We propose a deep learning solution to the problem of object detection in 3D CT images, i.e. the localization and classification of multiple structures. Supervised learning methods require large annotated datasets that are usually difficult to acquire. We thus develop a Cycle Generative Adversarial Network (CycleGAN) + You Only Look Once (YOLO) combined method for CT data augmentation using MRI source images to train a YOLO detector. This results in a fast and accurate detection with a mean average distance of 7.95 ± 6.2 mm, which is significantly better than detection without data augmentation. We show that the approach compares favorably to state-of-the-art detection methods for medical images.

The use of synthetic sequences is one of the most promising tools for advanced in silico evaluation of the quantification of cardiac deformation and strain through 3-D ultrasound (US) and magnetic resonance (MR) imaging. In this paper, we propose the first simulation framework which allows the generation of realistic 3-D synthetic cardiac US and MR (both cine and tagging) image sequences from the same virtual patient. A state-of-the-art electromechanical (E/M) model was exploited for simulating groundtruth cardiac motion fields ranging from healthy to various pathological cases, including both ventricular dyssynchrony and myocardial ischemia. The E/M groundtruth along with template MR/US images and physical simulators were combined in a unified framework for generating synthetic data. We efficiently merged several warping strategies to keep the full control of myocardial deformations while preserving realistic image texture. In total, we generated 18 virtual patients, each with synthetic 3-D US, cine MR, and tagged MR sequences. The simulated images were evaluated both qualitatively by showing realistic textures and quantitatively by observing myocardial intensity distributions similar to real data. In particular, the US simulation showed a smoother myocardium/background interface than the state-of-the-art. We also assessed the mechanical properties. The pathological subjects were discriminated from the healthy ones by both global indexes (ejection fraction and the global circumferential strain) and regional strain curves. The synthetic database is comprehensive in terms of both pathology and modality, and has a level of realism sufficient for validation purposes. All the 90 sequences are made publicly available to the research community via an open-access database.

In this paper, we propose a new emission strategy in the context of plane wave imaging. Plane wave imaging indeed implies compounding in order to preserve a good image quality. Such compounding is usually obtained using multiple, successive emissions, which in turn yields a decrease of the frame rate. As opposed to this approach, our method is based on the simultaneous emission of several coded plane waves. This allows the reconstruction of all the images corresponding to the different plane waves, by using an inverse problem approach. The proposed method is closely related to channel estimation in telecommunications, and coded excitation in synthetic aperture ultrasound imaging. In this paper, we extend these lines of work to the case of ultrasound plane wave imaging and evaluate the obtained performance from various numerical simulations.

In this paper we propose a framework for using duplex Doppler ultrasound systems. These type of systems need to interleave the acquisition and display of a B-mode image and of the pulsed Doppler spectrogram. In a recent study (Richy , 2013), we have shown that compressed sensing-based reconstruction of Doppler signal allowed reducing the number of Doppler emissions and yielded better results than traditional interpolation and at least equivalent or even better depending on the configuration than the study estimating the signal from sparse data sets given in Jensen, 2006. We propose here to improve over this study by using a novel framework for randomly interleaving Doppler and US emissions. The proposed method reconstructs the Doppler signal segment by segment using a block sparse Bayesian learning (BSBL) algorithm based CS reconstruction. The interest of using such framework in the context of duplex Doppler is linked to the unique ability of BSBL to exploit block-correlated signals and to recover non-sparse signals. The performance of the technique is evaluated from simulated data as well as experimental in vivo data and compared to the recent results in Richy , 2013.

In this paper we present a compressed sensing (CS) method adapted to 3D ultrasound imaging (US). In contrast to previous work, we propose a new approach based on the use of learned overcomplete dictionaries that allow for much sparser representations of the signals since they are optimized for a particular class of images such as US images. In this study, the dictionary was learned using the K-SVD algorithm and CS reconstruction was performed on the non-log envelope data by removing 20% to 80% of the original data. Using numerically simulated images, we evaluate the influence of the training parameters and of the sampling strategy. The latter is done by comparing the two most common sampling patterns, i.e., point-wise and line-wise random patterns. The results show in particular that line-wise sampling yields an accuracy comparable to the conventional point-wise sampling. This indicates that CS acquisition of 3D data is feasible in a relatively simple setting, and thus offers the perspective of increasing the frame rate by skipping the acquisition of RF lines. Next, we evaluated this approach on US volumes of several ex vivo and in vivo organs. We first show that the learned dictionary approach yields better performances than conventional fixed transforms such as Fourier or discrete cosine. Finally, we investigate the generality of the learned dictionary approach and show that it is possible to build a general dictionary allowing to reliably reconstruct different volumes of different ex vivo or in vivo organs.

Tracking of moving objects in an image sequence is an important task in many application (e.g. medical imaging, robotics). However, this task is usually difficult due to inherent problems that could happen in sequences (i.e. possible occlusion of the object, large interframe motion). In this paper, we describe a new approach to integrate a priori motion information into a level-set-based active contour approach. Specifically, we introduce a new constraint that enforces the conservation of the levels of the implicit function along the image sequence. This constraint is formulated as a motion prior energy and is used in a tracking algorithm. The method is validated quantitatively on a clinical application based on 10 echocar-diographic and 5 cine-MRI sequences (≈ 700 images).

In this paper we propose a compressed sensing (CS) method adapted to 3D ultrasound imaging (US). In contrast to previous work, we propose a new approach based on the use of learned overcomplete dictionaries. Such dictionaries allow for much sparser representations of the signals since they are optimized for a particular class of images such as US images. We will investigate two undersampling patterns of the 3D US imaging: a spatially uniform random acquisition and a line-wise random acquisition. The latter being extremely interesting for 3D imaging: it would indeed allow skipping the acquisition of many lines among the several thousands required in 3D acquisitions, thus, speeding up the whole acquisition process and incrementing the imaging rate. In this study, the dictionary was learned using the K-SVD algorithm on patches extracted from a training dataset constituted of simulated 3D non-log envelope US volumes. Experiments were performed on a testing dataset made of a simulated 3D US log-envelope volume not included in the testing dataset. CS reconstruction was performed by removing 20% to 80% of the original samples according to the two undersampling patterns. Reconstructions using a K-SVD dictionary previously trained dictionary indicate minimal information loss, thus showing the potential of the overcomplete dictionaries.

Quantification of regional myocardial motion and deformation from cardiac ultrasound is fostering considerable research efforts. Despite the tremendous improvements done in the field, all existing approaches still face a common limitation which is intrinsically connected with the formation of the ultrasound images. Specifically, the reduced lateral resolution and the absence of phase information in the lateral direction highly limit the accuracy in the computation of lateral displacements. In this context, this paper introduces a novel setup for the estimation of cardiac motion with ultrasound. The framework includes an unconventional beamforming technique and a dedicated motion estimation algorithm. The beamformer aims at introducing phase information in the lateral direction by producing transverse oscillations. The estimator directly exploits the phase information in the two directions by decomposing the image into two 2-D single-orthant analytic signals. An in silico evaluation of the proposed framework is presented on five ultra-realistic simulated echocardiographic sequences, where the proposed motion estimator is contrasted against other two phase-based solutions exploiting the presence of transverse oscillations and against block-matching on standard images. An implementation of the new beamforming strategy on a research ultrasound platform is also shown along with a preliminary in vivo evaluation on one healthy subject.

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.

Duplex ultrasonography is a mode of medical ultrasonography that allows one to visualize, at the same time, the inner structure of the body (B-mode) and the blood flow at a particular point in the body (Doppler mode). This mode requires a strategy for alternating B-mode and flow emissions. Traditional strategies either halve the maximum measurable velocity or introduce gaps in the flow data. The objective of this article is to propose a completely original method based on compressive sensing for reconstructing the Doppler signal segment by segment. Our approach is based on randomly alternating B-mode and flow emissions. The influence of the different parameters on the reconstruction quality is studied in detail. The technique is evaluated and its feasability is validated in simulation and from experimental in vivo data. It is also compared to the only method from the literature, proposed by Jensen, that reconstructs blood velocity estimates from sparse data sets.

We present a framework for the simulation of realistic cardiac ultrasound sequences. Both the visual aspect and the synthesized motion mimic a real echocardiography sequence used as template. The resulting simulation appears virtually indistinguishable from a real scan. As the true tissue motion is known, these synthetic sequences can provide a trustful benchmark for the analysis of heart motion. This possibility is illustrated by comparing the performance of two well known motion estimation algorithms.

A new formulation of active contours based on explicit functions has been recently suggested. This novel framework allows real-time 3-D segmentation since it reduces the dimensionality of the segmentation problem. In this paper, we propose a B-spline formulation of this approach, which further improves the computational efficiency of the algorithm. We also show that this framework allows evolving the active contour using local region-based terms, thereby overcoming the limitations of the original method while preserving computational speed. The feasibility of real-time 3-D segmentation is demonstrated using simulated and medical data such as liver computer tomography and cardiac ultrasound images.

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.

We introduce in this paper a formulation of region-based active contours using explicit surfaces. By formally relating the explicit formulation to the usual implicit formulation, we show that it allows to take advantage of the local or global region-based functionals designed in the level-set framework, while providing a fast algorithm yielding close to real-time segmentation of 3D objects. We compare the segmentation performance of the proposed method with the fast level-set based method of Shi [8]. In similar conditions, the proposed method is 26x to 46x faster. Furthermore, we show that it can handle challenging, inhomogeneous data, while keeping a low computational burden.

In this paper we address the segmentation of shapes which may be approximated by two elliptical contours. Such patterns are indeed recurrent in many image processing applications. In this context, we develop a level-set framework especially dedicated to the detection of annular-like shapes. The behavior of this approach is illustrated on images from different fields. An evaluation is then performed for myocardium detection in medical images.

This paper describes a free open source software in Matlab (named Creaseg, http://www.creatis.insa-lyon. fr/~bernard/creaseg) for the evaluation of the performance of different level-set based algorithms in the context of 2D image segmentation. The platform gives access to the implementation of six level-set methods that have been chosen in order to cover a wide range of data attachment terms (contour, region and localized approaches). The software also gives the possibility to compare the performance of the proposed algorithms on any kind of images. The performance can be evaluated either visually, or from similarity measurements between a reference and the results of the segmentation.

In the field of image segmentation, most level-set-based active-contour approaches take advantage of a discrete representation of the associated implicit function. We present in this paper a different formulation where the implicit function is modeled as a continuous parametric function expressed on a B-spline basis. Starting from the active-contour energy functional, we show that this formulation allows us to compute the solution as a restriction of the variational problem on the space spanned by the B-splines. As a consequence, the minimization of the functional is directly obtained in terms of the B-spline coefficients. We also show that each step of this minimization may be expressed through a convolution operation. Because the B-spline functions are separable, this convolution may in turn be performed as a sequence of simple 1-D convolutions, which yields an efficient algorithm. As a further consequence, each step of the level-set evolution may be interpreted as a filtering operation with a B-spline kernel. Such filtering induces an intrinsic smoothing in the algorithm, which can be controlled explicitly via the degree and the scale of the chosen B-spline kernel. We illustrate the behavior of this approach on simulated as well as experimental images from various fields.

This work presents an algorithm for the segmentation of myocardial regions in echocardiography imaging based on the statistics of the radiofrequency image. We formulate the problem of segmentation in a maximum likelihood framework using the Generalized Gaussian as an a priori distribution. We minimize the resulting functional using a radial basis functions-based multiphase level set model. Numerical results obtained on both simulation and in vivo data demonstrate the capacity of our approach to segment myocardial regions in echocardiography imaging.

We consider a recent parametric level-set segmentation approach where the implicit interface is the zero level of a continuous function expanded onto compactly supported radial basis functions, defined by their centers, coefficients and supports. We propose to introduce prior knowledge of the shape to be recovered by placing the centers quasi-uniformly over an uncertainty area.

The partial differential equation driving level-set evolution in segmentation is usually solved using finite differences schemes. In this paper, we propose an alternative scheme based on radial basis functions (RBFs) collocation. This approach provides a continuous representation of both the implicit function and its zero level set. We show that compactly supported RBFs (CSRBFs) are particularly well suited to collocation in the framework of segmentation. In addition, CSRBFs allow us to reduce the computation cost using a kd-tree-based strategy for neighborhood representation. Moreover, we show that the usual reinitialization step of the level set may be avoided by simply constraining the l1-norm of the CSRBF parameters. As a consequence, the final solution is topologically more flexible, and may develop new contours (i.e., new zero-level components), which are difficult to obtain using reinitialization. The behavior of this approach is evaluated from numerical simulations and from medical data of various kinds, such as 3-D CT bone images and echocardiographic ultrasound images.

Motion and deformation analysis of the myocardium are of utmost interest in cardiac imaging. Part of the, research is devoted to the estimation of the heart function by analysis of the shape changes of the left-ventricular endocardial surface. However, most clinically used shape-based approaches are often two-dimensional (2-D) and based on the analysis of the shape at only two cardiac instants. Three-dimensional (3-D) approaches generally make restrictive hypothesis about the actual endocardium motion to be able to recover a point-to-point correspondence between two surfaces. The present work is a first step toward the automatic spatio-temporal analysis and recognition of deformable surfaces. A curvature-based and easily interpretable description of the surfaces is derived. Based on this description, shape dynamics is first globally estimated through the temporal shape spectra. Second, a regional curvature-based tracking approach is proposed assuming a smooth deformation. It combines geometrical and spatial information in order to analyze a specific endocardial region. These methods are applied both on true 3-D X-ray data and on simulated normal and abnormal left ventricles. The results are coherent and easily interpretable. Shape dynamics estimations and comparisons between deformable object sequences are now possible through these techniques. This promising framework is a suitable tool for a complete regional description of deformable surfaces.