The advent of deep learning has pushed medical image analysis to new levels, rapidly replacing more traditional machine learning and computer vision pipelines. However segmenting and labelling anatomical regions remains challenging owing to appearance variations, imaging artifacts, the paucity and variability of annotated data, and the difficulty of fully exploiting domain constraints such as anatomical knowledge about inter-region relationships. We address the last point, improving the network's region-labeling consistency by introducing NonAdjLoss, an adjacency-graph based auxiliary training loss that penalizes outputs containing regions with anatomically-incorrect adjacency relationships. NonAdjLoss supports both fully-supervised training and a semi-supervised extension in which it is applied to unlabeled supplementary training data. The approach substantially reduces segmentation anomalies on the MICCAI-2012, IBSRv2 brain MRI datasets and the Anatomy3 whole body CT dataset, especially when semi-supervised training is included.

Simulation is a fast, controlled, and reproducible way to evaluate new algorithms for distributed computing platforms in a variety of conditions. However, the realism of simulations is rarely assessed, which critically questions the applicability of a whole range of findings. In this paper, we present our efforts to build platform models from application traces, to allow for the accurate simulation of file transfers across a distributed infrastructure. File transfers are key to performance, as the variability of file transfer times has important consequences on the dataflow of the application. We present a methodology to build realistic platform models from application traces and provide a quantitative evaluation of the accuracy of the derived simulations. Results show that the proposed models are able to correctly capture real-life variability and significantly outperform the state-of-the-art model.

In this paper, we propose a mean-shift formulation allowing spatiotemporal clustering of video streams, and possibly extensible to other multivariate evolving data. Our formulation enables causal or omniscient filtering of spatiotemporal data, which is robust to total object occlusions. It embeds a new clustering algorithm within the filtering procedure that will group samples and reduce their number over the iterations. Based on our formulation, we express similar approaches and assess their robustness on real video sequences.

Most time-series clustering methods, such as k-means or k-medoids, are initialized by prior knowledge about the number of classes or by a learning step. We propose an unsupervised clustering technique based on spatiotemporal mean-shift and optimal time series warping using dynamic time warping (DTW). Our main contribution consists in combining a spatiotemporal filtering technique, which gathers similar and synchronized temporal patterns in image sequences, with a clustering algorithm that applies a trajectory constraint on the DTW associations, thereby discriminating between similar time-series that are temporally shifted or warped. We assess the method's robustness on synthetic data, and demonstrate its versatility on brain magnetic resonance and multispectral satellite image sequences.

In this paper, we propose a new anisotropic diffusion formulation allowing non-linear spatiotemporal filtering of image sequences. We first formulate a multidimensional spatiotemporal diffusion equation based on Barash's iterative form, processing independently both spatial, temporal and intensity dimensions with their own diffusion functions and scale parameters. We then introduce a local regularization term designed to smooth the remaining spike noise. Experimental results processed on synthetic data and real MR images show that considering the temporal information and the regularization term improves the filtering quality. The method is also shown to be robust to noise, blur and temporal intensity evolution. Results are compared to the BM3D method with MSE and SSIM evaluation metrics.

This paper presents an end-to-end SimGrid-based simulation of a Monte-Carlo computation deployed on the European Grid Infrastructure. We describe a simulation framework allowing to replay executions from real traces. Middleware services, namely the file catalog, storage elements, the DIRAC pilot-job system, and the MOTEUR workflow engine are simulated by SimGrid processes. The deployment of pilot jobs, performed on EGI by the gLite WMS and batch queues, is simulated by a random selection of platform hosts, and a matching of their latencies and failure times with real traces. The Monte-Carlo application workflow is calibrated to address performance discrepancies between the real and simulated network and CPUs. The simulation is evaluated against real executions of the GATE Monte-Carlo application. Results show that SimGrid can be used to study the performance of applications running on EGI. Simulated and real make spans are consistent, and conclusions drawn in production about the influence of application parameters such as the check pointing frequency and the number of mergers are also made in simulation. This opens the door to better and faster experimentation on production grids.

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.

We propose a framework for the segmentation by region growing approach leveraging on feature space. It has the advantages to deal with multidimensional data and easily specify locally adaptive segmentation. It relies upon the definition of a robust neighborhood which drives the region growing. We propose two applications to illustrate this framework: a segmentation of physical parameters maps of MRI by using n-dimensional region growing and a segmentation of highly noisy image by using adaptive region growing.

Medical image simulation produces virtual images from software representations of imaging devices and virtual object models representing the human body. Object models consist of the geometry of the objects (e.g. organs, tissues, pathological structures, etc.) and of their physical parameters used for the simulation. The diversity of this information makes it difficult to share and reuse across simulation modalities and users. We address this issue by explicitly describing object models using a semantic approach. In particular, we developed an ontology that contains the relevant concepts and relations of the domain of object models for image simulation. This ontology is used to annotate object models and to describe their content and structure. In this paper, we present the construction steps of this ontology, the representation choices and we illustrate how it is used to annotate object models.

This paper describes a framework for the integration of medical image simulators in the Virtual Imaging Platform (VIP). Simulation is widely involved in medical imaging but its availability is hampered by the heterogeneity of software interfaces and the required amount of computing power. To address this, VIP defines a simulation workflow template which transforms object models from the IntermediAte Model Format (IAMF) into native simulator formats and parallelizes the simulation computation. Format conversions, geometrical scene definition and physical parameter generation are covered. The core simulator executables are directly embedded in the simulation workflow, enabling data parallelism exploitation without modifying the simulator. The template is instantiated on simulators of the four main medical imaging modalities, namely Positron Emission Tomography, Ultrasound imaging, Magnetic Resonance Imaging and Computed Tomography. Simulation examples and performance results on the European Grid Infrastructure are shown.

Mean shift is an efficient filtering algorithm processing multidimensional data as color images. Such algorithm needs few tuning parameters named scale parameters. In this paper, we study the impact of the color space used on the results quality. Two linear transformations of the RGB space (Y'UV and PC A) and a non linear one (L*a*b* color space) are addressed. The results quality is assessed using the PSNR and the SSIM, a consistent measure with human eye perception. To determine the optimal color space, we use an exhaustive search of the scale parameters. This study reminds that PC A transformation is useless for mean shift and shows (using 5 natural color images and 2 synthesized data) that optimizing the bandwidth parameters in the L*a*b* space helps in improving the mean shift filtering assessed by PSNR.

Medical imaging research deals with large, heterogeneous and fragmented amounts of medical images. The need for secure, federated and functional medical image databases is very strong within these research communities. This paper provides an overview of the different projects concerned with building medical image databases for medical imaging research. It also discusses the characteristics and requirements of this community and tries to determine to what extent existing solutions can answer these specific requirements.

Magnetic Resonance Imaging (MRI) is a medical imaging modality that provides structural and functional information. To improve the contrast of MR images, contrast agents, such as Ultrasmall SuperParamagnetic Iron Oxide particles (USPIO), are more commonly being used. Quantifying and locating these nanoparticles is of high interest. An efficient technique consists in images analyzing of the default field inhomogeneities induced by USPIOs. In a previous work, we have introduced such a quantification framework. Here, we improve our approach by deriving from two methods: Wiener and CLS filtering. Both methods have been evaluated on realistic data demonstrating that CLS filtering gives very promising quantification results.

We present an adaptive solution for guinea pig knee cartilage segmentation using a 3-D smoothing B-Spline active surface. An adaptive parametric combination of edge-based forces and balloon force solves the problem of capture range of external forces. The comparison between the results of the experiments using this method and previous 3-D validated snake segmentation shows that the accuracy and robustness are improved.

A new adaptive IIR filtering of surface meshes by smoothing B-spline is proposed. We define some ad hoc processing of borders and singular points for genus 0 meshes. By adapting the filter cut-off frequency to the local mesh sampling, it enables homogeneous smoothing of non-uniform meshes. We present genus 0 and genus 1 meshes filtering as well as volume segmentations. These segmentations have been obtained using the IIR filtering of deformation forces within a smoothing B-Spline active surface.

In this paper, we propose an original correction method for susceptibility artifact in spin echo and gradient echo MR images when a field map of the imaged object is known. The susceptibility artifact impact on MR images is modeled by means of a pixel shifting map and an intensity distortion map. An original scheme based on MRI simulation is used to build the intensity distortion map to take into account different complex phenomena such as intravoxel dephasing. The proposed method has been successfully tested on both spin echo and gradient echo MR images giving promising results.

In this paper, a new active surface based segmentation algorithm is proposed. We detail the regularization process that is conducted through a smoothing B-spline filtering of the mesh point displacements. The corresponding segmentation algorithm is presented and applied on medical data. The proposed segmentation method is fast and suitable to quadrangular meshes.

In this paper, we give an overview of a python language library named "DaVaW" designed to ease the development of medical image processing applications. The main originality of this library is to provide high level components like medical image browser, 3D viewers or ROI manager, as well as a programming pattern that enables communication and interaction between these components. This library has an open source distribution and is mainly based on VTK and WxPython libraries.

In this paper, we present the implementation of a Magnetic Resonance Imaging (MRI) simulator on a GRID computing architecture. The simulation process is based on the resolution of Bloch equation [1] in a 3D space. The computation kernel of the simulator is distributed to the grid nodes using MPICH-G2 [2]. The results presented show that simulation of 3D MRI data is achieved with a reasonable cost which gives new perspectives to MRI simulations usage.

In this paper, we propose new evaluation measures for scene segmentation results, which are based on computing the difference between a region extracted from a segmentation map and the corresponding one on an ideal segmentation. The proposed measures take into account separately both under and over detected pixels. It also associates in its computation the compactness of the region under investigation.

We propose a set of scalable discrepancy measures that may be applied for segmentation evaluation when a reference is known. The proposed measures take into account under and over detected points within an adjustable area. They give the intensity of the discrepancy and its relative position. Furthermore a scale parameter allows the accuracy of the measures to be adjusted.

In this paper, we consider the texture coding of quadrangular cells which compose active meshes. Active mesh algorithms which have been developed for region based video coding mainly use interpolation techniques to quantize the cell texture. We propose a new quantization scheme which associates image warping and an adaptive JPEG scheme. Our approach leads to an interesting trade-off between mesh complexity and image quality.

Two enhanced subband coding schemes using a regularized image restoration technique are proposed: the first controls the global regularity of the decompressed image; the second extends the first approach at each decomposition level. The quantization scheme incorporates scalar quantization (SQ) and pyramidal lattice vector quantization (VQ) with both optimal bit and quantizer allocation. Experimental results show that both the block effect due to VQ and the quantization noise are significantly reduced.

We consider the coding of a quadrangular mesh structure generated by a multiresolution deformable quadtree algorithm. Such a structure has been found efficient for region based video coding, but few works deal with the coding of the mesh structure. We propose to use the rules of the mesh generation to significantly reduce the number of bits required for the coding of the nodes position defining the mesh structure. This reduction induces an improvement of either the image quality or either the global bit rate.

The influence of the phase of conjugate quadrature filters (CQFs) on the performances of a subband coding scheme is analyzed. When the filter length is short, the phase characteristic has virtually no influence on the peak signal-to-noise ratio (PSNR)/bit rate results and on the performance of postprocessing algorithms such as edge detection.

In this study, we consider the DCT based subband decomposition and show the efficiency of this approach for low contrast images in comparison to 9-7 bi-orthogonal FIR filters generally used in such a scheme. First, we consider the theoretical coding gain for the two approaches. Afterwards, the distortion versus rate performances are analyzed and the perceptual qualities of the reconstructed images are discussed. Finally, an edge detection operator is applied on the reconstructed images in order to show the robustness of DCT based subband decomposition to post-processing.