malformations.

PURPOSE: Peripheral arteriovenous malformations (pAVMs) are congenital lesions characterised by abnormal high-flow, low-resistance vascular connections-the so-called nidus-between arteries and veins. The mainstay treatment typically involves the embolisation of the nidus, however the complexity of pAVMs often leads to uncertain outcomes. This study aims at developing a simple, yet effective computational framework to aid the clinical decision making around the treatment of pAVMs using routinely acquired clinical data. METHODS: A computational model was developed to simulate the pre-, intra-, and post-intervention haemodynamics of a patient-specific pAVM. A porous medium of varying permeability was employed to simulate the sclerosant effect on the nidus haemodynamics. Results were compared against clinical data (digital subtraction angiography, DSA, images) and experimental flow-visualization results in a 3D-printed phantom of the same pAVM. RESULTS: The computational model allowed the simulation of the pAVM haemodynamics and the sclerotherapy-induced changes at different interventional stages. The predicted inlet flow rates closely matched the DSA-derived data, although the post-intervention one was overestimated, probably due to vascular system adaptations not accounted for numerically. The nidus embolization was successfully captured by varying the nidus permeability and increasing its hydraulic resistance from 0.330 to 3970 mmHg s ml-1. The nidus flow rate decreased from 71% of the inlet flow rate pre-intervention to 1%: the flow completely bypassed the nidus post-intervention confirming the success of the procedure. CONCLUSION: The study demonstrates that the haemodynamic effects of the embolisation procedure can be simulated from routinely acquired clinical data via a porous medium with varying permeability as evidenced by the good qualitative agreement between numerical predictions and both in vivo and in vitro data. It provides a fundamental building block towards a computational treatment-planning framework for AVM embolisation.

A novel MRI-based data fusion methodology for efficient, personalised, compliant simulations of aortic haemodynamics.

We present a novel, cost-efficient methodology to simulate aortic haemodynamics in a patient-specific, compliant aorta using an MRI data fusion process. Based on a previously-developed Moving Boundary Method, this technique circumvents the high computational cost and numerous structural modelling assumptions required by traditional Fluid-Structure Interaction techniques. Without the need for Computed Tomography (CT) data, the MRI images required to construct the simulation can be obtained during a single imaging session. Black Blood MR Angiography and 2D Cine-MRI data were used to reconstruct the luminal geometry and calibrate wall movement specifically to each region of the aorta. 4D-Flow MRI and non-invasive pressure measurements informed patient-specific inlet and outlet boundary conditions. Luminal area closely matched 2D Cine-MRI measurements with a mean error of less than 4.6% across the cardiac cycle, while physiological pressure and flow distributions were simulated to within 3.3% of patient-specific targets. Moderate agreement with 4D-Flow MRI velocity data was observed. Despite lower peak velocity, an equivalent rigid-wall simulation predicted a mean Time-Averaged Wall Shear Stress (TAWSS) 13% higher than the compliant simulation. The agreement observed between compliant simulation results and MRI data is testament to the accuracy and efficiency of this MRI-based simulation technique.

A Combined In Vivo, In Vitro, In Silico Approach for Patient-Specific Haemodynamic Studies of Aortic Dissection.

The optimal treatment of Type-B aortic dissection (AD) is still a subject of debate, with up to 50% of the cases developing late-term complications requiring invasive intervention. A better understanding of the patient-specific haemodynamic features of AD can provide useful insights on disease progression and support clinical management. In this work, a novel in vitro and in silico framework to perform personalised studies of AD, informed by non-invasive clinical data, is presented. A Type-B AD was investigated in silico using computational fluid dynamics (CFD) and in vitro by means of a state-of-the-art mock circulatory loop and particle image velocimetry (PIV). Both models not only reproduced the anatomical features of the patient, but also imposed physiologically-accurate and personalised boundary conditions. Experimental flow rate and pressure waveforms, as well as detailed velocity fields acquired via PIV, are extensively compared against numerical predictions at different locations in the aorta, showing excellent agreement. This work demonstrates how experimental and numerical tools can be developed in synergy to accurately reproduce patient-specific AD blood flow. The combined platform presented herein constitutes a powerful tool for advanced haemodynamic studies for a range of vascular conditions, allowing not only the validation of CFD models, but also clinical decision support, surgical planning as well as medical device innovation.

An in silico study of the influence of vessel wall deformation on neointimal hyperplasia progression in peripheral bypass grafts.

Neointimal hyperplasia (NIH) is a major obstacle to graft patency in the peripheral arteries. A complex interaction of biomechanical factors contribute to NIH development and progression, and although haemodynamic markers such as wall shear stress have been linked to the disease, these have so far been insufficient to fully capture its behaviour. Using a computational model linking computational fluid dynamics (CFD) simulations of blood flow with a biochemical model representing NIH growth mechanisms, we analyse the effect of compliance mismatch, due to the presence of surgical stitches and/or to the change in distensibility between artery and vein graft, on the haemodynamics in the lumen and, subsequently, on NIH progression. The model enabled to simulate NIH at proximal and distal anastomoses of three patient-specific end-to-side saphenous vein grafts under two compliance-mismatch configurations, and a rigid wall case for comparison, obtaining values of stenosis similar to those observed in the computed tomography (CT) scans. The maximum difference in time-averaged wall shear stress between the rigid and compliant models was 3.4 Pa, and differences in estimation of NIH progression were only observed in one patient. The impact of compliance on the haemodynamic-driven development of NIH was small in the patient-specific cases considered.

Patient-specific haemodynamic simulations of complex aortic dissections informed by commonly available clinical datasets.

Patient-specific computational fluid-dynamics (CFD) can assist the clinical decision-making process for Type-B aortic dissection (AD) by providing detailed information on the complex intra-aortic haemodynamics. This study presents a new approach for the implementation of personalised CFD models using non-invasive, and oftentimes minimal, datasets commonly collected for AD monitoring. An innovative way to account for arterial compliance in rigid-wall simulations using a lumped capacitor is introduced, and a parameter estimation strategy for boundary conditions calibration is proposed. The approach was tested on three complex cases of AD, and the results were successfully compared against invasive blood pressure measurements. Haemodynamic results (e.g. intraluminal pressures, flow partition between the lumina, wall shear-stress based indices) provided information that could not be obtained using imaging alone, providing insight into the state of the disease. It was noted that small tears in the distal intimal flap induce disturbed flow in both lumina. Moreover, oscillatory pressures across the intimal flap were often observed in proximity to the tears in the abdominal region, which could indicate a risk of dynamic obstruction of the true lumen. This study shows how combining commonly available clinical data with computational modelling can be a powerful tool to enhance clinical understanding of AD.

A simplified method to account for wall motion in patient-specific blood flow simulations of aortic dissection: Comparison with fluid-structure interaction.

Aortic dissection (AD) is a complex and highly patient-specific vascular condition difficult to treat. Computational fluid dynamics (CFD) can aid the medical management of this pathology, yet its modelling and simulation are challenging. One aspect usually disregarded when modelling AD is the motion of the vessel wall, which has been shown to significantly impact simulation results. Fluid-structure interaction (FSI) methods are difficult to implement and are subject to assumptions regarding the mechanical properties of the vessel wall, which cannot be retrieved non-invasively. This paper presents a simplified 'moving-boundary method' (MBM) to account for the motion of the vessel wall in type-B AD CFD simulations, which can be tuned with non-invasive clinical images (e.g. 2D cine-MRI). The method is firstly validated against the 1D solution of flow through an elastic straight tube; it is then applied to a type-B AD case study and the results are compared to a state-of-the-art, full FSI simulation. Results show that the proposed method can capture the main effects due to the wall motion on the flow field: the average relative difference between flow and pressure waves obtained with the FSI and MBM simulations was less than 1.8% and 1.3%, respectively and the wall shear stress indices were found to have a similar distribution. Moreover, compared to FSI, MBM has the advantage to be less computationally expensive (requiring half of the time of an FSI simulation) and easier to implement, which are important requirements for clinical translation.

Virtual TEVAR: Overcoming the Roadblocks of In-Silico Tools for Aortic Dissection Treatment.

The use of in silico tools for the interventional planning of complex vascular conditions, such as Aortic Dissections has been often limited by high computational cost, involving long timescales for accurate results to be produced and low numbers of patients, precluding the use of statistical analyses to inform individual-level models. In the paper [Theranostics 2018; 8(20):5758-5771. doi:10.7150/thno.28944], Chen et al. proposed a novel algorithm to compute patient-specific 'virtual TEVAR' that will help clinicians to approach individual treatment and decision-making based on objective and quantifiable metrics and validated on a cohort of 66 patients in real time. This research will significantly impact the field and has the potential to transform the way clinical interventions will be approached in the future.

Computational tools for clinical support: a multi-scale compliant model for haemodynamic simulations in an aortic dissection based on multi-modal imaging data.

Aortic dissection (AD) is a vascular condition with high morbidity and mortality rates. Computational fluid dynamics (CFD) can provide insight into the progression of AD and aid clinical decisions; however, oversimplified modelling assumptions and high computational cost compromise the accuracy of the information and impede clinical translation. To overcome these limitations, a patient-specific CFD multi-scale approach coupled to Windkessel boundary conditions and accounting for wall compliance was developed and used to study a patient with AD. A new moving boundary algorithm was implemented to capture wall displacement and a rich in vivo clinical dataset was used to tune model parameters and for validation. Comparisons between in silico and in vivo data showed that this approach successfully captures flow and pressure waves for the patient-specific AD and is able to predict the pressure in the false lumen (FL), a critical variable for the clinical management of the condition. Results showed regions of low and oscillatory wall shear stress which, together with higher diastolic pressures predicted in the FL, may indicate risk of expansion. This study, at the interface of engineering and medicine, demonstrates a relatively simple and computationally efficient approach to account for arterial deformation and wave propagation phenomena in a three-dimensional model of AD, representing a step forward in the use of CFD as a potential tool for AD management and clinical support.

Gas transfer model to design a ventilator for neonatal total liquid ventilation.

The study was aimed to optimize the gas transfer in an innovative ventilator for neonatal Total Liquid Ventilation (TLV) that integrates the pumping and oxygenation functions in a non-volumetric pulsatile device made of parallel flat silicone membranes. A computational approach was adopted to evaluate oxygen (O2) and carbon dioxide (CO2) exchanges between the liquid perfluorocarbon (PFC) and the oxygenating gas, as a function of the geometrical parameter of the device. A 2D semi-empirical model was implemented to this purpose using Comsol Multiphysics to study both the fluid dynamics and the gas exchange in the ventilator. Experimental gas exchanges measured with a preliminary prototype were compared to the simulation outcomes to prove the model reliability. Different device configurations were modeled to identify the optimal design able to guarantee the desired gas transfer. Good agreement between experimental and simulation outcomes was obtained, validating the model. The optimal configuration, able to achieve the desired gas exchange (ΔpCO2 = 16.5 mmHg and ΔpO2 = 69 mmHg), is a device comprising 40 modules, 300 mm in length (total exchange area = 2.28 m(2)). With this configuration gas transfer performance is satisfactory for all the simulated settings, proving good adaptability of the device.