Data Augmentation for Automatic Identification of Spatiotemporal Dispersion Electrograms in Persistent Atrial Fibrillation Ablation Using Machine Learning.

Catheter ablation is increasingly used to treat atrial fibrillation (AF), the most common sustained cardiac arrhythmia encountered in clinical practice. A recent breakthrough finding in AF ablation consists in identifying ablation sites based on their spatiotemporal dispersion (STD). STD stands for a delay of the cardiac activation observed in intracardiac electrograms (EGMs) across contiguous leads. In practice, interventional cardiologists localize STD sites visually using the PentaRay multipolar mapping catheter. This work aims at automatically characterizing STD by classifying EGM data into STD vs. non STD groups using machine learning (ML) techniques. A dataset of 23082 multichannel EGM recordings acquired by the PentaRay coming from 16 persistent AF patients is included in this study. A major problem hampering the classification performance lies in the highly imbalanced dataset ratio. We suggest to tackle data imbalance using adapted data augmentation techniques including 1) undersampling 2) oversampling 3) lead shift 4) time reversing and 5) time shift. These tools are designed to preserve the integrity of the cardiac data and are validated by a partner cardiologist. They provide enhancement in classification performance in terms of sensitivity, which increases from 50% to 80% while maintaining accuracy and AUC around 90% with oversampling. Bootstrapping is applied to check the variability of the trained classifiers.Clinical relevance-The machine learning techniques developed in this contribution are expected to aid cardiologists in performing patient-tailored catheter ablation procedures for treating persistent AF.

OntoVIP: an ontology for the annotation of object models used for medical image simulation.

This paper describes the creation of a comprehensive conceptualization of object models used in medical image simulation, suitable for major imaging modalities and simulators. The goal is to create an application ontology that can be used to annotate the models in a repository integrated in the Virtual Imaging Platform (VIP), to facilitate their sharing and reuse. Annotations make the anatomical, physiological and pathophysiological content of the object models explicit. In such an interdisciplinary context we chose to rely on a common integration framework provided by a foundational ontology, that facilitates the consistent integration of the various modules extracted from several existing ontologies, i.e. FMA, PATO, MPATH, RadLex and ChEBI. Emphasis is put on methodology for achieving this extraction and integration. The most salient aspects of the ontology are presented, especially the organization in model layers, as well as its use to browse and query the model repository.

A virtual imaging platform for multi-modality medical image simulation.

This paper presents the Virtual Imaging Platform (VIP), a platform accessible at http://vip.creatis.insa-lyon.fr to facilitate the sharing of object models and medical image simulators, and to provide access to distributed computing and storage resources. A complete overview is presented, describing the ontologies designed to share models in a common repository, the workflow template used to integrate simulators, and the tools and strategies used to exploit computing and storage resources. Simulation results obtained in four image modalities and with different models show that VIP is versatile and robust enough to support large simulations. The platform currently has 200 registered users who consumed 33 years of CPU time in 2011.

Technical support for Life Sciences communities on a production grid infrastructure.

Production operation of large distributed computing infrastructures (DCI) still requires a lot of human intervention to reach acceptable quality of service. This may be achievable for scientific communities with solid IT support, but it remains a show-stopper for others. Some application execution environments are used to hide runtime technical issues from end users. But they mostly aim at fault-tolerance rather than incident resolution, and their operation still requires substantial manpower. A longer-term support activity is thus needed to ensure sustained quality of service for Virtual Organisations (VO). This paper describes how the biomed VO has addressed this challenge by setting up a technical support team. Its organisation, tooling, daily tasks, and procedures are described. Results are shown in terms of resource usage by end users, amount of reported incidents, and developed software tools. Based on our experience, we suggest ways to measure the impact of the technical support, perspectives to decrease its human cost and make it more community-specific.

NeuroLOG: sharing neuroimaging data using an ontology-based federated approach.

This paper describes the design of the NeuroLOG middleware data management layer, which provides a platform to share heterogeneous and distributed neuroimaging data using a federated approach. The semantics of shared information is captured through a multi-layer application ontology and a derived Federated Schema used to align the heterogeneous database schemata from different legacy repositories. The system also provides a facility to translate the relational data into a semantic representation that can be queried using a semantic search engine thus enabling the exploitation of knowledge embedded in the ontology. This work shows the relevance of the distributed approach for neurosciences data management. Although more complex than a centralized approach, it is also more realistic when considering the federation of large data sets, and open strong perspectives to implement multi-centric neurosciences studies.

Grid-wide neuroimaging data federation in the context of the NeuroLOG project.

Grid technologies are appealing to deal with the challenges raised by computational neurosciences and support multi-centric brain studies. However, core grids middleware hardly cope with the complex neuroimaging data representation and multi-layer data federation needs. Moreover, legacy neuroscience environments need to be preserved and cannot be simply superseded by grid services. This paper describes the NeuroLOG platform design and implementation, shedding light on its Data Management Layer. It addresses the integration of brain image files, associated relational metadata and neuroscience semantic data in a heterogeneous distributed environment, integrating legacy data managers through a mediation layer.

Towards production-level cardiac image analysis with grids.

Production exploitation of cardiac image analysis tools is hampered by the lack of proper IT infrastructure in health institutions, the non trivial integration of heterogeneous codes in coherent analysis procedures, and the need to achieve complete automation of these methods. HealthGrids are promising technologies to address these difficulties. This paper details how they can be complemented by high level problem solving environments such as workflow managers to improve the performance of applications both in terms of execution time and robustness of results. Two of the most important important cardiac image analysis tasks are considered, namely myocardium segmentation and motion estimation in a 4D sequence. Results are shown on the corresponding pipelines, using two different execution environments on the EGEE grid production infrastructure.

A distributed security policy for neuroradiology data sharing.

Grids are key technologies to federate data distributed in multiple neuroscience centers, thus enabling large scale multi-centric studies. However, the take up of these technologies is slow due to the difficulty to manipulate sensitive neuroradiological data in an open environment and the recognized risk of federated sites to loose control over their valuable data. In this paper we propose a distributed data access control policy, enabling the federation of existing data stores, where local security policies prevail, to supports multi-centric neuroscience studies. It achieves a compromise between enabling collaborative work through data sharing and preventing unauthorized access to data in a competitive environment.

NeuroLOG: a community-driven middleware design.

The NeuroLOG project designs an ambitious neurosciences middleware, gaining from many existing components and learning from past project experiences. It is targeting a focused application area and adopting a user-centric perspective to meet the neuroscientists expectations. It aims at fostering the adoption of HealthGrids in a pre-clinical community. This paper details the project's design study and methodology which were proposed to achieve the integration of heterogeneous site data schemas and the definition of a site-centric policy. The NeuroLOG middleware will bridge HealthGrid and local resources to match user desires to control their resources and provide a transitional model towards HealthGrids.

Bridging clinical information systems and grid middleware: a Medical Data Manager.

This paper describes the effort to deploy a Medical Data Management service on top of the EGEE grid infrastructure. The most widely accepted medical image standard, DICOM, was developed for fulfilling clinical practice. It is implemented in most medical image acquisition and analysis devices. The EGEE middleware is using the SRM standard for handling grid files. Our prototype is exposing an SRM compliant interface to the grid middleware, transforming on the fly SRM requests into DICOM transactions. The prototype ensures user identification, strict file access control and data protection through the use of relevant grid services. This Medical Data Manager is easing the access to medical databases needed for many medical data analysis applications deployed today. It offers a high level data management service, compatible with clinical practices, which encourages the migration of medical applications towards grid infrastructures. A limited scale testbed has been deployed as a proof of concept of this new service. The service is expected to be put into production with the next EGEE middleware generation.

Medical image registration algorithms assesment: Bronze Standard application enactment on grids using the MOTEUR workflow engine.

Medical image registration is pre-processing needed for many medical image analysis procedures. A very large number of registration algorithms are available today, but their performance is often not known and very difficult to assess due to the lack of gold standard. The Bronze Standard algorithm is a very data and compute intensive statistical approach for quantifying registration algorithms accuracy. In this paper, we describe the Bronze Standard application and we discuss the need for grids to tackle such computations on medical image databases. We demonstrate MOTEUR, a service-based workflow engine optimized for dealing with data intensive applications. MOTEUR eases the enactment of the Bronze Standard and similar applications on the EGEE production grid infrastructure. It is a generic workflow engine, based on current standards and freely available, that can be used to instrument legacy application code at low cost.

Performance evaluation of grid-enabled registration algorithms using bronze-standards.

Evaluating registration algorithms is difficult due to the lack of gold standard in most clinical procedures. The bronze standard is a real-data based statistical method providing an alternative registration reference through a computationally intensive image database registration procedure. We propose in this paper an efficient implementation of this method through a grid-interfaced workflow enactor enabling the concurrent processing of hundreds of image registrations in a couple of hours only. The performances of two different grid infrastructures were compared. We computed the accuracy of 4 different rigid registration algorithms on longitudinal MRI images of brain tumors. Results showed an average subvoxel accuracy of 0.4 mm and 0.15 degrees in rotation.

Authentication and authorisation prototype on the mu grid for medical data management.

This paper presents mu grid, a light weight middleware for grid applications, and focuses mainly on security issues--more specifically on the access control to resources--that are critical for the gridification of many medical applications. For this purpose, we use Sygn as a distributed, certificate based, and flexible access control mechanism, which has been fully integrated in mu grid. We discuss the advantages of the solution compared to classical grid approaches and the limitations of the final architecture.

4D deformable models with temporal constraints: application to 4D cardiac image segmentation.

Segmentation of time series of 3D cardiac images is clinically used for the assessment of the mechanical function of the left ventricle. To take into account the 4D (3D+T) nature of those images, we propose to extend the deformable surface framework by introducing time-dependent constraints. Thus, in addition to computing an internal force for enforcing the regularity of the deformable model, prior motion knowledge is introduced in the deformation process through either temporal smoothing or trajectory constraints. In this paper, deformable surfaces are represented as simplex meshes owing to their generality and their ability to compute mean curvature at each vertex. The segmentation accuracy of this 4D deformable model is estimated on synthetic SPECT image sequences for which a ground truth about the LV volume is known. Segmentation of non-synthetic SPECT and other modalities 4D images is also discussed.