Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 15
Encontrada(s) 1 página(s)

Repositorio Digital de la UPF
Publication Contribución de congreso (ConferenceOutput)

In-silico analysis of the influence of pulmonary vein configuration on left atrial haemodynamics and thrombus formation in a large cohort

  • Mill, Jordi
  • Harrison, Josquin
  • Legghe, Benoit
  • Olivares, Andy L.
  • Morales, Xabier
  • Noailly, Jérôme
  • Iriart, Xavier
  • Cochet, Hubert
  • Sermesant, Maxime
  • Camara, Oscar
Comunicació presentada a: FIMH 2021 11th International Conference, celebrada del 21 al 25 de juny de 2021 a Stanford, CA, USA., Atrial fibrillation (AF) is considered the most common human arrhythmia. Around 99% of thrombi in non-valvular AF are formed in the left atrial appendage (LAA). Studies suggest that abnormal LAA haemodynamics and the subsequently stagnated flow are the factors triggering clot formation. However, the relation between LAA morphology, the blood pattern and the triggering is not fully understood. Moreover, the impact of structures such as the pulmonary veins (PVs) on LA haemodynamics has not been thoroughly studied due to the difficulties of acquiring appropriate data. On the other hand, in-silico studies and flow simulations allow a thorough analysis of haemodynamics, analysing the 4D nature of blood flow patterns under different boundary conditions. However, the reduced number of cases reported on the literature of these studies has been a limitation. The main goal of this work was to study the influence of PVs on left atrium (LA) and LAA haemodynamics. Computational fluid dynamics simulations were run on 52 patients, the largest cohort so far in the literature, where different parameters were individually studied: pulmonary veins orientation and configuration; LAA and LA volumes and its ratio; and flow velocities. Our computational analysis showed how the right pulmonary vein height and angulation have a great influence on LA haemodynamics. Additionally, we found that LAA with great bending with its tip pointing towards the mitral valve could contribute to favour flow stagnation., This work was supported by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the the Grants for the Contracting of New Research Staff Programme - FI (2020 FI B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502) and the Retos Investigaci´on project (RTI2018-101193-B-I00). Additionally, this work was supported by the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1- DTH-06-2020; grant agreement No. 101016496) and the European project PARIS (ID35).
Repositorio Digital de la UPF
Publication Contribución de congreso (ConferenceOutput)

Impact of blood rheological strategies on the optimization of patient-specific LAAO configurations for thrombus assessment

  • Albors, Carlos
  • Olivares Miyares, Andy Luis
  • Iriart, Xavier
  • Cochet, Hubert
  • Mill, Jordi
  • Camara, Oscar
Comunicació presentada a 12th International Conference, Functional Imaging and Modeling of the Heart, FIMH 2023, celebrada del 19 al 22 de juny de 2023 a Lió, França., Left atrial appendage occlusion devices (LAAO) are a feasible alternative for non-valvular atrial fibrillation (AF) patients at high risk of thromboembolic stroke and contraindication to antithrombotic therapies. However, optimal LAAO device configurations (i.e., size, type, location) remain unstandardized due to the large anatomical variability of the left atrial appendage (LAA) morphology, leading to a 4–6% incidence of device-related thrombus (DRT). In-silico simulations can be used to estimate the risk of DRT and identify the critical parameters, such as suboptimal device positioning. However, simulation outcomes depend a lot on a series of modelling assumptions such as blood behaviour. Therefore, in this work, we present fluid simulations results computed on two patient-specific LA geometries, using two different commercially available LAAO devices, located in two positions: 1) mimicking the real post-LAAO intervention configuration; and 2) an improved one better covering the pulmonary ridge for DRT prevention. Different blood modeling strategies were also tested. The results show flow re-circulations at low velocities with significant platelet accumulation in LAA-deep device positioning uncovering the pulmonary ridge, potentially leading to thrombus formation. In addition, assuming Newtonian blood behaviour may result in an overestimation of DRT risk., This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement, No 101016496 (SimCardioTest).
Repositorio Digital de la UPF
Publication Contribución de congreso (ConferenceOutput)

Unsupervised machine learning exploration of morphological and haemodynamic indices to predict thrombus formation in the left atrial appendage

  • Saiz-Vivó, Marta
  • Mill, Jordi
  • Harrison, Josquin
  • Jimenez-Perez, Guillermo
  • Legghe, Benoit
  • Iriart, Xavier
  • Cochet, Hubert
  • Piella Fenoy, Gemma
  • Sermesant, Maxime
  • Camara, Oscar
Comunicació presentada a 13th International Workshop, STACOM 2022, conjuntament amb MICCAI 2022, celebrada el 18 de setembre de 2022 a Singapur., Atrial Fibrillation (AF) is the most common cardiac arrhythmia, and it is associated with an increased risk of embolic stroke. It is known that AF-related thrombus formation occurs predominantly in the left atrial appendage (LAA). However, it is still unknown the structural and functional characteristics of the left atria (LA) that promote low velocities and stagnated blood flow, thus a high risk of thrombogenesis. In this work, we investigated morphological and in-silico haemodynamic indices of the LA and LAA with unsupervised machine learning (ML) techniques, to identify the most relevant features that could subsequently be used to generate thrombus prediction models. A fully automatic pipeline was implemented to extract multiple morphological parameters from a 3D mesh of a LA. Morphological parameters were then combined with particle flow parameters from in-silico fluid simulations. Unsupervised multiple kernel learning (MKL) was used for dimensionality reduction, resulting in a latent space positioning patients based on feature similarity. Clustering applied to the MKL output space estimated clusters with different proportion of thrombus cases. The cluster with the highest risk of thrombus formation was characterised by high values of LAA height, tortuosity and ostium perimeter, as well as total number of flow particles in the LAA and low angle between the LAA and the left superior pulmonary vein, proving the usefulness of unsupervised ML techniques to extract knowledge from the data, and early identify AF patients at higher risk of thrombus formation., This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101016496 (SimCardioTest).
Repositorio Digital de la UPF
Publication Contribución de congreso (ConferenceOutput)

Geometric deep learning for the assessment of thrombosis risk in the left atrial appendage

  • Morales, Xabier
  • Mill, Jordi
  • Simeon, Guillem
  • Juhl, Kristine Aavild
  • De Backer, Ole
  • Paulsen, Rasmus R.
  • Camara, Oscar
Comunicació presentada a: FIMH 2021 11th International Conference, celebrada del 21 al 25 de juny de 2021 a Stanford, CA, USA., The assessment of left atrial appendage (LAA) thrombogenesis has experienced major advances with the adoption of patient-specific computational fluid dynamics (CFD) simulations. Nonetheless, due to the vast computational resources and long execution times required by fluid dynamics solvers, there is an ever-growing body of work aiming to develop surrogate models of fluid flow simulations based on neural networks. The present study builds on this foundation by developing a deep learning (DL) framework capable of predicting the endothelial cell activation potential (ECAP), linked to the risk of thrombosis, solely from the patient-specific LAA geometry. To this end, we leveraged recent advancements in Geometric DL, which seamlessly extend the unparalleled potential of convolutional neural networks (CNN), to non-Euclidean data such as meshes. The model was trained with a dataset combining 202 synthetic and 54 real LAA, predicting the ECAP distributions instantaneously, with an average mean absolute error of 0.563. Moreover, the resulting framework manages to predict the anatomical features related to higher ECAP values even when trained exclusively on synthetic cases., This work was supported by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the the Grants for the Contracting of New Research Staff Programme - FI (2020 FI B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502) and the Retos Investigaci´on project (RTI2018-101193-B-I00). Additionally, this work was supported by the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1- DTH-06-2020; grant agreement No. 101016496).
Repositorio Digital de la UPF
Publication Contribución de congreso (ConferenceOutput)

Sensitivity analysis of left atrial wall modeling approaches and inlet/outlet boundary conditions in fluid simulations to predict thrombus formation

  • Albors, Carlos
  • Mill, Jordi
  • Kjeldsberg, Henrik A
  • Viladés-Medel, David
  • Olivares Miyares, Andy Luis
  • Valen-Sendstad, Kristian
  • Camara, Oscar
Comunicació presentada a 13th International Workshop, STACOM 2022, conjuntament amb MICCAI 2022, celebrada el 18 de setembre de 2022 a Singapur., In-silico fluid simulations of the left atria (LA) in atrial fibrillation (AF) patients can help to describe and relate patient-specific morphologies and complex flow haemodynamics with the pathophysiological mechanisms behind thrombus formation. Even in AF patients, LA wall motion plays a non-negligible role in LA function and blood flow patterns. However, obtaining 4D LA wall dynamics from patient-specific data is not an easy task due to current image resolution limitations. Therefore, several approaches have been proposed in the literature to include left atrial wall motion in fluid simulations, being necessary to benchmark them in relation to their estimations on thrombogenic risk. In this work, we present results obtained with computational fluid dynamic simulations of LA geometries from a control and an AF patient with different left atrial wall motion approaches: 1) assuming rigid walls; 2) with a passive movement of the mitral valve annulus plane from a dynamic mesh approach based on springs (DM-SB); and 3) imposing LA wall deformation extracted from dynamic computed tomography (DM-dCT) images. Different strategies for the inlet/outlet boundary conditions were also tested. The DM-dCT approach was the one providing simulation results closer to velocity curves extracted from the reference echocardiographic data, whereas the rigid wall strategy over-estimated the risk of thrombus formation., This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101016496 (SimCardioTest).
Repositorio Digital de la UPF
Publication Contribución de congreso (ConferenceOutput)

Towards real-time optimization of left atrial appendage occlusion device placement through physics-informed neural networks

  • Morales, Xabier
  • Albors, Carlos
  • Mill, Jordi
  • Camara, Oscar
Comunicació presentada a 13th International Workshop, STACOM 2022, conjuntament amb MICCAI 2022, celebrada el 18 de setembre de 2022 a Singapur., The adoption of patient-specific computational fluid dynamics (CFD) simulations has been instrumental toward a better understanding of the mechanisms underlying thrombogenesis in the left atrial appendage. Such simulations can help optimize the placement of left atrial appendage occlusion (LAAO) devices in atrial fibrillation patients and avoid the generation of device-related thrombosis. However, integrating conventional solvers into clinical practice is cumbersome, as even the slightest change in model geometry involves computing the entire simulation from scratch. In contrast, neural networks can entirely circumvent this issue by transferring knowledge across models targeted at similar physical domains. Thus, in the present study, we introduced a neural network capable of predicting left atrial hemodynamics under different occlusion device configurations, relying solely on a single finite element simulation for training. To this end, we leveraged physics-informed neural networks (PINN), which embed the physical laws governing the domain of interest into the model, exhibiting far superior generalization capabilities than conventional data-driven models. Several device types and positions have been tested in two distinct left atrial geometries. By employing a single reference simulation per patient the network can predict the updated hemodynamics for a variety of device types and positions, orders of magnitude faster than with conventional CFD solvers., This work was funded by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the Grants for the Contracting of New Research Staff Programme - FI (2020 FI_B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), and the Retos Investigación project (RTI2018-101193-B-I00), and the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1-DTH-06-2020; grant agreement No. 101016496). Additionally, this research was supported by grants from NVIDIA and utilized NVIDIA RTX A6000.
Repositorio Digital de la UPF
Publication Contribución de congreso (ConferenceOutput)

In-silico analysis of device-related thrombosis for different left atrial appendage occluder settings

  • Planas, Eric
  • Mill, Jordi
  • Olivares, Andy L.
  • Morales, Xabier
  • Pons Vidal, Maria Isabel
  • Iriart, Xavier
  • Cochet, Hubert
  • Camara, Oscar
Comunicació presentada a: 12th International Workshop, STACOM 2021 celebrada del 27 de setembre de 2021 a Strasbourg, França., Atrial fibrillation (AF) is one of the most common cardiac arrhythmias and is associated to an increasing risk of stroke. Most AF- related strokes are formed in the left atrial appendage (LAA). To prevent thrombus formation, LAA occlusion (LAAO) is considered a suitable alternative for AF patients with contraindications for anti-coagulation treatment. Nevertheless, LAAO is linked to a non-negligible risk of gener- ating thrombus at the surface near the device (i.e., device-related throm- bus, DRT), depending on the implantation settings. For instance, it has been shown that not covering the pulmonary ridge (PR) with the LAAO increases the risk of DRT. In-silico analysis is a useful tool to better understand the blood flow patterns after LAAO and predict the risk of DRT for a given patient and device configuration. In the present work we designed a modelling pipeline based on fluid simulations, including a thrombus model using discrete phase modelling, to analyse the risk of DRT in six patient-specific geometries for different LAAO settings. In particular, we studied the possible incidence of DRT depending on the device positioning (covering/uncovering the PR) and type (Amplatzer Amulet and Watchman FLX). The resulting in-silico indices demon- strated that covering the PR entails less thrombogenic patterns than uncovering it. In our study, disk-based devices had better adaptability to complex LAA morphologies and a slightly minor associated risk of DRT than non-disk devices., This work was supported by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the the Grants for the Contracting of New Research Staff Programme - FI (2020 FI B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502) and the Retos Inves- tigaci ́on project (RTI2018-101193-B-I00). Additionally, this work was supported by the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1-DTH-06-2020; grant agreement No. 101016496).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Contribución de congreso (ConferenceOutput)

Simulación de los efectos de fármacos antiarrítmicos en la electrofisiología auricular

  • Puche-García, Violeta|||0000-0002-3386-1478
  • Romero Pérez, Lucia|||0000-0003-4605-8630
  • Saiz Rodríguez, Francisco Javier|||0000-0002-9850-0825
[ES] La terapia farmacológica es la primera línea contra la fibrilación auricular; aunque los mecanismos no se comprenden por completo. Nuestro objetivo es comparar datos experimentales a nivel unicelular de varios fármacos con la respuesta de tres modelos computacionales diferentes: Courtemanche (CRN), Grandi y Skibsbye (SKI) empleando un modelo de bloqueo de poro simple, así como una variante de modelo dinámico para los fármacos flecainida y vernakalant en CRN (+INa,Marvok). La mayoría de las simulaciones consiguen reproducir la dependencia del aumento del APD90 y la disminución de la dV/dtmax con la frecuencia y la concentración. Además, las respuestas de CRN, CRN+INa,Marvok y SKI reprodujeron de manera más adecuada las tendencias encontradas en los resultados experimentales. El análisis realizado contribuye a la mejora de la simulación de fármacos antiarrítmicos., Este trabajo ha sido financiado por el Programa Prometeo de la Generalitat Valenciana (2020/043) y por el programa de investigación e innovación Horizon 2020 de la Unión Europea bajo subvención No 101016496 (SimCardioTest).
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Sensitivity analysis of left atrial wall modeling approaches and inlet/outlet boundary conditions in fluid simulations to predict thrombus formation

  • Albors, Carlos
  • Mill, Jordi
  • Kjeldsberg, Henrik A
  • Viladés-Medel, David
  • Olivares Miyares, Andy Luis
  • Valen-Sendstad, Kristian
  • Camara, Oscar
Comunicació presentada a 13th International Workshop, STACOM 2022, conjuntament amb MICCAI 2022, celebrada el 18 de setembre de 2022 a Singapur., In-silico fluid simulations of the left atria (LA) in atrial fibrillation (AF) patients can help to describe and relate patient-specific morphologies and complex flow haemodynamics with the pathophysiological mechanisms behind thrombus formation. Even in AF patients, LA wall motion plays a non-negligible role in LA function and blood flow patterns. However, obtaining 4D LA wall dynamics from patient-specific data is not an easy task due to current image resolution limitations. Therefore, several approaches have been proposed in the literature to include left atrial wall motion in fluid simulations, being necessary to benchmark them in relation to their estimations on thrombogenic risk. In this work, we present results obtained with computational fluid dynamic simulations of LA geometries from a control and an AF patient with different left atrial wall motion approaches: 1) assuming rigid walls; 2) with a passive movement of the mitral valve annulus plane from a dynamic mesh approach based on springs (DM-SB); and 3) imposing LA wall deformation extracted from dynamic computed tomography (DM-dCT) images. Different strategies for the inlet/outlet boundary conditions were also tested. The DM-dCT approach was the one providing simulation results closer to velocity curves extracted from the reference echocardiographic data, whereas the rigid wall strategy over-estimated the risk of thrombus formation., This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101016496 (SimCardioTest).
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

In-silico analysis of the influence of pulmonary vein configuration on left atrial haemodynamics and thrombus formation in a large cohort

  • Mill, Jordi
  • Harrison, Josquin
  • Legghe, Benoit
  • Olivares, Andy L.
  • Morales, Xabier
  • Noailly, Jérôme
  • Iriart, Xavier
  • Cochet, Hubert
  • Sermesant, Maxime
  • Camara, Oscar
Comunicació presentada a: FIMH 2021 11th International Conference, celebrada del 21 al 25 de juny de 2021 a Stanford, CA, USA., Atrial fibrillation (AF) is considered the most common human arrhythmia. Around 99% of thrombi in non-valvular AF are formed in the left atrial appendage (LAA). Studies suggest that abnormal LAA haemodynamics and the subsequently stagnated flow are the factors triggering clot formation. However, the relation between LAA morphology, the blood pattern and the triggering is not fully understood. Moreover, the impact of structures such as the pulmonary veins (PVs) on LA haemodynamics has not been thoroughly studied due to the difficulties of acquiring appropriate data. On the other hand, in-silico studies and flow simulations allow a thorough analysis of haemodynamics, analysing the 4D nature of blood flow patterns under different boundary conditions. However, the reduced number of cases reported on the literature of these studies has been a limitation. The main goal of this work was to study the influence of PVs on left atrium (LA) and LAA haemodynamics. Computational fluid dynamics simulations were run on 52 patients, the largest cohort so far in the literature, where different parameters were individually studied: pulmonary veins orientation and configuration; LAA and LA volumes and its ratio; and flow velocities. Our computational analysis showed how the right pulmonary vein height and angulation have a great influence on LA haemodynamics. Additionally, we found that LAA with great bending with its tip pointing towards the mitral valve could contribute to favour flow stagnation., This work was supported by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the the Grants for the Contracting of New Research Staff Programme - FI (2020 FI B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502) and the Retos Investigaci´on project (RTI2018-101193-B-I00). Additionally, this work was supported by the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1- DTH-06-2020; grant agreement No. 101016496) and the European project PARIS (ID35).
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

In-silico analysis of device-related thrombosis for different left atrial appendage occluder settings

  • Planas, Eric
  • Mill, Jordi
  • Olivares, Andy L.
  • Morales, Xabier
  • Pons Vidal, Maria Isabel
  • Iriart, Xavier
  • Cochet, Hubert
  • Camara, Oscar
Comunicació presentada a: 12th International Workshop, STACOM 2021 celebrada del 27 de setembre de 2021 a Strasbourg, França., Atrial fibrillation (AF) is one of the most common cardiac arrhythmias and is associated to an increasing risk of stroke. Most AF- related strokes are formed in the left atrial appendage (LAA). To prevent thrombus formation, LAA occlusion (LAAO) is considered a suitable alternative for AF patients with contraindications for anti-coagulation treatment. Nevertheless, LAAO is linked to a non-negligible risk of gener- ating thrombus at the surface near the device (i.e., device-related throm- bus, DRT), depending on the implantation settings. For instance, it has been shown that not covering the pulmonary ridge (PR) with the LAAO increases the risk of DRT. In-silico analysis is a useful tool to better understand the blood flow patterns after LAAO and predict the risk of DRT for a given patient and device configuration. In the present work we designed a modelling pipeline based on fluid simulations, including a thrombus model using discrete phase modelling, to analyse the risk of DRT in six patient-specific geometries for different LAAO settings. In particular, we studied the possible incidence of DRT depending on the device positioning (covering/uncovering the PR) and type (Amplatzer Amulet and Watchman FLX). The resulting in-silico indices demon- strated that covering the PR entails less thrombogenic patterns than uncovering it. In our study, disk-based devices had better adaptability to complex LAA morphologies and a slightly minor associated risk of DRT than non-disk devices., This work was supported by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the the Grants for the Contracting of New Research Staff Programme - FI (2020 FI B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502) and the Retos Inves- tigaci ́on project (RTI2018-101193-B-I00). Additionally, this work was supported by the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1-DTH-06-2020; grant agreement No. 101016496).
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Impact of blood rheological strategies on the optimization of patient-specific LAAO configurations for thrombus assessment

  • Albors, Carlos
  • Olivares Miyares, Andy Luis
  • Iriart, Xavier
  • Cochet, Hubert
  • Mill, Jordi
  • Camara, Oscar
Comunicació presentada a 12th International Conference, Functional Imaging and Modeling of the Heart, FIMH 2023, celebrada del 19 al 22 de juny de 2023 a Lió, França., Left atrial appendage occlusion devices (LAAO) are a feasible alternative for non-valvular atrial fibrillation (AF) patients at high risk of thromboembolic stroke and contraindication to antithrombotic therapies. However, optimal LAAO device configurations (i.e., size, type, location) remain unstandardized due to the large anatomical variability of the left atrial appendage (LAA) morphology, leading to a 4–6% incidence of device-related thrombus (DRT). In-silico simulations can be used to estimate the risk of DRT and identify the critical parameters, such as suboptimal device positioning. However, simulation outcomes depend a lot on a series of modelling assumptions such as blood behaviour. Therefore, in this work, we present fluid simulations results computed on two patient-specific LA geometries, using two different commercially available LAAO devices, located in two positions: 1) mimicking the real post-LAAO intervention configuration; and 2) an improved one better covering the pulmonary ridge for DRT prevention. Different blood modeling strategies were also tested. The results show flow re-circulations at low velocities with significant platelet accumulation in LAA-deep device positioning uncovering the pulmonary ridge, potentially leading to thrombus formation. In addition, assuming Newtonian blood behaviour may result in an overestimation of DRT risk., This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement, No 101016496 (SimCardioTest).
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Towards real-time optimization of left atrial appendage occlusion device placement through physics-informed neural networks

  • Morales, Xabier
  • Albors, Carlos
  • Mill, Jordi
  • Camara, Oscar
Comunicació presentada a 13th International Workshop, STACOM 2022, conjuntament amb MICCAI 2022, celebrada el 18 de setembre de 2022 a Singapur., The adoption of patient-specific computational fluid dynamics (CFD) simulations has been instrumental toward a better understanding of the mechanisms underlying thrombogenesis in the left atrial appendage. Such simulations can help optimize the placement of left atrial appendage occlusion (LAAO) devices in atrial fibrillation patients and avoid the generation of device-related thrombosis. However, integrating conventional solvers into clinical practice is cumbersome, as even the slightest change in model geometry involves computing the entire simulation from scratch. In contrast, neural networks can entirely circumvent this issue by transferring knowledge across models targeted at similar physical domains. Thus, in the present study, we introduced a neural network capable of predicting left atrial hemodynamics under different occlusion device configurations, relying solely on a single finite element simulation for training. To this end, we leveraged physics-informed neural networks (PINN), which embed the physical laws governing the domain of interest into the model, exhibiting far superior generalization capabilities than conventional data-driven models. Several device types and positions have been tested in two distinct left atrial geometries. By employing a single reference simulation per patient the network can predict the updated hemodynamics for a variety of device types and positions, orders of magnitude faster than with conventional CFD solvers., This work was funded by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the Grants for the Contracting of New Research Staff Programme - FI (2020 FI_B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), and the Retos Investigación project (RTI2018-101193-B-I00), and the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1-DTH-06-2020; grant agreement No. 101016496). Additionally, this research was supported by grants from NVIDIA and utilized NVIDIA RTX A6000.
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Geometric deep learning for the assessment of thrombosis risk in the left atrial appendage

  • Morales, Xabier
  • Mill, Jordi
  • Simeon, Guillem
  • Juhl, Kristine Aavild
  • De Backer, Ole
  • Paulsen, Rasmus R.
  • Camara, Oscar
Comunicació presentada a: FIMH 2021 11th International Conference, celebrada del 21 al 25 de juny de 2021 a Stanford, CA, USA., The assessment of left atrial appendage (LAA) thrombogenesis has experienced major advances with the adoption of patient-specific computational fluid dynamics (CFD) simulations. Nonetheless, due to the vast computational resources and long execution times required by fluid dynamics solvers, there is an ever-growing body of work aiming to develop surrogate models of fluid flow simulations based on neural networks. The present study builds on this foundation by developing a deep learning (DL) framework capable of predicting the endothelial cell activation potential (ECAP), linked to the risk of thrombosis, solely from the patient-specific LAA geometry. To this end, we leveraged recent advancements in Geometric DL, which seamlessly extend the unparalleled potential of convolutional neural networks (CNN), to non-Euclidean data such as meshes. The model was trained with a dataset combining 202 synthetic and 54 real LAA, predicting the ECAP distributions instantaneously, with an average mean absolute error of 0.563. Moreover, the resulting framework manages to predict the anatomical features related to higher ECAP values even when trained exclusively on synthetic cases., This work was supported by the Agency for Management of University and Research Grants of the Generalitat de Catalunya under the the Grants for the Contracting of New Research Staff Programme - FI (2020 FI B 00608) and the Spanish Ministry of Economy and Competitiveness under the Programme for the Formation of Doctors (PRE2018-084062), the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502) and the Retos Investigaci´on project (RTI2018-101193-B-I00). Additionally, this work was supported by the H2020 EU SimCardioTest project (Digital transformation in Health and Care SC1- DTH-06-2020; grant agreement No. 101016496).
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Unsupervised machine learning exploration of morphological and haemodynamic indices to predict thrombus formation in the left atrial appendage

  • Saiz-Vivó, Marta
  • Mill, Jordi
  • Harrison, Josquin
  • Jimenez-Perez, Guillermo
  • Legghe, Benoit
  • Iriart, Xavier
  • Cochet, Hubert
  • Piella Fenoy, Gemma
  • Sermesant, Maxime
  • Camara, Oscar
Comunicació presentada a 13th International Workshop, STACOM 2022, conjuntament amb MICCAI 2022, celebrada el 18 de setembre de 2022 a Singapur., Atrial Fibrillation (AF) is the most common cardiac arrhythmia, and it is associated with an increased risk of embolic stroke. It is known that AF-related thrombus formation occurs predominantly in the left atrial appendage (LAA). However, it is still unknown the structural and functional characteristics of the left atria (LA) that promote low velocities and stagnated blood flow, thus a high risk of thrombogenesis. In this work, we investigated morphological and in-silico haemodynamic indices of the LA and LAA with unsupervised machine learning (ML) techniques, to identify the most relevant features that could subsequently be used to generate thrombus prediction models. A fully automatic pipeline was implemented to extract multiple morphological parameters from a 3D mesh of a LA. Morphological parameters were then combined with particle flow parameters from in-silico fluid simulations. Unsupervised multiple kernel learning (MKL) was used for dimensionality reduction, resulting in a latent space positioning patients based on feature similarity. Clustering applied to the MKL output space estimated clusters with different proportion of thrombus cases. The cluster with the highest risk of thrombus formation was characterised by high values of LAA height, tortuosity and ostium perimeter, as well as total number of flow particles in the LAA and low angle between the LAA and the left superior pulmonary vein, proving the usefulness of unsupervised ML techniques to extract knowledge from the data, and early identify AF patients at higher risk of thrombus formation., This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101016496 (SimCardioTest).