0000000000954641
AUTHOR
Miguel Lozano
Automatic estimation of Purkinje-Myocardial junction hot-spots from noisy endocardial samples: A simulation study
The reconstruction of the ventricular cardiac conduction system (CCS) from patient-specific data is a challenging problem. High-resolution imaging techniques have allowed only the segmentation of proximal sections of the CCS from images acquired ex vivo. In this paper, we present an algorithm to estimate the location of a set of Purkinje-myocardial junctions (PMJs) from electro-anatomical maps, as those acquired during radio-frequency ablation procedures. The method requires a mesh representing the myocardium with local activation time measurements on a subset of nodes. We calculate the backwards propagation of the electrical signal from the measurement points to all the points in the mesh …
Towards reactive navigation and attention skills for 3D intelligent characters
This paper presents a neural design which is able to provide the necessary reactive navigation and attention skills for 3D embodied agents (virtual humanoids or characters). Based on Grossberg's neural model of conditioning [6], as recently implemented by Chang and Gaudiando [7], and according to the Adaptative Resonance Theory (ART) and the neuroscientific concepts associated, the neural design introduced has been divided in two main phases. Firstly, an environmentcategorization phase, where an on-line pattern recognition and categorization of the current agent sensory input data is carried out by a self organizing neural network, which will finally provide the agent's short term memory la…
Simulating socially intelligent agents in semantic virtual environments
AbstractThe simulation of synthetic humans inhabiting virtual environments is a current research topic with a great number of behavioral problems to be tackled. Semantical virtual environments (SVEs) have recently been proposed not only to ease world modeling but also to enhance the agent–object and agent–agent interaction. Thus, we propose the use of ontologies to define the world’s knowledge base and to introduce semantic levels of detail that help the sensorization of complex scenes—containing lots of interactive objects. The object taxonomy also helps to create general and reusable operativity for autonomous characters—for example, liquids can be poured from containers such as bottles. …
MARL-Ped: A multi-agent reinforcement learning based framework to simulate pedestrian groups
Abstract Pedestrian simulation is complex because there are different levels of behavior modeling. At the lowest level, local interactions between agents occur; at the middle level, strategic and tactical behaviors appear like overtakings or route choices; and at the highest level path-planning is necessary. The agent-based pedestrian simulators either focus on a specific level (mainly in the lower one) or define strategies like the layered architectures to independently manage the different behavioral levels. In our Multi-Agent Reinforcement-Learning-based Pedestrian simulation framework (MARL-Ped) the situation is addressed as a whole. Each embodied agent uses a model-free Reinforcement L…
A Distributed Framework for Scalable Large-Scale Crowd Simulation
Emerging applications in the area of Emergency Response and Disaster Management are increasingly demanding interactive capabilities to allow for the quick understanding of a critical situation, in particular in urban environments. A key component of these interactive simulations is how to recreate the behavior of a crowd in real- time while supporting individual behaviors. Crowds can often be unpredictable and present mixed behaviors such as panic or aggression, that can very rapidly change based on unexpected new elements introduced into the environment. We present preliminary research specifically oriented towards the simulation of large crowds for emergency response and rescue planning s…
Automatic Location of Sources of Electrical Activation from Electroanatomical Maps
Electro-anatomical mapping is a widely used technique used by electrophysiologists to understand patient's activation pattern. The system measures activation time at different locations but does not provide information on underlying electrical pathways or triggering points, such as Purkinje-myocardial junctions or ectopic foci. We present a method to estimate the locations of Purkinje-myocardial junctions from a discrete set of endocardial samples. Using less than 1000 endocardial samples it can recover locations and activation times of the most influencing Purkinje myocardial junctions from Purkinje trees of up to 500 junctions. A simulation study revealed that using the estimated Purkinje…
Clinically-Driven Virtual Patient Cohorts Generation: An Application to Aorta
The combination of machine learning methods together with computational modeling and simulation of the cardiovascular system brings the possibility of obtaining very valuable information about new therapies or clinical devices through in-silico experiments. However, the application of machine learning methods demands access to large cohorts of patients. As an alternative to medical data acquisition and processing, which often requires some degree of manual intervention, the generation of virtual cohorts made of synthetic patients can be automated. However, the generation of a synthetic sample can still be computationally demanding to guarantee that it is clinically meaningful and that it re…
A GPU-Based Multi-agent System for Real-Time Simulations
The huge number of cores existing in current Graphics Processor Units (GPUs) provides these devices with computing capabilities that can be exploited by distributed applications. In particular, these capabilites have been used in crowd simulations for enhancing the crowd rendering, and even for simulating continuum crowds. However, GPUs have not been used for simulating large crowds of complex agents, since these simulations require distributed architectures that can support huge amounts of agents. In this paper, we propose a GPU-based multi-agent system for crowd simulation. Concretely, we propose the use of an on-board GPU to implement one of the main tasks that a distributed server for c…
Performance improvements of real-time crowd simulations
The current challenge for crowd simulations is the design and development of a scalable system that is capable of simulating the individual behavior of millions of complex agents populating large scale virtual worlds with a good frame rate. In order to overcome this challenge, this thesis proposes different improvements for crowd simulations. Concretely, we propose a distributed software architecture that can take advantage of the existing distributed and multi-core architectures. In turn, the use of these distributed architectures requires partitioning strategies and workload balancing techniques for distributed crowd simulations. Also, these architectures allow the use of GPUs not only fo…
Tuning Java to Run Interactive Multiagent Simulations over Jason
Java-based simulation environments are currently used by many multiagent systems (MAS), since they mainly provide portability as well as an interesting reduction of the development cost. However, this kind of MAS are rarely considered when developing interactive applications with time response constraints. This paper analyses the performance provided by Jason, a well-known Java-based MAS platform, as a suitable framework for developing interactive multiagent simulations. We show how to tune both the heap size and the garbage collection of the Java Virtual Machine in order to achieve a good performance while executing a simple locomotion benchmark based on crowd simulations. Furthermore, the…
Coordination and Sociability for Intelligent Virtual Agents
This paper presents a multi-agent framework designed to simulate synthetic humans that properly balance task oriented and social behaviors. The work presented in this paper focuses on the social library integrated in BDI agents to provide socially acceptable decisions. We propose the use of ontologies to define the social relations within an artificial society and the use of a market based mechanism to reach sociability by means of task exchanges. The social model balances rationality, to control the global coordination of the group, and sociability, to simulate relations (e.g. friendliness) and reciprocity among agents. The multi-agent framework has been tested successfully in dynamic envi…
Sociable Behaviors in Virtual Worlds
When simulating three-dimensional environments populated by virtual humanoids, immersion requires the simulation of consistent social behaviors to keep the attention of the user/s while displaying realistic scenes. However, intelligent virtual actors still lack a kind of collective or social intelligence necessary to reinforce the roles they are playing in the simulated environment (e.g. a waiter, a guide, etc). Decision making for virtual agents has been traditionally modeled under self interested assumptions, which are not suitable for social multi-agent domains. Instead, artificial society models should be introduced to provide virtual actors with socially acceptable decisions, which are…
A J-MADeM Agent-Based Social Simulation to Model Urban Mobility
The mobility models followed within metropolitan areas, mainly based on the massive use of the car instead of the public transportation, will soon become unsustainable unless there is a change of citizens’ minds and transport policies. The main challenge related to urban mobility is that of getting free-flowing greener cities, which are provided with a smarter and accessible urban transport system. In this paper, we present an agent-based social simulation approach to tackle this kind of social-ecological systems. The Jason Multi-modal Agent Decision Making (JMADeM) library enable us to model and implement the social decisions made by each habitant about how to get to work every day, e.g., …
J-MADeM, a market-based model for complex decision problems
This paper presents J-MADeM, a multi-modal decision making mechanism to provide agents in a Multi-Agent Systems (MAS) with a market-based model for complex decision problems. J-MADeM is now available as an open source library fully integrated into Jason, the successful interpreter for the AgentSpeak programming language. The aim of this work is to improve Jason by incorporating an agent decision-making module able to merge multiple information sources received from the rest of the agents. This information is modeled as a set of utility functions expressing the preferences of the agents for a specific problem. Then, J-MADeM agents use one-round sealed-bid combinatorial auctions as the main p…
A scalable multiagent system architecture for interactive applications
Interactive applications like crowd simulations need to properly render the virtual world while simulating the interaction of thousands of agents at the same time. The computational workload generated by these two tasks highly increases with the number of the simulated agents, requiring a scalable design of the multiagent system. In this paper, we present, in an unified manner, a distributed multiagent system architecture that can manage large crowds of autonomous agents at interactive rates while rendering multiple views of the virtual world being simulated. This architecture consists of a distributed multiagent system and a complementary distributed visualization subsystem. We also presen…
A distributed visualization system for crowd simulations1
The visualization system of large-scale crowd simulations should scale up with both the number of visuals views of the virtual world and the number of agents displayed in each visual. Otherwise, we could have large scale crowd simulations where only a small percentage of the population is displayed. Several approaches have been proposed in order to efficiently render crowds of animated characters. However, these approaches either render crowds animated with simple behaviors or they can only support a few hundreds of user-driven entities. In this paper, we propose a distributed visualization system for large crowds of autonomous agents that allows the visualization of crowds animated with co…
Calibrating a Motion Model Based on Reinforcement Learning for Pedestrian Simulation
In this paper, the calibration of a framework based in Multi-agent Reinforcement Learning (RL) for generating motion simulations of pedestrian groups is presented. The framework sets a group of autonomous embodied agents that learn to control individually its instant velocity vector in scenarios with collisions and friction forces. The result of the process is a different learned motion controller for each agent. The calibration of both, the physical properties involved in the motion of our embodied agents and the corresponding dynamics, is an important issue for a realistic simulation. The physics engine used has been calibrated with values taken from real pedestrian dynamics. Two experime…
An Automata-Based Cardiac Electrophysiology Simulator to Assess Arrhythmia Inducibility
Personalized cardiac electrophysiology simulations have demonstrated great potential to study cardiac arrhythmias and help in therapy planning of radio-frequency ablation. Its application to analyze vulnerability to ventricular tachycardia and sudden cardiac death in infarcted patients has been recently explored. However, the detailed multi-scale biophysical simulations used in these studies are very demanding in terms of memory and computational resources, which prevents their clinical translation. In this work, we present a fast phenomenological system based on cellular automata (CA) to simulate personalized cardiac electrophysiology. The system is trained on biophysical simulations to re…
A Multiagent Framework to Animate Socially Intelligent Agents
This paper presents a multiagent framework designed to animate groups of synthetic humans that properly balance task oriented and social behaviors. The work presented in this paper focuses on the BDI agents and the social model integrated to provide socially acceptable decisions. The social model provides rationality, to control the global coordination of the group, and sociability, to simulate relations (e.g. friends) and reciprocity between members. The multiagent based framework has been tested successfully in dynamic environments while simulating a virtual university bar, where several types of agents (groups of waiters and customers) can interact and finally display complex social beha…
Modeling, evaluation, and scale on artificial pedestrians: a literature review
Modeling pedestrian dynamics and their implementation in a computer are challenging and important issues in the knowledge areas of transportation and computer simulation. The aim of this article is to provide a bibliographic outlook so that the reader may have quick access to the most relevant works related to this problem. We have used three main axes to organize the article's contents: pedestrian models, validation techniques, and multiscale approaches. The backbone of this work is the classification of existing pedestrian models; we have organized the works in the literature under five categories, according to the techniques used for implementing the operational level in each pedestrian …
Design of an ICT Tool for Decision Making in Social and Health Policies
The governance requires technical support regarding the complexity in deciding health policies to assist people who require long-term care. Long-term care policies require the use of ICT simulation tools that can provide policy makers with the option of going into a decision theatre and virtually knowing the consequences of different policies prior to finally determining the real policy to be adopted. In this sense, there is an absence of simulation tools for decision making about long-term care policies. In this chapter, the authors propose the foundations and guidelines of SSIMSOWELL, a new scalable, multiagent simulation tool that increases the prediction capacity of governance in the lo…
An efficient synthetic vision system for 3D multi-character systems
This poster deals with the problem of sensing virtual environments for 3D intelligent multi-character simulations. As these creatures should display reactive skills (navigation or gazing), together with the necessary planning processes, required to animate their behaviours, we present an efficient and fully scalable sensor system designed to provide this information (low level + high level) to different kinds of 3D embodied agents (games, storytelling, etc).
Training machine learning models with synthetic data improves the prediction of ventricular origin in outflow tract ventricular arrhythmias
In order to determine the site of origin (SOO) in outflow tract ventricular arrhythmias (OTVAs) before an ablation procedure, several algorithms based on manual identification of electrocardiogram (ECG) features, have been developed. However, the reported accuracy decreases when tested with different datasets. Machine learning algorithms can automatize the process and improve generalization, but their performance is hampered by the lack of large enough OTVA databases. We propose the use of detailed electrophysiological simulations of OTVAs to train a machine learning classification model to predict the ventricular origin of the SOO of ectopic beats. We generated a synthetic database of 12-l…
Accelerating collision detection for large-scale crowd simulation on multi-core and many-core architectures
The computing capabilities of current multi-core and many-core architectures have been used in crowd simulations for both enhancing crowd rendering and simulating continuum crowds. However, improving the scalability of crowd simulation systems by exploiting the inherent parallelism of these architectures is still an open issue. In this paper, we propose different parallelization strategies for the collision check procedure that takes place in agent-based simulations. These strategies are designed for exploiting the parallelism in both multi-core and many-core architectures like graphic processing units (GPUs). As for the many-core implementations, we analyse the bottlenecks of a previous G…
Non-invasive localization of atrial ectopic beats by using simulated body surface P-wave integral maps
Non-invasive localization of continuous atrial ectopic beats remains a cornerstone for the treatment of atrial arrhythmias. The lack of accurate tools to guide electrophysiologists leads to an increase in the recurrence rate of ablation procedures. Existing approaches are based on the analysis of the P-waves main characteristics and the forward body surface potential maps (BSPMs) or on the inverse estimation of the electric activity of the heart from those BSPMs. These methods have not provided an efficient and systematic tool to localize ectopic triggers. In this work, we propose the use of machine learning techniques to spatially cluster and classify ectopic atrial foci into clearly diffe…
Emergent Collective Behaviors in a Multi-agent Reinforcement Learning Pedestrian Simulation: A Case Study
In this work, a Multi-agent Reinforcement Learning framework is used to generate simulations of virtual pedestrians groups. The aim is to study the influence of two different learning approaches in the quality of generated simulations. The case of study consists on the simulation of the crossing of two groups of embodied virtual agents inside a narrow corridor. This scenario is a classic experiment inside the pedestrian modeling area, because a collective behavior, specifically the lanes formation, emerges with real pedestrians. The paper studies the influence of different learning algorithms, function approximation approaches, and knowledge transfer mechanisms on performance of learned ped…
A new system architecture for crowd simulation
Crowd simulation requires both rendering visually plausible images and managing the behavior of autonomous agents. Therefore, these applications need an efficient design that allows them to simultaneously handle these two requirements. Although several proposals have focused on the software architectures for these systems, no proposals have focused on the computer systems supporting them. In this paper, we analyze the computer architectures used in the literature to support distributed virtual environments. Also, we propose a distributed computer architecture which is efficient enough to support simulations of thousand of autonomous agents. This proposal consists of a cluster of interconnec…
Integrating miniMin-HSP agents in a dynamic simulation framework
In this paper, we describe the framework created for implementing AI-based animations for artificial actors in the context of IVE (Intelligent Virtual Environments). The minMin-HSP (Heuristic Search Planner) planner presented in [12] has been updated to deal with 3D dynamic simulation environments, using the sensory/actuator system fully implemented in UnrealTM and presented in [10]. Here, we show how we have integrated these systems to handle the necessary balance between the reactive and deliberative skills for 3D Intelligent Virtual Agents (3DIVAs). We have carried out experiments in a multi-agent 3D blocks world, where 3DIVAs will have to interleave sensing, planning and execution to be…
Integrating Social Skills in Task-Oriented 3D IVA
This paper presents a set of mechanisms oriented to incorporate social information into the decision taking of task-oriented 3DIVA. The aim of this approach is to integrate collaborative skills in different character's roles (seller/buyer, worker, pedestrian, etc.) in order to enhance its behavioral animation. The collective intelligence expected in this kind of multi-character domains (e.g. storytelling, urban simulation, interactive games, etc.) requires agents able to dialogue/interact with other characters, to autonomously group/ungroup (according to their goals), or to distribute tasks and coordinate their execution for solving possible conflicts. The social model implemented follows t…
Estimation of Location and Activation Time of Purkinje Myocardial Junctions from Sparse and Noisy Endocardial Electrical Samples
The activation of the myocardial muscle is triggered by Purkinje-myocardial junctions (PMJs), which are the terminal sites of the specialised cardiac conduction system (CCS). Obtaining the location of the PMJs and other sources of endocardial ectopic activity would be desirable for building computer models of cardiac electrophysiology and planning ablation interventions. We present a method to estimate the location and activation times of endocardial electrical sources in a 3D model of the ventricles. The algorithm requires a set of discrete electrical samples on the endocardium, which can include errors in location and activation time. The estimated sources are properly placed with a locat…
A comparative study of partitioning methods for crowd simulations
The simulation of large crowds of autonomous agents with realistic behavior is still a challenge for several computer research communities. In order to handle large crowds, some scalable architectures have been proposed. Nevertheless, the effective use of distributed systems requires the use of partitioning methods that can properly distribute the workload generated by agents among the existing distributed resources. In this paper, we analyze the use of irregular shape regions (convex hulls) for solving the partitioning problem. We have compared a partitioning method based on convex hulls with two techniques that use rectangular regions. The performance evaluation results show that the conv…
Estimation of Personalized Minimal Purkinje Systems From Human Electro-Anatomical Maps
The Purkinje system is a heart structure responsible for transmitting electrical impulses through the ventricles in a fast and coordinated way to trigger mechanical contraction. Estimating a patient-specific compatible Purkinje Network from an electro-anatomical map is a challenging task, that could help to improve models for electrophysiology simulations or provide aid in therapy planning, such as radiofrequency ablation. In this study, we present a methodology to inversely estimate a Purkinje network from a patient's electro-anatomical map. First, we carry out a simulation study to assess the accuracy of the method for different synthetic Purkinje network morphologies and myocardial junct…
A Saturation Avoidance Technique for Peer-to-Peer Distributed Virtual Environments
This paper presents a multi-agent framework oriented to animate groups of synthetic humans that properly balance task-oriented and social behaviors. We mainly focus on the social model designed for BDI-agents to display socially acceptable decisions. This model is based on an auction mechanism used to coordinate the group activities derived from the character's roles. The model also introduces reciprocity relations between the members of a group and allows the agents to include social tasks to produce realistic behavioral animations. Furthermore, a conversational library provides the set of plans to manage social interactions and to animate from simple chats to more complex negotiations. Th…
Inverse estimation of terminal connections in the cardiac conduction system
Modeling the cardiac conduction system is a challenging problem in the context of computational cardiac electrophysiology. Its ventricular section, the Purkinje system, is responsible for triggering tissue electrical activation at discrete terminal locations, which subsequently spreads throughout the ventricles. In this paper, we present an algorithm that is capable of estimating the location of the Purkinje system triggering points from a set of random measurements on tissue. We present the properties and the performance of the algorithm under controlled synthetic scenarios. Results show that the method is capable of locating most of the triggering points in scenarios with a fair ratio bet…
Atrial Fibrosis Hampers Non-invasive Localization of Atrial Ectopic Foci From Multi-Electrode Signals: A 3D Simulation Study
[EN] Introduction: Focal atrial tachycardia is commonly treated by radio frequency ablation with an acceptable long-term success. Although the location of ectopic foci tends to appear in specific hot-spots, they can be located virtually in any atrial region. Multi-electrode surface ECG systems allow acquiring dense body surface potential maps (BSPM) for non-invasive therapy planning of cardiac arrhythmia. However, the activation of the atria could be affected by fibrosis and therefore biomarkers based on BSPM need to take these effects into account. We aim to analyze the effect of fibrosis on a BSPM derived index, and its potential application to predict the location of ectopic foci in the …
Towards a Simulator of Integrated Long-Term Care Systems for Elderly People
In this paper, we propose a simulator for integrated long-term care systems using as a starting point a holistic model of care systems for people that need long term care, the Sustainable Socio-Health Model (SSHM). The implementation of the simulator on the Jason multi-agent platform allows the tool to include the human interactions, preferences, and social abilities that take place between elderly people and the staff of healthcare systems (doctors, social workers and nurses). In addition, the use of this multi-agent platform provides the required scalability for simulating population sizes of different orders of magnitude. The paper shows the model to be implemented in the simulator, the…
Combining Biophysical Modeling and Machine Learning to Predict Location of Atrial Ectopic Triggers
The search for focal ectopic activity in the atria triggered from non-standard regions can be time consuming. The use of body surface potential maps to plan the intervention can be helpful, but require an advance processing of the data, that usually involves to solve an ill-posed inverse problem. In addition, changes in maps due to pathological substrate such as fibrosis might affect the expected electrical patterns. In this work, we use a machine learning approach to relate ectopic focus activity in different atrial regions with body surface potential maps, and consider the effects of fibrosis in various densities and distributions. Results show that as fibrosis increases over 15% the syst…
Social Animation in Complex Environments
This work presents a market-based social model to produce good quality behavioral animations for groups of intelligent virtual agents. The social model coordinates the activities of groups of virtual characters and also includes social actions in the agent decision-making. We follow the Multi-Agent Resource Allocation approach presented in [2], where agents express their preferences using utility functions. The dynamics of social interactions is inspired by the theory of Piaget [3] over which we have implemented reciprocal task exchanges.
Simplified Electrophysiology Modeling Framework to Assess Ventricular Arrhythmia Risk in Infarcted Patients
Patients that have suffered a myocardial infarction are at lifetime high risk for sudden cardiac death (SCD). Personalized 3D computational modeling and simulation can help to find non-invasively arrhythmogenic features of patients’ infarcts, and to provide additional information for stratification and planning of radiofrequency ablation (RFA). Currently, multiscale biophysical models require high computational resources and long simulations times, which make them impractical for clinical environments. In this paper, we develop a phenomenological solver based on cellular automata to simulate cardiac electrophysiology, with results comparable to those of biophysical models. The solver can ru…
Using Inverse Reinforcement Learning with Real Trajectories to Get More Trustworthy Pedestrian Simulations
Reinforcement learning is one of the most promising machine learning techniques to get intelligent behaviors for embodied agents in simulations. The output of the classic Temporal Difference family of Reinforcement Learning algorithms adopts the form of a value function expressed as a numeric table or a function approximator. The learned behavior is then derived using a greedy policy with respect to this value function. Nevertheless, sometimes the learned policy does not meet expectations, and the task of authoring is difficult and unsafe because the modification of one value or parameter in the learned value function has unpredictable consequences in the space of the policies it represents…
Adding Synthetic Detail to Natural Terrain Using a Wavelet Approach
Terrain representation is a basic topic in the field of interactive graphics. The amount of data required for good quality terrain representation offers an important challenge to developers of such systems. For users of these applications the accuracy of geographical data is less important than their natural visual appearance. This makes it possible to mantain a limited geographical data base for the system and to extend it generating synthetic data.In this paper we combine fractal and wavelet theories to provide extra data which keeps the natural essence of actual information available. The new levels of detail(LOD) for the terrain are obtained applying an inverse Wavelet Transform (WT) to…
Emergent behaviors and scalability for multi-agent reinforcement learning-based pedestrian models
This paper analyzes the emergent behaviors of pedestrian groups that learn through the multiagent reinforcement learning model developed in our group. Five scenarios studied in the pedestrian model literature, and with different levels of complexity, were simulated in order to analyze the robustness and the scalability of the model. Firstly, a reduced group of agents must learn by interaction with the environment in each scenario. In this phase, each agent learns its own kinematic controller, that will drive it at a simulation time. Secondly, the number of simulated agents is increased, in each scenario where agents have previously learnt, to test the appearance of emergent macroscopic beha…