A5037356492- Medical Image Segmentation Techniques
- Ultrasound Imaging and Elastography
- Cardiovascular Function and Risk Factors
- Cardiac Valve Diseases and Treatments
- Advanced MRI Techniques and Applications
- Cardiac Imaging and Diagnostics
- Photoacoustic and Ultrasonic Imaging
- Ultrasonics and Acoustic Wave Propagation
- Medical Imaging Techniques and Applications
- Pediatric Hepatobiliary Diseases and Treatments
- Radiomics and Machine Learning in Medical Imaging
- Liver Disease and Transplantation
- Advanced Neural Network Applications
- Image and Object Detection Techniques
- Organ Transplantation Techniques and Outcomes
- Cardiovascular Health and Disease Prevention
- Image and Signal Denoising Methods
- Sparse and Compressive Sensing Techniques
- Coronary Interventions and Diagnostics
- Liver Disease Diagnosis and Treatment
- Advanced Vision and Imaging
- AI in cancer detection
- Advanced Numerical Analysis Techniques
- Non-Invasive Vital Sign Monitoring
- Electrical and Bioimpedance Tomography
Inserm
2013-2024
Université Claude Bernard Lyon 1
2014-2024
Centre de Recherche en Acquisition et Traitement de l'Image pour la Santé
2015-2024
Centre National de la Recherche Scientifique
2012-2024
Laboratoire d’Imagerie Biomédicale
2015-2024
Institut Universitaire de France
2024
Université de Montpellier
2024
Bicêtre Hospital
1997-2023
Institut National des Sciences Appliquées de Lyon
2013-2023
Université de Poitiers
2009-2023
Delineation of the left ventricular cavity, myocardium, and right ventricle from cardiac magnetic resonance images (multi-slice 2-D cine MRI) is a common clinical task to establish diagnosis. The automation corresponding tasks has thus been subject intense research over past decades. In this paper, we introduce "Automatic Cardiac Diagnosis Challenge" dataset (ACDC), largest publicly available fully annotated for purpose MRI (CMR) assessment. contains data 150 multi-equipments CMRI recordings...
Delineation of the cardiac structures from 2D echocardiographic images is a common clinical task to establish diagnosis. Over past decades, automation this has been subject intense research. In paper, we evaluate how far state-of-the-art encoder-decoder deep convolutional neural network methods can go at assessing images, i.e., segmenting and estimating indices, on dataset, especially, designed answer objective. We, therefore, introduce acquisitions for multi-structure ultrasound...
Plane-Wave imaging enables very high frame rates, up to several thousand frames per second. Unfortunately the lack of transmit focusing leads reduced image quality, both in terms resolution and contrast. Recently, numerous beamforming techniques have been proposed compensate for this effect, but comparing different methods is difficult due appropriate tools. PICMUS, Imaging Challenge Medical Ultrasound aims provide these This paper describes PICMUS challenge, its motivation, implementation, metrics.
Progressive familial intrahepatic cholestasis (PFIC) is a lethal inherited childhood of hepatocellular origin. Different subtypes PFIC have been described according to serum gamma–glutamyl transpeptidase (GGT) activity. There currently no effective medical therapy available for children with PFIC. We report on 39 patients who received ursodeoxycholic acid (UDCA) orally (20–30 mg/kg b.w./day) period 2 4 years. Group 1 (n = 26) consisted normal GGT activity, and group 13) high Within 1, liver...
In the field of image segmentation, most level-set-based active-contour approaches take advantage a discrete representation associated implicit function. We present in this paper different formulation where function is modeled as continuous parametric expressed on B-spline basis. Starting from energy functional, we show that allows us to compute solution restriction variational problem space spanned by B-splines. As consequence, minimization functional directly obtained terms coefficients....
Convolutional neural networks (CNN) have had unprecedented success in medical imaging and, particular, image segmentation. However, despite the fact that segmentation results are closer than ever to inter-expert variability, CNNs not immune producing anatomically inaccurate segmentations, even when built upon a shape prior. In this paper, we present framework for cardiac maps guaranteed respect pre-defined anatomical criteria, while remaining within variability. The idea behind our method is...
Quantification of cardiac deformation and strain with 3D ultrasound takes considerable research efforts. Nevertheless, a widespread use these techniques in clinical practice is still held back due to the lack solid verification process quantify compare performance. In this context, fully synthetic sequences has become an established tool for initial silico evaluation. realism existing simulation too limited represent reliable benchmarking data. Moreover, fact that different centers typically...
Real-time 3D Echocardiography (RT3DE) has been proven to be an accurate tool for left ventricular (LV) volume assessment. However, identification of the LV endocardium remains a challenging task, mainly because low tissue/blood contrast images combined with typical artifacts. Several semi and fully automatic algorithms have proposed segmenting in RT3DE data order extract relevant clinical indices, but systematic fair comparison between such methods so far impossible due lack publicly...
Volume and ejection fraction (EF) measurements of the left ventricle (LV) in 2-D echocardiography are associated with a high uncertainty not only due to interobserver variability manual measurement, but also ultrasound acquisition errors such as apical foreshortening. In this work, real-time fully automated EF measurement foreshortening detection method is proposed. The uses several deep learning components, view classification, cardiac cycle timing, segmentation landmark extraction, measure...
Convolutional neural networks (CNN) have demonstrated their ability to segment 2D cardiac ultrasound images. However, despite recent successes according which the intra-observer variability on end-diastole and end-systole images has been reached, CNNs still struggle leverage temporal information provide accurate temporally consistent segmentation maps across whole cycle. Such consistency is required accurately describe function, a necessary step in diagnosing many cardiovascular diseases. In...
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 problem. In this paper, we propose a B-spline approach, which further improves computational efficiency algorithm. We also show that evolving contour using local region-based terms, thereby overcoming limitations original method while preserving speed. The feasibility is demonstrated simulated and medical data...
This paper presents the Virtual Imaging Platform (VIP), a platform accessible at http://vip.creatis.insa-lyon.fr to facilitate sharing of object models and medical image simulators, provide access distributed computing storage resources. A complete overview is presented, describing ontologies designed share in common repository, workflow template used integrate tools strategies exploit Simulation results obtained four modalities with different show that VIP versatile robust enough support...
We present a method for the analysis of heart motion from medical images. The algorithm exploits monogenic signal theory, recently introduced as an N-dimensional generalization analytic signal. displacement is computed locally by assuming conservation phase over time. A local affine model considered to account typical motions contraction/expansion and shear. coarse-to-fine B-spline scheme allows robust effective computation model's parameters, pyramidal refinement helps handle large motions....
In this paper we present a compressed sensing (CS) method adapted to 3D ultrasound imaging (US). contrast previous work, propose new approach based on the use of learned overcomplete dictionaries that allow for much sparser representations signals since they are optimized particular class images such as US images. study, dictionary was using K-SVD algorithm and CS reconstruction performed non-log envelope data by removing 20% 80% original data. Using numerically simulated images, evaluate...
Purpose: Deformable registration generally relies on the assumption that sought spatial transformation is smooth. Yet, breathing motion involves sliding of lung with respect to chest wall, causing a discontinuity in field, and smoothness can lead poor matching accuracy. In response, alternative methods have been proposed, several which rely prior segmentations. We propose an original method for automatically extracting particular segmentation, called mask, from CT image thorax. Methods: The...
Cardiac volume/function assessment remains a critical step in daily cardiology, and 3-D ultrasound plays an increasingly important role. Fully automatic left ventricular segmentation is, however, challenging task due to the artifacts low contrast-to-noise ratio of imaging. In this paper, fast fully framework for full-cycle endocardial ventricle is proposed. This approach couples advantages B-spline explicit active surfaces framework, purely image information approach, those statistical shape...
We present the UltraSound ToolBox (USTB), a processing framework for ultrasound signals. USTB aims to facilitate comparison of imaging techniques and dissemination research results. It fills void tools algorithm sharing verification, enables solid assessment correctness relevance new approaches. also boost productivity by cutting down implementation time code maintenance. is MATLAB toolbox 2D 3D data, supporting both C++ implementations. Channel data from any origin, simulated experimental,...
Intraventricular vector flow mapping (iVFM) seeks to enhance and quantify color Doppler in cardiac imaging.In this study, we propose novel alternatives the traditional iVFM optimization scheme by utilizing physicsinformed neural networks (PINNs) a physics-guided nnU-Net-based supervised approach.When evaluated on simulated images derived from patientspecific computational fluid dynamics model vivo acquisitions, both approaches demonstrate comparable reconstruction performance original...