Barcelona VPH Summer School 2018

June 18th - 22nd, 2018

The 3rd VPH Summer School will be held in Barcelona, Spain, on June 18-22, 2018. This 3rd edition will focus on data integration, model verification and validation. Leave your e-mail here to get to know the updates in the program.

The VPH Summer School series is co-organized by the Barcelona MedTech - Universitat Pompeu Fabra and by the Virtual Physiological Human Institute. It aims to provide junior engineers and medical doctors with a complete overview of state-of-the-art VPH research, following a complete pipeline from basic science and clinical needs, to model application.The Summer School is largely supported by the Marie Curie ITN CardioFunXion, collaborates with CompBioMed Centre of Excellence in Computational Biomedicine (talks and hands-on), the Chair UPF-QUAES: Computational Technologies for health (Talks and hands-on award) and VHeartSN, the Spanish Network of Research in Cardiac Computational Modeling (talks).

Location: UPF Poble Nou Campus (Roc Boronat, 138, Barcelona)

This Summer School provides a thorough overview and hands-on experience in state-of-the-art Virtual Physiological Human (VPH) research. The key concepts of this methodological and technological framework will be presented using illustrative cases and enriched with hands-on analysis under supervision of the experts.

The Summer School includes a poster session with a best poster award (sponsored by Simula last year), and a hands-on award provided by Chair QUAES-UPF.

This 3rd edition will focus on data integration, model verification and validation.

PROGRAM OVERVIEW

Hands-on sessions (each participant selects one)

 
 

 

 

DETAILED PROGRAM

Speakers

Day 1: Pathophysiology. Basic Science and Clinical Understanding

Special focus 3rd Edition: Biological and clinical problems that require collecting, handling and interpreting large amounts of data

Keynote speaker 1 (9:30h - 11:00h). Using omics to understand common problems in medicine: Chronic pain

 
Frances Williams (King's College London / Guy’s and St Thomas’ NHS Foundation Trust)
 
Abstract

Modern medicine is currently ill equipped to deal with the highly prevalent symptom of chronic pain. Seen in all medical specialties, chronic pain syndromes are poorly understood. They have, however, been shown to have a heritable basis. The use of omic data derived from identical and non-identical same sex twins allows a greater understanding of the role of genetic and environmental factors in chronic pain conditions.

Chronic widespread musculoskeletal pain (CWP) and fibromyalgia are associated with a number of other traits such as anxiety and depression, pain catastrophizing and other medical and psychological diagnoses. In this talk the role of genetics in CWP will be discussed and the results of recent large scale epidemiological studies will be covered. Particular reference will be made to new work exploring the of association of CWP with cardiovascular disease.

Short bio

I have been at King’s College London for 10 years and my position allows me to combine the study of chronic pain with the practice of musculoskeletal medicine. In particular, I run clinics in General Rheumatology and Metabolic bone disease. I also have an interest in occupational conditions and hold a monthly clinic for Musicians and Performing Artists with musculoskeletal complaints, which is unique in the UK and is funded by the NHS.

The aim of my group is to understand better the pathogenic processes underlying common complex traits. In particular we focus on those presenting to rheumatology such as chronic widespread pain and chronic back pain. The study of a broad range of omics, in isolation and in combination, will reveal more about the pathways involved in chronic pain, in particular the transition from acute to chronic pain. I lead the EU FP7 project PainOmics which examines biomarkers of the transition from acute to chronic low back pain. I am also contributing to several CHARGE consortium efforts to identify genetic variants contributing to back pain and age related hearing impairment.

I am married with 3 children and in my spare time I enjoy walking the dog, running and going to the opera.

--------------------------------

Talk 1 (11:30-12:30): The ins and outs of cardiac myocytes

 

 

Gudrun Antoons (Maastricht University)

Short bio

Gudrun Antoons studied biology, and specialized in cellular cardiac electrophysiology during her PhD training at the University of Leuven. She received her doctoral degree in Medical Sciences in 2003. She continued her career at Utrecht University, where she received a VENI fellowship to investigate cellular mechanisms of electrical instability related to calcium abnormalities as biomarkers of pro-arrhythmia. Her post-doc work also included the investigation of Na+/Ca2+ exchanger and Na+ channels as anti-arrhythmic targets in various forms of ventricular arrhythmias. In 2011, she was appointed as associated professor at the Medical University of Graz and participated as investigator in a EU program on atrial fibrillation. At the end of 2014, she joined the cardiovascular research institute in Maastricht and continued working on basic mechanisms of atrial and ventricular arrhythmias.

--------------------------------

Talk 2 (12:30-13:30): Pathophysiology & clinical management of epilepsy

Rodrigo Rocamora (Epilepsy Reference Center, Hospital del Mar)

Abstract

Epileptic disorders are devastating multi-causal chronic disorders characterized by recurrent spontaneous seizures affecting around 60 million people worldwide. Approximately one third of these patients present drug resistant epilepsy (DRE). The burden of epilepsy stems from a multitude of issues including DRE and comorbidities leading to a huge societal cost. The most efficient treatment of DRE is resective surgery. Surgical resection of the epileptogenic area, a localized region where seizures arise offers potentially curative procedures for DRE, but for a relevant number of these patients the correct localization of the area to target remains a challenging problem. Moreover, epilepsy surgery remains underused and has not improved its outcome rate of 50% of seizure-free patients for the past 50 years. Although substantial progress has been made, the precise localization of the areas to target still remain a big challenge also because of the lack of mechanistic insights and precise biomarkers. The development of personalized brain computer models promises to change this poor patient management success rate.
 
Short bio
 
Prof. Dr. Rocamora founded in 2009 the Epilepsy Monitoring Unit of the Hospital del Mar (EMU-HM). Under his direction, the EMU-HM became one of the most advanced epilepsy centres in Spain. Through the fast structural and scientific development, CatSalut awarded the Epilepsy Unit in 2012 as a regional reference centre (UFCE) for complex epilepsies in Catalonia. In 2013, the EMU-HM was recognized by the Spanish Health Ministry (MSSSI) as
national reference centre (CSUR) for epilepsy surgery being until now the only unit at Hospital del Mar with this qualification of excellence.

--------------------------------

Presentation of sponsors and supporters

Presentation of VPHI, ESB and Chair UPF-QUAES (13:30 - 13:45)

Presentation Master Computational Biomedical Engineering (14:45 - 14:55)

Day 2: Input measurements. Acquisition, processing, quantification

Special focus 3rd Edition: Quality, heterogeneity, collection, accessibility and treatment of data

 

Keynote speaker 1 (9:30h - 11:00h): Computational Challenges in Personalized Medicine

 

Alfonso Valencia (Barcelona Supercomputing Center)

Abstract

Personalized Medicine represents the adoption of genomics and other –OMIC technologies to the diagnosis and treatment of diseases. PerMed is arguably the more promising developments resulting from the genomic revolution.

The treatment and analysis of genomic information is tremendously challenging for a number of reasons that include the diversity and heterogeneity of the data, strong dependence of the associated metadata, very fast evolution of the experimental methods and the large size and confidential nature of the data. Furthermore, the lack of a sufficiently developed conceptual framework in molecular biology makes data interpretation even more challenging.

The use of OMIC information in medical applications will require the appropriate combination with descriptions of diseases, symptoms, drugs and treatments, as well as with information obtained from medical devices, medical images and Electronic Medical Records. Finally, the actual application of these developments in medicine will require the integrated use of machine learning methods to make all this information accessible to the adequate simulation frameworks.

At BSC, we see the interpretation of genomic and other medical information represents the key computational and methodological challenges of our time. 

Short bio

Alfonso Valencia is a Spanish biologist, the current director of the Spanish National Bioinformatics Institute and the Life Sciences department at Barcelona Supercomputing Center-Centro Nacional de Supercomputación. As of 2015, he is also President of the International Society for Computational Biology. His research is focused on the study of protein families and their interaction networks.

--------------------------------

Talk 1 (11:30-12:30): Multi-omics data acquisition and analysis in drug discovery: a Systems Pharmacology approach

Leonidas Alexopoulos (National Technical University of Athens)

Abstract

A major challenge for bringing safe and effective new treatment to patients is the deep understanding of a disease. Here, we describe multi-omics data acquisition technologies and systems biology algorithms for tackling major questions in the drug discovery and development pipeline: (i) construction of pathways and comparison between normal or diseased cells, (ii) identification of drug mode of action (MoA), and (iii) prediction of drug toxicity and efficacy. Our network-based approach is able to predict the drugs' main target and uncover off-target effects. Subsequently, machine learning algorithms can select MoAs with reduced toxicity, increased efficacy and tailor drugs to specific disease mechanisms. So far, we have applied our approach to liver cancer, osteoarthritis, multiple sclerosis, non-alcoholic fat liver disease, chronic kidney disease and more recently in melanoma. Our pathway analysis algorithms and high throughput multiplex platform pave the road for new solutions in early drug discovery.

Short bio

Dr. Leonidas Alexopoulos is Assistant Professor in the Department of Mechanical Engineering at the National Technical University of Athens and co-founder of ProtATonce (now Protavio) , a successful biotech startup company.

Dr Alexopoulos has studied at Duke (PhD 2004), MIT, and Harvard Medical School in the areas of Biomedical Engineering, Systems Biology, and Biological Engineering.  He combines strong academic research with translational capacity and entrepreneurial opportunities.

He has worked and collaborated with leading companies in the area of biotechnology, drug discovery and Systems Pharmacology including Pfizer, Boehringer Ingelheim, Vertex, and Roche.

He holds strong publication and grant record with several international awards and honors for top academic performance. Leonidas Alexopoulos is directing a multidisciplinary research laboratory composed of engineers, biologists and physicians working in the area of Systems Biology, Bioengineering, and Medical Devices. We integrate novel biological and engineering appoaches for proteomic profiling, modeling biological systems, and assay automation with applications in drug development.

Check the recording of his talk here: https://youtu.be/11R26Sx7Vfw

--------------------------------

Talk 2 (12:30-13:30). Concepts and tools from nonlinear dynamics: Helpful to understand and diagnose epilepsy?

Ralph Andrzejak (DTIC, Universitat Pompeu Fabra)

Abstract

Nonlinear time series analysis allows characterizing dynamical systems in which nonlinearity gives rise to a complex, seemingly irregular temporal evolution. Importantly, these nonlinear techniques can extract information from signals that cannot be resolved by classical linear techniques, such as spectral analysis. Nonlinear time series analysis can, for example, help to discriminate nonlinear deterministic and linear stochastic dynamics or to characterize directional interactions between dynamics.

In this seminar, I will at first illustrate these strengths of nonlinear time series analysis using some simple examples based on signals from model systems. I will then discuss applications of nonlinear time series analysis to electroencephalographic recordings from epilepsy patients. We will for example see how this analysis can contribute to the localization of the seizure generating brain area without the necessity of observing actual seizure activity – a finding of clear clinical relevance.

From this top-down data analysis, we will then turn to some bottom-up modeling of dynamical systems. In particular, we will consider networks of coupled oscillators that can show intriguingly complex states in which synchronization and de-synchronization co-exist. The link to real-world biomedical applications will again be drawn to epilepsy. I will point to an analogy between the sudden collapse of the co-existence of synchronization and de-synchronization to a fully coherent state on the one hand and the onset of epileptic seizures on the other hand. Such studies may advance our understanding of epileptic seizures from a dynamical systems point of view.

Short bio

Ralph Gregor Andrzejak was born in Germany (1970) and studied physics at the University Bonn, Germany. He wrote his Diploma (1997) and PhD thesis in physics (2001). During these theses he was member of the Neurophysics group of K. Lehnertz and C.E. Elger (Department of Physics and Department of Epileptology, University Bonn, Germany). During his PhD he spent a research stay at the Neurodynamics research group of S.J. Schiff at the George Mason University, Fairfax, USA (1999). After his PhD he carried out a first postdoc with the Complex Systems research group of P. Grassberger at the Research Centre Jülich, Germany (2002-2004). Awarded with a Feodor Lynen-fellowship from the German Alexander von Humboldt-Foundation, he carried out a second postdoc with the Computational Neuroscience group headed by G. Deco (2005-2006) at the Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona. Subsequently, he successfully carried out a tenure-track appointment at this Department granted by the Spanish Ramón y Cajal program (2007-2011). Since 2011, he is an associate professor at the Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona.

A total of 64 publications of Ralph G. Andrzejak are indexed in the ISI Web of Science. These are 45 journal articles published in leading journals of physics, neuroscience, neurology, and engineering as well as 17 conference contributions and 2 editorial articles. In the ISI-Web of Science this work receives more than 3200 citations (h-index: 26; ISI Researcher ID: H-7923-2012). In Google Scholar the publications of Dr. Andrzejak reach more than 5800 citations (h-index: 31). Dr. Andrzejak has made more than 50 conference presentations, in the majority talks. He was furthermore co-author of some 60 conference talks and posters. Altogether, his work was presented at 73 conferences in 18 different countries. In addition to his conference talks, Dr. Andrzejak gave more than 30 invited scientific seminars or tutorial lectures at international universities, research institutions, and international advanced schools.

Check the recording of his talk here: https://youtu.be/DKkv9G5kss8

Day 3: Multiphysics / Multiscale Models

Organ, cell, molecular

Special focus 3rd Edition: Integration of data for creation and parametrization of models at different scales and across the scales

 

Keynote speaker 1 (9:30h - 11:00h). Particle-based computational models and optical microscopy-based measurements of cell-matrix mechanical interactions

 

Hans van Oosterwyck (KU Leuven)

Abstract

Cells mechanically interact with their extracellular matrix (ECM) by adhering and applying force to it. These cellular forces play an important role in the dynamics of cell adhesion and migration as part of single cell or multicellular, collective behaviour and more in general can modulate cell signaling in a process called mechanotransduction. We have worked on particle-based computational models and optical microscopy-based methods of quantifying these cellular forces and studying their role during single endothelial cell migration and vascular invasion.

Computational models of single cell mechanics were created, by defining and solving force balances at subcellular scale, taking into account among others external forces from contact and friction between a cells and its ECM, as well as internal forces related to cell cortex viscoelasticity and bending rigidity. Models were extended with a description of the discrete, spatially non-uniform nature of focal adhesion turnover, protrusion and stress fiber formation, making them suitable to study the dynamics of cell adhesion and migration. Models are currently being tuned with respect to experimental data, stemming from 2D endothelial cell migration assays on polyacrylamide gels that among others enable us to quantify the distribution and dynamics of cellular tractions by means of traction force microscopy.

Computational models of cellular invasion into a viscoelastic, degradable ECM were also implemented, based on smoothed particle hydrodynamics (SPH). The cell model again accounted for cell cortex viscoelasticity and contractility, protrusion and adhesion dynamics. Computational model simulations were combined with in vitro models of 3D vascular invasion assays, compatible with live cell optical microscopy imaging. Non-rigid image registration was used to dynamically measure 3D deformation fields of ECM-mimicking collagen gels around vascular sprouts. By fitting the measured deformation fields, the computational model is able to estimate the location and magnitude of cellular tractions.

Acknowledgements: The research leading to these results has received funding from the European Research Council under the European Union's Seventh Framework Programme (FP7/2007-2013)/ ERC Grant Agreement n° 308223).

Short bio

Hans Van Oosterwyck (DOB 02.02.1972) is a professor and the chair of the Biomechanics section (Mechanical Engineering Department) at KU Leuven, where he is heading the Mechanobiology and Tissue Engineering research group (www.mech.kuleuven.be/mechanobiology). He holds an MSc degree in Materials Engineering (1995) and a PhD degree in Engineering (2000), both obtained at KU Leuven (Leuven, Belgium). He has been a postdoctoral fellow at the AO Research Institute (Davos, Switzerland) in 2004-2005 and a visiting scientist at the University of Zaragoza (Spain) in 2009. He is a member of Prometheus, the Leuven R&D Division for Skeletal Tissue Engineering. In 2012 he was awarded an ERC Starting Grant on the role of cell-matrix interaction in angiogenesis. His research focuses on the development of quantitative tools for unraveling the role of the physical microenvironment for cell fate, in particular the development of multiscale computational models and optical microscopy-based techniques for studying the importance of cell-matrix mechanics and mass transport for angiogenesis and bone regeneration. His research group is strongly interdisciplinary and combines computational modelling with experimental techniques, adopted from various fields, such as cell and tissue mechanics, cell biology, biomaterials and biophysics.

Hans Van Oosterwyck has been a Council Member of the European Society of Biomechanics (ESB) between 2006 and 2014. He has been the President of the ESB between 2012-2014.

A full list of publications can be found on https://lirias.kuleuven.be/cv?u=U0030372.

--------------------------------

Talk 1 (11:30-12:30) Modelling approaches for Systems Toxicology

Laura Furlong (IMIM - UPF)

Abstract

Drug safety is one of the major causes of failure of drug candidates during clinical trials, while drug adverse reactions (ADRs) are still a big concern in the clinical setting during the treatment of patients. Despite their clinical and economic relevance, the prediction of ADRs during early stages of the drug development process is poor. For instance, pre-clinical animal experiments fail to predict human ADRs in 30% of the cases. Our understanding of the molecular events elicited by the treatment of drugs, either the therapeutic effects or the adverse drug reactions, is still very limited. In the last years, Systems Toxicology approaches have emerged to improve our understanding of drug side effects and provide better prediction tools for drug toxicity. These approaches are based on the integration of classical toxicology with quantitative analysis of networks of molecular interactions occurring across multiple levels of biological organization. In this talk I will provide an overview of modelling approaches applied in the context of Systems Toxicology, with a special emphasis on the exploitation of omics data for the modelling of drug toxicities at the cellular and subcellular scale.

Short bio

Laura I. Furlong is a Miguel Servet Researcher at of the Hospital del Mar Medical Research Institute (IMIM), Associate Professor at the Universitat Pompeu Fabra (UPF) and Head of the Integrative Biomedical Informatics Group of the Research Programme on Biomedical Informatics, GRIB (IMIM-UPF) at Barcelona (Spain). She has a PhD in Biology from the University of Buenos Aires, Argentina and a Msc in Bioinformatics by University Pompeu Fabra. She has a broad expertise covering molecular biology, computational systems biology and text mining. Her current research is focused in developing bioinformatic approaches to unravel the mechanisms underlying human diseases and to understand why drugs produce undesired effects. Director of 4 doctoral theses. She has published over 58 peer-reviewed articles, and act as reviewer for the journals Bioinformatics, BMC Bioinformatics, BMC Systems Biology, and PLOS journals. She is Member of the Scientific Committee of the MARBiobanc, Parc de Salut Mar. She has participated in several FP7 projects (@neurist, EU-ADR) and is currently involved in the IMI (Innovative Medicines Initiative) projects transQST,  eTOXEMIFiPiE and the H2020 projects MedBioinformatics and  Elixir-Excelerate.

Check the recording of her talk here: https://youtu.be/Cb--9_6tmuI

--------------------------------

Talk 2 (12:30-13:30). A thorough analysis of the physiology of coronary arteries: the role of the basal tone and residual stress

Jean Louis Martiel (Univ. Grenoble Alpes, CNRS, CHU Grenoble Alpes, INSERM)

 

Abstract

We analyse the mechanics of coronary arteries to understand the existence of residual mechanical stress, its consequences for the physiological functioning of the artery and its role in driving the pathological evolution of the wall (formation of plaques).

Arteries are made of three layers, namely, the intima (a unicellular layer), the media and the adventitia. Here, we assume that the intima and the media do form a unique layer with similar mechanical properties; from now on we refer to this intima-media layer as the media. In physiological conditions, the active smooth muscle cells in the media develop an active stress that vasodilates or constricts the artery. Conversely, during one cardiac cycle, the artery wall is subjected to blood pressure, in the range 80 − 120 mmHg (mean at 100 mmHg), that tends to radially expand the tube. Classically, these two antagonist responses of the artery to stress are captured by the compliance, Cp, and the vasodilatation ratio, Rt, variables. These variables measure, respectively, the relative variation of the lumen area during one cardiac cycle (area at diastolic pressure minus area at systolic pressure divided by the pressure change times the area at systolic pressure) and the ratio of the lumen area without and with the basal muscle tone developed in the media. Additionally, the artery wall at rest (no activation, no pressure) is not stress free, as shown in experiments where a longitudinally cut artery tube opens spontaneously, without external loading, or, when the three layers of an open tube are separated.

In this paper, we show how a correct prediction of Cp and Rt, in combination with the wall residual stress and the basal tone in the media, is a crucial condition in modeling healthy and pathological arteries. To model arteries, we start from

• the geometrical characterization of arterial segments isolated from the heart (internal and external radii and the radius at the media-adventitia transition),

• the existence of a stress free configuration for the whole wall, characterized by an opening angle denoted α,

• the existence of an energy density function, W, for arteries (one function per layer), that yields the constitutive relation between the Cauchy stress and the deformation.

We derived the kinematics of the deformation from the stress free artery configuration, in which the there layers are separated, to the open configuration, which corresponds to the longitudinally cut artery segment characterized by a unique opening angle α. Also, we considered the deformation from the stress-free configuration to the closed tube at rest (no pressure loading, no active stress in the media). By solving the mechanical equilibrium equations, we were able to reconstruct the unknown stress-free configuration from the opening angle α and the radii of the resting configuration. In a second step, we determined the unknown smooth muscle cells tension (T0, activity in the media) and the in situ axial stretch λ∗, so that both parameters Cp and Rt have the physiological values measured in vivo.

In healthy arteries, the constrains to satisfy both Cp and Rt are compatible with a quasi-uniform circumferential Cauchy stress distribution, as already found in other studies. Moreover, our model predicts that, in this physiological state, the elastic energy stored in the wall deformation (measured with respect to the stress-free configuration) is at a minimum. By analyzing the role of the opening angle α, we also predicted that Cp and Rt clearly differ from their physiological value, hereby showing the importance of the residual stress in the artery dynamics. For example, in the absence of residual stress (α = 0), the artery response to pressure or activity change is greatly altered. Finally, we used the energy functions that were determined for pathological arteries to understand how the mechanical changes of both the media and the adventitia affect the artery performance.

Check the recording of his talk here: https://youtu.be/CT6kVDf92zI

Honorary VPH Lecture

Natalia Trayanova. Computational Cardiology

 

Wednesday June 20th 2018 - 18:00h - 19:30h

 

 

Natalia Trayanova (Johns Hopkins University). Computational Cardiology

 

 

Abstract

Sudden cardiac death (SCD) from arrhythmias is a leading cause of mortality. For patients at high SCD risk, prophylactic insertion of implantable cardioverter-defibrillators (ICDs) reduces mortality. Current approaches to identify patients at risk for arrhythmia are, however, of low sensitivity and specificity, which results in a low rate of appropriate ICD therapy. There is a critical clinical need to develop risk metrics that directly assess the interplay between abnormal myocardial structure and electrical instability in the heart, that together predispose to SCD. Here we present a novel non-invasive personalized approach to assess SCD risk in post-infarction patients based on cardiac imaging and computational modeling. This is an example of the emerging field of computational cardiology.

In computational cardiology, we construct personalized 3D computer models of post-infarction hearts from patients’ clinical magnetic resonance imaging data. Each heart model incorporates not only myocardial structure, but electrophysiological functions from the sub-cellular to the organ, allowing for representation of electrical instability. Thus the interplay between abnormal myocardial structure and electrical instability in the heart that predisposes to SCD can be directly assessed. In each heart model, we conduct a virtual multi-site delivery of electrical stimuli from ventricular locations at different distances to remodeled tissue so that the patient’s heart propensity to develop infarct-related ventricular arrhythmias can be comprehensively evaluated.  Simulations are conducted for each virtual heart, probing its propensity to develop infarct-related ventricular arrhythmia. We term this non-invasive SCD risk assessment approach VARP, virtual-heart arrhythmia risk predictor.

In a proof-of-concept retrospective study, we assessed the predictive capability of the VARP approach as compared to that of other clinical metrics in a cohort of 41 patients. Statistical analysis demonstrated that a positive VARP test was significantly associated with the primary endpoint, with a four-fold higher arrhythmia risk than patients with negative VARP test. Our results also demonstrate that VARP significantly outperformed clinical metrics in predicting future arrhythmic events.

The robust and non-invasive VARP approach has the potential to prevent SCD and eliminate unnecessary ICD implantations in post-infarction patients.  Importantly, the methodology could be applied to patients with prior MI but preserved ejection fraction who could also be at significant risk for arrhythmia because of their remodeled myocardium, but are generally not targeted for therapy under current clinical recommendations. Finally, the same approach is easily extendable to other heart diseases.

Short bio

Dr. Natalia Trayanova is the Murray B. Sachs Professor in the Department of Biomedical Engineering and the Institute for Computational Medicine, and directs the Computational Cardiology Laboratory at Johns Hopkins University. She is also a Professor in the Department of Medicine. In 2013, she received the NIH Director’s Pioneer Award for her project “Virtual Electrophysiology Laboratory”. Dr. Trayanova was also the inaugural William R. Brody Faculty Scholar at Johns Hopkins University. She is a Fellow of the Heart Rhythm Society, American Heart Association, Biomedical Engineering Society, and the American Institute for Medical and Biological Engineering.

 

Day 4: Computational Aspects. Implementation, validation, coupling.

Special focus 3rd Edition: Sensitivity analyses, Stochastic simulations, model verification, direct and indirect validation

 

Keynote speaker 1 (9:30h - 11:00h). Numerical methods for the simulation of the heart function

 

Alfio Quarteroni (Politecnico di Milano / EPFL, Lausanne - honorary professor)

Abstract

Mathematical models based on first principles allow the description of the blood motion in the human circulatory system, as well as the interaction between electrical, mechanical and fluid-dynamical processes occurring in the heart. This is a classical environment where multi-physics processes have to be addressed.

Appropriate numerical strategies can be devised to allow for an effective description of the fluid in large and medium size arteries, the analysis of physiological and pathological conditions, and the simulation, control and shape optimization of assisted devices or surgical prostheses.

This presentation will address some of these issues and a few representative applications of clinical interest.

Short bio

Alfio Quarteroni is Honorary Professor of Mathematics at the EPFL, Lausanne  and Professor of Numerical Analysis at the Politecnico di Milano since 1989.

He is member of the Italian Academy of Science, the European Academy of Science, and the Academia Europaea.

He got the NASA Group Achievement Award for the pioneering work in Computational Fluid Dynamics in 1992, the International Galileo Galilei prize for Sciences 2015, thedoctorate Honoris Causa in Naval Engineering from University of Trieste, 2003; he is the Recipient of the Galileian Chair from the Scuola Normale Superiore, Pisa, Italy ,2001,

He is author of 22 books, editor of 5 books, author of more than 300 research papers.

His research interests concern Mathematical Modeling, Numerical Analysis, Scientific Computing, and Application to: fluid mechanics, geophysics, medicine and the improvement of sports performance. His Group has carried out the mathematical simulation for the optimisation of performances of the Alinghi yacht, the winner of two editions (2003 and 2007) of the America’s Cup.

--------------------------------

Talk 1 (11:30-12:30). Understanding heart function through combined computational, experimental and clinical research

  

Esther Pueyo (University of Zaragoza / Biomedical Research Networking Centre on Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN).

Abstract

Cardiovascular diseases (CVDs) are responsible for more than 4 million deaths per year in Europe. The single major risk factor for CVD is age, which leads to a progressive decline in the physiological functions of the body, with very notable effects on the heart. These effects are associated with enhanced predisposition to cardiac arrhythmias. Also, CVDs like heart failure, with high prevalence in the elderly population, are linked to increased risk for  arrhythmia development, in some cases leading to sudden cardiac death.
 
In this presentation, integrative approaches combining in silico modeling with in vitro cell and tissue analysis and in vivo electrocardiographic evaluation will be presented. Application of such approaches to investigate how aging and cardiovascular diseases manifest at a range of scales, covering from ion channels in the cell membrane to whole-body surface potentials, will be shown. The role of electrical, structural and autonomic alterations in contributing to such manifestations will be explored. Advances towards the proposal of non-invasive markers able to identify aged and diseased individuals at high arrhythmic risk will be presented.
 

Short bio

Esther Pueyo is Associate Professor at the University of Zaragoza and Senior Research Scientist at the Biomedical Research Networking Centre on Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN). She has established research experience in biomedical signal processing, the field where she pursued her PhD at the University of Zaragoza, and in electrophysiological modeling and simulation, after having completed a three-year postdoctoral period at the University of Oxford (UK). Esther Pueyo has additionally worked in the areas of cardiology and experimental physiology during research visits and in collaborations with different institutions in Spain, UK, Sweden, Hungary and the United States. Her investigations involve active collaborations with a wide network of research groups including the University of Oxford, University College London, University of Szeged, University of Debrecen and FDA, among others.

Esther Pueyo has participated as Principal Investigator (PI) or Collaborator in a large number of international, European, national and regional research projects. She is currently PI of the ERC Starting Grant project MODELAGE, funded with 1.5 M€, and is / has been PI of 7 other grants funded by regional, national and international agencies. She has and is actively participating in 3 projects led by the European Space Agency, 2 Marie Curie Innovative Training Networks under H2020 EU funding and was invited to join the Oxford team to contribute to the EU-funded FP7 project PREDiCT. She has also participated in more than 28 other projects, with broad international collaborations.

Esther Pueyo has authored/co-authored over 100 peer-reviewed publications, including 52 full articles in high-impact journals, 12 journal abstracts and 38 conference papers. More than 75% of her publications are in the top third of JCR ranking. She regularly delivers invited talks in scientific conferences and symposiums, is a reviewer for twenty different JCR journals, Editorial Board member of Physiological Measurement and Plos One as well as Associate Editor of Frontiers in Physiology. She has supervised / is supervising 6 postdoctoral researchers and 9 PhD students, three of the latter jointly with FDA (USA), University of Oxford (UK) and Politecnico di Milano (Italy). She has additionally supervised 4 Master Theses, 7 Graduate Theses and 3 Industrial Fellowships.

--------------------------------

Talk 2 (12:30-13:30): Systems biology & experiments for tissue engineering

 

Liesbet Geris (University of Liége)

Abstract

Computer modeling and simulation can be used in the context of tissue engineering to help quantify and optimize the products and processes and can help in predicting or interpreting the biological response after implantation. The Biomechanics Research Unit has developed a suite of models that capture various aspects of the tissue engineering life cycle.  From gene regulatory network models describing culture strategies over data-driven models for biomaterial optimization and mechanistic models for bioreactor control, to in silico clinical trials of treatment strategies for pediatric orphan indications. Every model that has been developed goes through its own model development: model establishment and selection, parameter optimization, sensitivity analysis, model adaptation. Intertwined with this process is the establishment of model credibility with the verification, validation and uncertainty quantification. The establishment of guidelines and standards is a crucial step in the transition of models from bench to bedside.

In this talk I will address, by means of various examples from the bone tissue engineering field, the different model development steps as well as different approaches to establishing model credibility. Finally, I will give an overview of the ongoing efforts within the in silico medicine community, including the Virtual Physiological Human Institute, the Avicenna Alliance and the FDA, to establish the aforementioned guidelines and standards.

Short bio

Liesbet Geris (DOB 04.06.1979) is Professor in Biomechanics and Computational Tissue Engineering at the Department of Aerospace and Mechanical Engineering at the university of Liège and Associate Professor at the Department of Mechanical Engineering of the KU Leuven, Belgium. From the KU Leuven, she received her MSc degree in Mechanical Engineering in 2002 and her PhD degree in Engineering in 2007, both summa cum laude. In 2007 she worked as a postdoctoral researcher at the Centre of Mathematical Biology of Oxford University.

Her research interests encompass the mathematical modeling of bone regeneration during fracture healing, implant osseointegration and tissue engineering applications. The phenomena described in these mathematical models reach from the tissue level, over the cell level, down to the molecular level. She works in close collaboration with experimental and clinical researchers from the university hospitals Leuven, focusing on the development of mathematical models of impaired healing situations and the in silico design of novel treatment strategies. She is scientific coordinator of Prometheus, the skeletal tissue engineering division of the KU Leuven. Her research is financed by European, regional and university funding. In 2011 she was awarded an ERC starting grant to pursue her research and in 2017 she was awarded an ERC consolidator grant.

Day 5: Applications. Understanding therapy decision support

Special focus 3rd Edition: Interpretable machine learning, metamodeling, success and failure stories

 

Keynote speaker 1 (9:30h - 11:00h). Digging into the information of the diversity of the human genome: from the reconstruction of human origins to the adaptation to the environment

 

Jaume Bertranpetit (Universitat Pompeu Fabra)

Short bio

Professor of Biology at the Pompeu Fabra University (Barcelona). Group leader in the Evolutionary Biology and Complex Systems Program in this University. Promoter of the Institute for Evolutionary Biology, IBE (UPF-CSIC). His research field is in different aspects on the study of the human genome variation and diversity: human population genetics, molecular evolution, comparative genomics and the interaction between human evolutionary biology and other fields, including medicine, genetic of complex diseases, statistical genetics and others. Recent publications are mainly on the footprint of natural selection in the human genome and the emerging field of Evolutionary Systems Biology, with the relationship of molecular networks and adaptation in genome-wide perspective. He has published over 350 research papers, most of them since his major dedication to genome studies (since 1992). Director of ICREA (Institució Catalana de Recerca i Estudis Avançats) till 2015.

--------------------------------

Talk 1 (11:30-12:30): Bioinformatics for gene panelling

 

Sergio Lois (Sistemas Genómicos)

Abstract

The development of high-throughput technologies such as next-generation sequencing (NGS) has allowed for thousands of DNA loci to be interrogated simultaneously in a fast and economical method for the detection of clinically deleterious variants. Whenever a clinical diagnosis is known, a targeted NGS approach involving the use of disease-specific gene panels can be employed. This approach is often valuable as it allows for a more specific and clinically relevant interpretation of results. Here, we describe the role of bioinformatics in the customization, optimization and utilization of gene panels for the scalability of clinical diagnostic cardiovascular genetic tests. Success story: Incorporation of custom copy number analysis pipeline results in an increase in the diagnostic yield of cardiovascular diseases.

Short bio

(Sergio Lois) was born in Barcelona (1980) and studied Human Biology at University Pompeu Fabra, Spain. He holds a PhD degree onBiochemistry and Molecular Biology by University of Barcelona, where he entered the field of bioinformatics applied to epigenetics. He joined Predictive and Personalized Cancer Insititute (IMPPC) in 2008, where he gained experience in next-generation sequencing analysis applied to DNA methylation. In 2012, he joined Translational Bioinformatics group at Vall d’Hebron Hospital (VHIR), where he developed bioinformatic pipelines applied to the identification of genomic and epigenomic alterations in cancer. Currently, he is working at the Bioinformatics Unit of Sistemas Genomicos, where he is focused on the development of bioinformatic tools aimed at understanding the molecular basis of disease.

--------------------------------

Talk 2 (12:30-13:30). Combining multi-scale computer models and machine learning for planning radio-frequency ablation interventions

Rafael Sebastián (Universitat de València)

Abstract

Cardiac arrhythmias, such as focal atrial tachycardia or reentrant tachycardias are commonly treated by radio frequency ablation with an acceptable long-term success. The catheter ablation treatment targets the arrhythmogenic electrical drivers and terminates them by localized energy delivery. The end point of catheter based ablation is to eliminate the triggers with the least amount of ablation necessary. Although the location of ectopic foci tends to appear in specific hot-spots, and reentrant ventricular tachycardias are usually triggered from peri-infarct areas, they can be virtually located in any atrial region.

We aim to develop indices based on non-invasive data that have the potential to predict the location of arrhythmic drivers in the atria and ventricles.

To achieve that goal, we make use of detailed 3D heart models that allow simulating different scenarios, with different degrees of pathological remodelling to learn how they affect non-invasive indices. Those models combined with machine learning techniques have the potential to assist electrophysiologist in therapy planning of ablation interventions.  As a results, we can improve the success of ablation interventions and shorten their durations.

Short bio

Rafael Sebastian is assistant professor at the department of computer science of Universitat de Valencia. He holds a PhD on Computer Sciences by University of Valencia. As postdoctoral researcher, he joined the Center for Computational Imaging & Simulation Technologies in Biomedicine at the Universitat Pompeu Fabra, supported by the Spanish government with a Juan de la Cierva fellowship. He was responsible for the Computer Cardiac Modelling  and Simulation  Section  leading  several  work  packages  in  national  and European projects such as the euHeart project. He returned to the Universitat de Valencia to start the Computational Modelling Simulation Lab (CoMMLab), which focuses on modelling of cellular, tissue and organ  structures  of  the  heart. Currently, he is interested in using machine learning techniques combined with biophysical simulation to study cardiac arrhythmia, and aid in therapy planning and optimisation. He has  published  30  papers  in  peer-review  journals  and more  than  50 conferences papers. 

--------------------------------

Presentation of awards (16:30 - 17:00)

Best Poster Award 

Best Hands-on Award by Chair UPF-QUAES

--------------------------------

Hands-on sessions (each participant selects one)

The participants will work in groups on implementing the whole EP simulation pipeline starting from image segmentation, fiber extraction, mesh generation and finishing with the electro-physiology simulations. The participants are expected to learn the basic steps and requirements for setting up the simulations and get a hands on experience with running simulations on the HPC. The simulations will be carried out using Alya multiphysics simulator developed by BSC.

Chronic modification of the mechanical fields on the articular cartilage can trigger/accelerate catabolic responses in chondrocytes. In this hands on motion analyses will be done in the UPF MoCA LAB, and knee joint forces will be calculated from inverse dynamic analyses (The gait of two volunteers selected among the participants will be analysed - Volunteers will need to wear stretch sport clothes to allow proper movement analysis). These forces will be further translated as boundary conditions on a cartilage tissue finite element model. Multiphysics simulations performed with FEBio will give the possibility to introduce the effect of preexisting tissue damage, e.g. osteoarthritis. The calculated mechanical fields will be further used to feed an agent-based model of cartilaginous cells, developed in NetLOGO, so as to explore the catabolic response of these cells in terms of inflammation and protein expressions. For the smooth implementation of the workshop, students will be divided into three focus groups.

This computational fluid dynamics (CFD) hands on focusses on the simulation of 3D flow conditions by using the Finite Volume Method. The students will be guided through the modelling of most aspects of fluid flow, i.e. non-Newtonian, multiphase, porous media or particles (Lagrange) flows in laminar or turbulent flow regimes. All modelling approaches will include the following steps: (i) selecting the appropriate description of the flow (single-phase or two-phase, laminar or turbulent flows); (ii) creating or importing the 3D model geometry; (iii) defining the fluid properties; (iv) performing steady or transient simulations. The hand will exploit left atrium models to focus on a specific application case.

--------------------------------

 

VPHi Best Poster Award

All VPHi student members bringing a poster to the VPH Summer will be considered for the VPHi Best Poster Award that consists in a 1.000 € grant for the participation to scientific conferences or workshop from September 2018 to December 2019. Posters brought to the VPH Summer School will be exposed in the coffee area of the hands-on session building from Tuesday 19th to Thursday 21st (installation on Monday, removal on Friday) and will be evaluated by an independent scientific committee that will include various of the VPH Summer School speakers. If you are not a VPHi Student Member, but want to be eligible for the VPHi Best Poster Award but, please apply here (add link to the application form).