BACKGROUND:Transcatheter aortic valve (TAV) failure may require redo transcatheter aortic valve replacement (redo-TAVR). The multiple layers of metal after redo-TAVR may jeopardize coronary artery patency and accessibility and result in prosthesis-patient mismatch. Dedicated software may use multi-slice computed tomography (MSCT) imaging to simulate TAV deployment relative to the prior TAV and the individual's anatomy and predict coronary accessibility and TAV expansion. AIM:We aimed to validate redo-TAVR simulations created in FEops HEARTguide. METHODS:This multicenter observational study included redo-TAVR patients who had MSCT imaging available at baseline, after index-TAVR and after redo-TAVR. Patient-specific redo-TAVR simulations were created. Simulated valve-to-aorta distances (a proxy for coronary accessibility), neoskirt height, TAV expansion at multiple frame levels, and residual valve area after redo-TAVR were compared to follow-up MSCT. RESULTS:Among 25 patients, HEARTguide simulations of valve-to-coronary (VTC) distance (left VTC 6.0 ± 2.2 vs. 6.4 ± 2.3 mm, ICC = 0.856, p < 0.001; right VTC 5.8 ± 1.9 vs. 5.6 ± 1.9 mm, ICC = 0.740, p < 0.001) showed good agreement while valve-to-sinotubular-junction (VTSTJ) distance showed moderate agreement with follow-up MSCT (left VTSTJ 2.5 ± 1.8 vs. 3.3 ± 1.8 mm, ICC = 0.614, p = 0.004; right VTSTJ 2.8 ± 1.8 vs. 2.9 ± 1.2 mm, ICC = 0.613, p = 0.003). Redo-TAVR frame expansion simulations (mean frame area 434.8 ± 141.6 vs. 401.6 ± 142.9 mm, ICC = 0.930, p < 0.001; 393.5 ± 63.2 vs. 383.1 ± 82.5 mm, ICC = 0.822, p < 0.001, respectively) and simulations of residual valve area after redo-TAVR showed good agreement for self-expanding and balloon-expandable TAV (407.8 ± 67.8 vs. 389.6 ± 50.4 mm, ICC = 0.837, p < 0.001; 358.0 ± 66.1 vs. 336.1 ± 79.9 mm, ICC = 0.823, p < 0.001). CONCLUSIONS:In this pilot study, MSCT-derived simulations of redo-TAVR provided accurate insights into coronary accessibility, TAV expansion, and residual valve area.

BACKGROUND:This study aims to assess how Watchman FLX (Boston Scientific) and Amplatzer Amulet (Abbott) devices for left atrial appendage occlusion (LAAO) impact the flow dynamics in the LA and the risk of device-related thrombosis (DRT). METHODS:Flow simulations were carried out for 12 AF patients before and after undergoing LAAO. Six patients had Watchman FLX while the other six had Amulet devices, of which three in each group had DRT. The flow parameters investigated included the LA velocities, pressure, time averaged wall shear stress (TAWSS), oscillatory shear index (OSI), and relative residence time (RRT). Low TAWSS and high OSI and RRT are generally associated with flow stasis and thrombosis. RESULTS:LA flow dynamics differed significantly between the two groups. On average, after LAAO, the Amulet patients exhibited a 25% increase in TAWSS and 24% decrease in RRT while the Watchman FLX patients exhibited, on average, a 24% decrease in TAWSS and a 23% increase in RRT. Both populations experienced an increase in LA pressure (~10% for Amulet patients and 16% for Watchman FLX patients). Increased DRT volume seems to be correlated with low TAWSS and elevated RRT. CONCLUSION:This preliminary study suggests that the Amulet device might be associated with improved LA flow dynamics and a potentially lower incidence of DRT compared with the Watchman. However, a larger clinical study is needed to confirm these findings.

BackgroundCoronary obstruction (CO) is a rare but devasting complication during transcatheter aortic valve replacement (TAVR).ObjectivesWe aim to demonstrate that the predicted distance between the coronary ostia and the closest structure derived with patient-specific computer simulation is associated with CO risk during TAVR.MethodsWe retrospectively analysed 14 aortic stenosis patients who underwent TAVR through finite element simulation. The frame deformation predicted with patient-specific computer simulation was qualitatively and quantitatively compared to the post-operative device deformation. The minimum distance between each coronary ostium and the closest structure was calculated and compared in patients who developed CO, at high risk of CO, and at no risk of CO.ResultsFour patients experienced CO during TAVR, 5 patients were at high risk of CO, and the remaining 5 patients had no risk of CO. A high coefficient of determination was obtained for all measurements extracted from the simulated device and the post-operative device (≥0.95). Simulations predicted shorter distance between the coronary ostium and the closest structure in patients who experienced CO, compared to patients at high risk of CO or who did not experience this complication (right coronary: 5.9 vs. 6.8 vs. 8.8 mm, left coronary: 3.0 vs. 3.3 vs. 6.5 mm respectively).ConclusionsThe distance between the coronary ostium and the closest structure was lower in patients who experienced CO during TAVR through patient-specific computational simulation. This technology enables coronary obstruction analysis before TAVR in the future.

BACKGROUND:The exclusion/occlusion of the left atrial appendage (LAA) is a treatment option for atrial fibrillation (AF) patients who are at high risk of stroke and high risk of bleeding. As the role of the LAA is not well understood or explored, this study aims to assess its role on flow dynamics in the left atrium. METHODS:Computational fluid dynamics (CFD) simulations were carried out for nine AF patients before and after LAA exclusion. The flow parameters investigated included the LA velocities, Time Averaged Wall Shear Stress (TAWSS), Oscillatory Shear Index (OSI), Relative Residence Time (RRT), and Pressure in the LA. RESULTS:This study shows that, on average, a decrease in TAWSS (1.82 ± 1.85 Pa to 1.27 ± 0.96 Pa, p < 0.05) and a slight increase in OSI (0.16 ± 0.10 to 0.17 ± 0.10, p < 0.05), RRT (1.87 ± 1.84 Pa-1 to 2.11 ± 1.78 Pa-1, p < 0.05), and pressure (-19.2 ± 6.8 mmHg to -15.3 ± 8.3 mmHg, p < 0.05) were observed in the LA after the exclusion of the LAA, with a decrease in low-magnitude velocities. CONCLUSION:The exclusion of the LAA seems to be associated with changes in LA flow dynamics. Further studies are needed to elucidate the clinical implications of these changes.

BACKGROUND Device-related thrombosis (DRT) occurs in up to 4% of patients undergoing left atrial appendage occlusion (LAAO) and is associated with substantial morbidity and mortality. However, its pathophysiology, predictors, and optimal management remain unclear. OBJECTIVES This study aims to assess flow dynamic factors correlating to DRT. METHODS A multicenter registry of patients who underwent LAAO and had pre-and post-computed tomography imaging was used. Patient-specific 3-dimensional digital models of the left atrium were created, and finite element simulations were performed to implant an LAAO device into each model in a position that matched the clinical deployment. Computational fluid dynamic simulations were performed to quantify the following flow dynamic parameters: time averaged wall shear stress, oscillatory shear index, and endothelial cell activation potential. RESULTS A total of 38 patients (19 with DRT and 19 without DRT) were included. Left atrium volumes and mitral valve areas were larger in the DRT cohort compared with controls. Patients with DRT had a significantly lower time averaged wall shear stress (1.76 +/- 1.24 Pa vs 2.90 +/- 2.70 Pa), a higher oscillatory shear index (0.19 +/- 0.11 vs 0.17 +/- 0.11), and a higher endothelial cell activation potential (0.23 +/- 0.58 Pa-1 vs 0.17 +/- 0.30 Pa-1) than the controls (P < 0.001 for all). Thrombus locations identified from in-vivo images correlated well with the flow dynamic parameters tested. CONCLUSIONS Flow dynamic parameters may be able to predict the risk of DRT after LAAO. Further investigation with a larger patient cohort and long-term follow-up is needed to assess the role of computational fluid dynamics in the risk stratification of patients considered for LAAO.

BackgroundThree-dimensional transesophageal echocardiography (3D-TEE) is the primary imaging tool for left atrial appendage closure planning. The utility of cardiac computed tomography angiography (CCTA) and patient-specific computational models is unknown.ObjectivesThe purpose of this study was to evaluate the accuracy of the FEops HEARTguide patient-specific computational modeling in predicting appropriate device size, location, and compression of the WATCHMAN FLX compared to intraprocedural 3D-TEE.MethodsPatients with both preprocedural and postprocedural CCTA and 3D-TEE imaging of the LAA who received a WATCHMAN FLX left atrial appendage closure device were studied (n = 22). The FEops HEARTguide platform used baseline CCTA imaging to generate a prediction of device size(s), device position(s), and device dimensions. Blinded (without knowledge of implanted device size/position) and unblinded (implant device size/position disclosed) simulations were evaluated.ResultsIn 16 (72.7%) patients, the blind simulation predicted the final implanted device size. In these patients, the 3D-TEE measurements were not significantly different and had excellent correlation (Pearson correlation coefficient (r) ≥ 0.90). No patients had peridevice leak after device implant. In the 6 patients for whom the model did not predict the implanted device size, a larger device size was ultimately implanted as per operator preference. The model measurements of the unblinded patients demonstrated excellent correlation with 3D-TEE.ConclusionsThis is the first study to demonstrate that the FEops HEARTguide model accurately predicts WATCHMAN FLX device implantation characteristics. Future studies are needed to evaluate if computational modeling can improve confidence in sizing, positioning, and compression of the device without compromising technical success.

We have investigated a specially grown Cu(In,Ga)Se2 (CIGSe) absorber, which was deposited by co-evaporation of Cu, In, Ga, and Se using a modified three stage process. Prior to the growth, the molybdenum-coated glass substrate was covered by a bent shroud made from tantalum (Ta), leading to a wedged absorber structure with a width of about 2 mm where the film thickness varies from 0 to 2 μm. In this region of interest the thickness dependency of morphology, concentration ratios and electronic properties was studied with secondary electron microscopy (SEM), X-ray fluorescence (XRF) and hard X-ray photoelectron spectroscopy (HAXPES), probing the CIGSe sample along the thickness gradient. The evidence of the thickness gradient itself was proven with SEM measurements in cross section geometry. By using XRF it was found that with decreasing film thickness the Cu concentration decreases significantly. This finding was also verified by HAXPES measurements. Furthermore, an enrichment of Ga towards the Mo back contact was found using the same technique. Besides these results the formation of a molybdenum selenide (MoSe) phase was observed on the fully covered part of the Mo coated substrate indicating a high mobility of Se on Mo under the given temperature conditions of the modified three stage deposition process.

We sought to assess the amount and distribution of force on the valve frame after transcatheter aortic valve replacement (TAVR) via patient-specific computer simulation. Patients successfully treated with the self-expanding Venus A-Valve and multislice computed tomography (MSCT) pre- and post-TAVR were retrospectively included. Patient-specific finite element models of the aortic root and prosthesis were constructed. The force (in Newton) on the valve frame was derived at every 3 mm from the inflow and at every 22.5° on each level. Twenty patients of whom 10 had bicuspid aortic valve (BAV) were analyzed. The total force on the frame was 74.9 N in median (interquartile range 24.0). The maximal force was observed at level 5 that corresponds with the nadir of the bioprosthetic leaflets and was 9.9 (7.1) N in all patients, 10.3 (6.6) N in BAV and 9.7 (9.2) N for patients with tricuspid aortic valve (TAV). The level of maximal force located higher from the native annulus in BAV and TAV patients (8.8 [4.8] vs. 1.8 [7.4] mm). The area of the valve frame at the level of maximal force decreased from 437.4 (239.7) mm2 at the annulus to 377.6 (114.3) mm2 in BAV, but increased from 397.5 (114.3) mm2 at the annulus to 406.7 (108.9) mm2 in TAV. The maximum force on the bioprosthetic valve frame is located at the plane of the nadir of the bioprosthetic leaflets. It remains to be elucidated whether this may be associated with bioprosthetic frame and leaflet integrity and/or function.

Transcatheter aortic valve implantation (TAVI) is associated with conduction abnormalities and the mechanical interaction between the prosthesis and the atrioventricular (AV) conduction path cause these life-threatening arrhythmias. Pre-operative assessment of the location of the AV conduction path can help to understand the risk of post-TAVI conduction abnormalities. As the AV conduction path is not visible on cardiac CT, the inferior border of the membranous septum can be used as an anatomical landmark. Detecting this border automatically, accurately and efficiently would save operator time and thus benefit pre-operative planning. This preliminary study was performed to identify the feasibility of 3D landmark detection in cardiac CT images with curriculum deep Q-learning. In this study, curriculum learning was used to gradually teach an artificial agent to detect this anatomical landmark from cardiac CT. This agent was equipped with a small field of view and burdened with a large action-space. Moreover, we introduced two novel action-selection strategies: a-decay and action-dropout. We compared these two strategies to the already established e-decay strategy and observed that a-decay yielded the most accurate results. Limited computational resources were used R ensure reproducibility. In order to maximize the amount of patient data, the method was cross-validated with k-folding for all three action-selection strategies. An inter-operator variability study was conducted to assess the accuracy of the method.

The mitral valve (MV) apparatus is a complex mechanical structure including annulus, valve leaflets, papillary muscles (PMs) and connected chordae tendineae. Chordae anchor to the papillary muscles to help the valve open and close properly during one cardiac cycle. It is of paramount importance to understand the functional, mechanical, and microstructural properties of mitral valve chordae and connecting PMs. In particular, little is known about the biomechanical properties of the anterior and posterior papillary muscle and corresponding chords. In this work, we performed uniaxial and biaxial tensile tests on the anterolateral (APM) and posteromedial papillary muscle (PPM), and their respective corresponding chordae tendineae, chordaeAPM and chordaePPM, in porcine hearts. Histology was carried out to link the microstructure and macro-mechanical behavior of the chordae and PMs. Our results demonstrate that chordaePPM are less in number, but significantly longer and stiffer than chordaeAPM. These different biomechanical properties may be partially explained by the higher collagen core ratio and larger collagen fibril density of chordaePPM. No significant mechanical or microstructural differences were observed along the circumferential and longitudinal directions of APM and PPM samples. Data measured on chordae and PMs were further fitted with the Ogden and reduced Holzapfel - Ogden strain energy functions, respectively. This study presents the first comparative anatomical, mechanical, and structural dataset of porcine mitral valve chordae and related PMs. Results indicate that a PM based classification of chordae will need to be considered in the analysis of the MV function or planning a surgical treatment, which will also help developing more precise computational models of MV.

Anatomic landmark detection is crucial during preoperative planning of transcatheter aortic valve implantation (TAVI) to select the proper device size and assess the risk of complications. The detection is currently a time-consuming manual process influenced by the image quality and subject to operator variability. In this work, we propose a novel automatic method to detect the relevant aortic landmarks from MDCT images using deep learning techniques. We trained three convolutional neural networks (CNNs) with 344 multidetector computed tomography (MDCT) acquisitions to detect five anatomical landmarks relevant for TAVI planning: the three basal attachment points of the aortic valve leaflets and the left and right coronary ostia. The detection strategy used these three CNN models to analyse a single MDCT image and yield three segmentation volumes as output. These segmentation volumes were averaged into one final segmentation volume, and the final predicted landmarks were obtained during a postprocessing step. Finally, we constructed the aortic annular plane, defined by the three predicted hinge points, and measured the distances from this plane to the predicted coronary ostia (i.e., coronary height). The methodology was validated on 100 patients. The automatic landmark detection was able to detect all the landmarks and showed high accuracy as the median distance between the ground truth and predictions is lower than the interobserver variations (1.5 mm [1.1–2.1], 2.0 mm [1.3–2.8] with a paired difference −0.5 ± 1.3 mm and p value <0.001). Furthermore, a high correlation is observed between predicted and manually measured coronary heights (for both R2 = 0.8). The image analysis time per patient was below one second. The proposed method is accurate, fast, and reproducible. Embedding this tool based on deep learning in the preoperative planning routine may have an impact in the TAVI environments by reducing the time and cost and improving accuracy.

The number of transcatheter aortic valve implantation (TAVI) procedures is expected to increase significantly in the coming years. Improving efficiency will become essential for experienced operators performing large TAVI volumes, while new operators will require training and may benefit from accurate support. In this work, we present a fast deep learning method that can predict aortic annulus perimeter and area automatically from aortic annular plane images. We propose a method combining two deep convolutional neural networks followed by a postprocessing step. The models were trained with 355 patients using modern deep learning techniques, and the method was evaluated on another 118 patients. The method was validated against an interoperator variability study of the same 118 patients. The differences between the manually obtained aortic annulus measurements and the automatic predictions were similar to the differences between two independent observers (paired diff. of 3.3 ± 16.8 mm2 vs. 1.3 ± 21.1 mm2 for the area and a paired diff. of 0.6 ± 1.7 mm vs. 0.2 ± 2.5 mm for the perimeter). The area and perimeter were used to retrieve the suggested prosthesis sizes for the Edwards Sapien 3 and the Medtronic Evolut device retrospectively. The automatically obtained device size selections accorded well with the device sizes selected by operator 1. The total analysis time from aortic annular plane to prosthesis size was below one second. This study showed that automated TAVI device size selection using the proposed method is fast, accurate, and reproducible. Comparison with the interobserver variability has shown the reliability of the strategy, and embedding this tool based on deep learning in the preoperative planning routine has the potential to increase the efficiency while ensuring accuracy.

Transcatheter aortic valve implantation has become an established procedure to treat severe aortic stenosis. Correct device sizing/positioning is crucial for optimal outcome. Lotus valve sizing is based upon multiple aortic root dimensions. Hence, it often occurs that two valve sizes can be selected. In this study, patient-specific computer simulation is adopted to evaluate the influence of Lotus size/position on paravalvular aortic regurgitation (AR) and conduction abnormalities, in patients with equivocal aortic root dimensions. First, simulation was performed in 62 patients to validate the model in terms of predicted AR and conduction abnormalities using postoperative echocardiographic, angiographic and ECG-based data. Then, two Lotus sizes were simulated at two positions in patients with equivocal aortic root dimensions. Large valve size and deep position were associated with higher contact pressure, while only large size, not position, significantly reduced the predicted AR. Despite general trends, simulations revealed that optimal device size/position is patient-specific.

PURPOSE:Ankle foot orthosis (AFO) stiffness is a key characteristic that determines how much support or restraint an AFO can provide. Thus, the goal of the current study is twofold: (1) to quantify AFO prescriptions for a group of patients; (2) to evaluate what impact these AFO have on the push-off phase.METHOD:Six patients were included in the study. Three patients were prescribed an AFO for ankle support and three patients were prescribed an AFO for ankle and knee support. Two types of AFO - a traditional polypropylene AFO (AFOPP) and a novel carbon-selective laser sintered polyamide AFO (AFOPA), were produced for each patient. AFO ankle stiffness was measured in a dedicated test rig. Gait analysis was performed under shod and orthotic conditions.RESULTS:Patient mass normalized AFOPP stiffness for ankle support ranged from 0.042 to 0.069 N·m·deg-1·kg-1, while for ankle and knee support it ranged from 0.081 to 0.127 N·m·deg-1·kg-1. On the group level, the ankle range of motion and mean ankle velocity in the push-off phase significantly decreased in both orthotic conditions, while peak ankle push-off power decreased non-significantly. Accordingly, on the group level, no significant improvements in walking speed were observed. However, after patient differentiation into good and bad responders it was found that in good responders peak ankle push-off power tended to be preserved and walking speed tended to increase.CONCLUSIONS:Quantification of AFO stiffness may help to understand why certain orthotic interventions are successful (unsuccessful) and ultimately lead to better AFO prescriptions. Implications for rehabilitation AFO ankle stiffness is key characteristic that determines how much support or restraint an AFO can provide. In a typical clinical setting, AFO ankle stiffness is not quantified. AFO has to meet individual patient's biomechanical needs. More objective AFO prescription and more controlled AFO production methods are needed to increase AFO success rate.

•Transcatheter mitral valve interventions are emerging as alternative for patients at high risk.•Anatomical analysis is vital and time-consuming during preoperative planning.•Automatic anatomical analysis is presented and validated on 71 patients.•The validation confirms the accuracy and efficiency of the automatic method.•An implementation is proposed to move towards safe and efficient preoperative planning.

Purpose This study proposes a new framework to optimize the design of a transcatheter aortic valve through patient-specific finite element and fluid dynamics simulation. Methods Two geometrical parameters of the frame, the diameter at ventricular inflow and the height of the first row of cells, were examined using the central composite design. The effect of those parameters on postoperative complications was investigated by response surface methodology, and a Nonlinear Programming by Quadratic Lagrangian algorithm was used in the optimization. Optimal and initial devices were then compared in 12 patients. The comparison was made in terms of device performance [i.e., reduced contact pressure on the atrioventricular conduction system and paravalvular aortic regurgitation (AR)]. Results Results suggest that large diameters and high cells favor higher anchoring of the device within the aortic root reducing the contact pressure and favor a better apposition of the device to the aortic root preventing AR. Compared to the initial device, the optimal device resulted in almost threefold lower predicted contact pressure and limited AR in all patients. Conclusions In conclusion, patient-specific modelling and simulation could help to evaluate device performance prior to the actual first-in-human clinical study and, combined with device optimization, could help to develop better devices in a shorter period.

Introduction Patients with severe functional and/or organic mitral regurgitation (MR) deemed unfit for traditional surgical mitral valve (MV) repair or MV replacement may be eligible for percutaneous edge-to-edge repair with the MitraClip® system (Abbott Vascular Structural Heart, Menlo Park, CA, USA)1. Procedural challenges such as optimal clip location placement, and the ideal number of clips to be placed are met using trial and error and are susceptible to procedural inefficacy resulting in residual MR or mitral stenosis. In this study we present a novel computer-based imaging software (FEops, Ghent, Belgium), which allows the creation of a patient-specific three-dimensional reconstruction of the MV. Using preprocedural and post-procedural transoesophageal echocardiography (TOE) images of two patients, we demonstrate an accurate model, which shows valve anatomy across the cardiac cycle, including mitral valve regurgitation. Using this software, future MitraClip procedures may be preceded by a virtual analysis of the MV, which will allow a careful assessment of the feasibility of repair, ideal clip location, and number of clips to be placed.

Heart failure is a progressive chronic condition in which the heart undergoes detrimental changes in structure and function across multiple scales in time and space. Multiscale models of cardiac growth can provide a patient-specific window into the progression of heart failure and guide personalized treatment planning. Yet, the predictive potential of cardiac growth models remains poorly understood. Here, we quantify predictive power of a stretch-driven growth model using a chronic porcine heart failure model, subject-specific multiscale simulation, and machine learning techniques. We combine hierarchical modeling, Bayesian inference, and Gaussian process regression to quantify the uncertainty of our experimental measurements during an 8-week long study of volume overload in six pigs. We then propagate the experimental uncertainties from the organ scale through our computational growth model and quantify the agreement between experimentally measured and computationally predicted alterations on the cellular scale. Our study suggests that stretch is the major stimulus for myocyte lengthening and demonstrates that a stretch-driven growth model alone can explain [Formula: see text] of the observed changes in myocyte morphology. We anticipate that our approach will allow us to design, calibrate, and validate a new generation of multiscale cardiac growth models to explore the interplay of various subcellular-, cellular-, and organ-level contributors to heart failure. Using machine learning in heart failure research has the potential to combine information from different sources, subjects, and scales to provide a more holistic picture of the failing heart and point toward new treatment strategies.

The purpose of this study was to create and validate a standardized framework for the evaluation of the ankle stiffness of two designs of 3D printed ankle foot orthoses (AFOs). The creation of four finite element (FE) models allowed patient-specific quantification of the stiffness and stress distribution over their specific range of motion during the second rocker of the gait. Validation was performed by comparing the model outputs with the results obtained from a dedicated experimental setup, which showed an overall good agreement with a maximum relative error of 10.38% in plantarflexion and 10.66% in dorsiflexion. The combination of advanced computer modelling algorithms and 3D printing techniques clearly shows potential to further improve the manufacturing process of AFOs.