scholarly journals Deep Learning Framework for Real-Time Estimation of in-silico Thrombotic Risk Indices in the Left Atrial Appendage

2021 ◽  
Vol 12 ◽  
Author(s):  
Xabier Morales Ferez ◽  
Jordi Mill ◽  
Kristine Aavild Juhl ◽  
Cesar Acebes ◽  
Xavier Iriart ◽  
...  
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Boundary Conditions ◽  
In Silico ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
Atrial Appendage ◽  
Superior Performance ◽  
Patient Specific ◽  
Cfd Simulations

Patient-specific computational fluid dynamics (CFD) simulations can provide invaluable insight into the interaction of left atrial appendage (LAA) morphology, hemodynamics, and the formation of thrombi in atrial fibrillation (AF) patients. Nonetheless, CFD solvers are notoriously time-consuming and computationally demanding, which has sparked an ever-growing body of literature aiming to develop surrogate models of fluid simulations based on neural networks. The present study aims at developing a deep learning (DL) framework capable of predicting the endothelial cell activation potential (ECAP), an in-silico index linked to the risk of thrombosis, typically derived from CFD simulations, solely from the patient-specific LAA morphology. To this end, a set of popular DL approaches were evaluated, including fully connected networks (FCN), convolutional neural networks (CNN), and geometric deep learning. While the latter directly operated over non-Euclidean domains, the FCN and CNN approaches required previous registration or 2D mapping of the input LAA mesh. First, the superior performance of the graph-based DL model was demonstrated in a dataset consisting of 256 synthetic and real LAA, where CFD simulations with simplified boundary conditions were run. Subsequently, the adaptability of the geometric DL model was further proven in a more realistic dataset of 114 cases, which included the complete patient-specific LA and CFD simulations with more complex boundary conditions. The resulting DL framework successfully predicted the overall distribution of the ECAP in both datasets, based solely on anatomical features, while reducing computational times by orders of magnitude compared to conventional CFD solvers.

scholarly journals Sensitivity Analysis of In Silico Fluid Simulations to Predict Thrombus Formation after Left Atrial Appendage Occlusion

Mathematics ◽  
2021 ◽  
Vol 9 (18) ◽  
pp. 2304
Author(s):  
Jordi Mill ◽  
Victor Agudelo ◽  
Andy L. Olivares ◽  
Maria Isabel Pons ◽  
Etelvino Silva ◽  
...  
Keyword(s):  
Sensitivity Analysis ◽  
Boundary Conditions ◽  
In Silico ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
Thrombus Formation ◽  
Atrial Appendage ◽  
Patient Specific ◽  
Ground Truth Data ◽  
Left Atrial Appendage Occlusion

Atrial fibrillation (AF) is nowadays the most common human arrhythmia and it is considered a marker of an increased risk of embolic stroke. It is known that 99% of AF-related thrombi are generated in the left atrial appendage (LAA), an anatomical structure located within the left atrium (LA). Left atrial appendage occlusion (LAAO) has become a good alternative for nonvalvular AF patients with contraindications to anticoagulants. However, there is a non-negligible number of device-related thrombus (DRT) events, created next to the device surface. In silico fluid simulations can be a powerful tool to better understand the relation between LA anatomy, haemodynamics, and the process of thrombus formation. Despite the increasing literature in LA fluid modelling, a consensus has not been reached yet in the community on the optimal modelling choices and boundary conditions for generating realistic simulations. In this line, we have performed a sensitivity analysis of several boundary conditions scenarios, varying inlet/outlet and LA wall movement configurations, using patient-specific imaging data of six LAAO patients (three of them with DRT at follow-up). Mesh and cardiac cycle convergence were also analysed. The boundary conditions scenario that better predicted DRT cases had echocardiography-based velocities at the mitral valve outlet, a generic pressure wave from an AF patient at the pulmonary vein inlets, and a dynamic mesh approach for LA wall deformation, emphasizing the need for patient-specific data for realistic simulations. The obtained promising results need to be further validated with larger cohorts, ideally with ground truth data, but they already offer unique insights on thrombogenic risk in the left atria.

scholarly journals In silico Optimization of Left Atrial Appendage Occluder Implantation Using Interactive and Modeling Tools

2019 ◽  
Vol 10 ◽  
Author(s):  
Ainhoa M. Aguado ◽  
Andy L. Olivares ◽  
Carlos Yagüe ◽  
Etelvino Silva ◽  
Marta Nuñez-García ◽  
...  
Keyword(s):  
In Silico ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
Atrial Appendage ◽  
Modeling Tools

P2467Fusion imaging (Anatomical Intelligence) enables automated left atrial appendage sizing in real-time a single center pilot study

2019 ◽  
Vol 40 (Supplement_1) ◽  
Author(s):  
S Afzal ◽  
V Veulemans ◽  
K Hellhammer ◽  
K Piayda ◽  
N Nijhof ◽  
...  
Keyword(s):  
Real Time ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
Fusion Imaging ◽  
Atrial Appendage ◽  
Patient Specific ◽  
Heart Model ◽  
Landing Zone ◽  
Device Selection ◽  
The Mean

Abstract Background and purpose Percutaneous left atrial appendage occlusion (LAAO) depicts an alternative treatment for patients with atrial fibrillation who are deemed for long-term oral anticoagulation therapy. In order to perform a successful LAAO accurate sizing of left atrial appendage (ostium, landing zone and depth) for device selection is essential. Echo-Fluoro fusion imaging in real-time offers with its latest prototype a patient-specific segmented automated 3D heart model and sizing of left atrial appendage (LAA). We therefore aimed to evaluate the automated segmented LAA sizing by comparing to 2D transesophageal (TOE) and MSCT measurements as gold standard. Methods We studied prospectively data of 8 consecutive patients who were admitted to our clinic for left atrial appendage closure. MSCT was performed preprocedural and analyzed with commercially available 3mensio software (Pie medical imaging). 2D TOE measurements and automated segmentation of the LAA and sizing were performed during the procedure by a highly experienced team of periinterventional cardiac imaging specialist and structural heart disease interventionalist who were blinded to the prior MSCT analysis. Dimension of ostium, landing zone (10 mm into the LAA parallel to the ostial plane at the level of the left circumflex for Amplatzer device) and depth (perpendicular to the ostial plane) were obtained in different TOE views according to instructions for use of Amulet Occluder. In order to generate an automated 3 D heart model, a high-quality 3D TOE image of the LAA volume and surrounding structures was acquired. After successful ECG-gated segmentation a 3 D heart model was generated. Automated LAA sizing followed in real-time. All measurements were taken into consideration before device selection. A Kruskal Wallis test was used to compare mean ranks of independent samples. A concordance analysis according to Kendall W was carried out to investigate reliability. Results The mean age of the patients was 82,6±4.15 years and half of the patients were female. All procedures were conducted successfully. The mean values of ostium and landing zone were comparable in TOE, automated sizing and MSCT sizing (ostium: 23,78±2,15 mm vs 25,71±5,25 mm vs 27,35±3,3 mm; (p=0,175); landing zone 22,13±3,18 mm vs 23,57±3,31 mm vs 24,00±3,51 mm; (p=0,377)). Furthermore, a significant concordance between the measurements was shown (ostium W= 0,991; p=0.045, landing zone W=0,835, p=0.014). Conclusion Automated LAA sizing acquired by fusion imaging may be an elegant real-time alternative for precise LAA Occluder device selection and needs to be investigated further.

Patient-specific catheter shaping for the minimally invasive closure of the left atrial appendage

2018 ◽  
Vol 13 (6) ◽  
pp. 837-846
Author(s):  
Eva C. Graf ◽  
Ilka Ott ◽  
Julian Praceus ◽  
Felix Bourier ◽  
Tim C. Lueth
Keyword(s):  
Minimally Invasive ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
Atrial Appendage ◽  
Patient Specific

Patient-specific 3D-printed Cardiac Model for Percutaneous Left Atrial Appendage Occlusion

2018 ◽  
Vol 71 (9) ◽  
pp. 762-764 ◽  
Author(s):  
Beatriz Vaquerizo ◽  
Carmen Escabias ◽  
Daniela Dubois ◽  
Gorka Gómez ◽  
Manuel Barreiro-Pérez ◽  
...  
Keyword(s):  
Left Atrial ◽  
Left Atrial Appendage ◽  
Atrial Appendage ◽  
Patient Specific ◽  
Left Atrial Appendage Occlusion ◽  
Cardiac Model ◽  
3D Printed

scholarly journals The Value of 3D Printing Models of Left Atrial Appendage Using Real-Time 3D Transesophageal Echocardiographic Data in Left Atrial Appendage Occlusion: Applications toward an Era of Truly Personalized Medicine

Cardiology ◽  
2016 ◽  
Vol 135 (4) ◽  
pp. 255-261 ◽  
Author(s):  
Peng Liu ◽  
Rijing Liu ◽  
Yan Zhang ◽  
Yingfeng Liu ◽  
Xiaoming Tang ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
3D Printing ◽  
Real Time ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
3D Models ◽  
Atrial Appendage ◽  
Patient Specific ◽  
The Real ◽  
Echocardiographic Image

Aims and Objectives: The objective of this study was to assess the clinical feasibility of generating 3D printing models of left atrial appendage (LAA) using real-time 3D transesophageal echocardiogram (TEE) data for preoperative reference of LAA occlusion. Background: Percutaneous LAA occlusion can effectively prevent patients with atrial fibrillation from stroke. However, the anatomical structure of LAA is so complicated that adequate information of its structure is essential for successful LAA occlusion. Emerging 3D printing technology has the demonstrated potential to structure more accurately than conventional imaging modalities by creating tangible patient-specific models. Typically, 3D printing data sets are acquired from CT and MRI, which may involve intravenous contrast, sedation, and ionizing radiation. It has been reported that 3D models of LAA were successfully created by the data acquired from CT. However, 3D printing of the LAA using real-time 3D TEE data has not yet been explored. Methods: Acquisition of 3D transesophageal echocardiographic data from 8 patients with atrial fibrillation was performed using the Philips EPIQ7 ultrasound system. Raw echocardiographic image data were opened in Philips QLAB and converted to ‘Cartesian DICOM' format and imported into Mimics® software to create 3D models of LAA, which were printed using a rubber-like material. The printed 3D models were then used for preoperative reference and procedural simulation in LAA occlusion. Results: We successfully printed LAAs of 8 patients. Each LAA costs approximately CNY 800-1,000 and the total process takes 16-17 h. Seven of the 8 Watchman devices predicted by preprocedural 2D TEE images were of the same sizes as those placed in the real operation. Interestingly, 3D printing models were highly reflective of the shape and size of LAAs, and all device sizes predicted by the 3D printing model were fully consistent with those placed in the real operation. Also, the 3D printed model could predict operating difficulty and the presence of a peridevice leak. Conclusions: 3D printing of the LAA using real-time 3D transesophageal echocardiographic data has a perfect and rapid application in LAA occlusion to assist with physician planning and decision making.

scholarly journals Demonstration of Patient-Specific Simulations to Assess Left Atrial Appendage Thrombogenesis Risk

2021 ◽  
Vol 12 ◽  
Author(s):  
Manuel García-Villalba ◽  
Lorenzo Rossini ◽  
Alejandro Gonzalo ◽  
Davis Vigneault ◽  
Pablo Martinez-Legazpi ◽  
...  
Keyword(s):  
Residence Time ◽  
Left Atrial ◽  
Left Atrial Appendage ◽  
Blood Stasis ◽  
Bleeding Risk ◽  
Atrial Appendage ◽  
Patient Specific ◽  
Immersed Boundary ◽  
Moving Wall ◽  
And Function

Atrial fibrillation (AF) alters left atrial (LA) hemodynamics, which can lead to thrombosis in the left atrial appendage (LAA), systemic embolism and stroke. A personalized risk-stratification of AF patients for stroke would permit improved balancing of preventive anticoagulation therapies against bleeding risk. We investigated how LA anatomy and function impact LA and LAA hemodynamics, and explored whether patient-specific analysis by computational fluid dynamics (CFD) can predict the risk of LAA thrombosis. We analyzed 4D-CT acquisitions of LA wall motion with an in-house immersed-boundary CFD solver. We considered six patients with diverse atrial function, three with either a LAA thrombus (removed digitally before running the simulations) or a history of transient ischemic attacks (LAAT/TIA-pos), and three without a LAA thrombus or TIA (LAAT/TIA-neg). We found that blood inside the left atrial appendage of LAAT/TIA-pos patients had marked alterations in residence time and kinetic energy when compared with LAAT/TIA-neg patients. In addition, we showed how the LA conduit, reservoir and booster functions distinctly affect LA and LAA hemodynamics. Finally, fixed-wall and moving-wall simulations produced different LA hemodynamics and residence time predictions for each patient. Consequently, fixed-wall simulations risk-stratified our small cohort for LAA thrombosis worse than moving-wall simulations, particularly patients with intermediate LAA residence time. Overall, these results suggest that both wall kinetics and LAA morphology contribute to LAA blood stasis and thrombosis.

Sign in / Sign up

Export Citation Format

Share Document