Research Interests:
Research Interests:
Research Interests:
Research Interests: CNN and Data Sampling
Research Interests:
International audienceThe detection of intracranial aneurysms from Magnetic Resonance Angiography images is a problem of rapidly growing clinical importance, but also extremely challenging to automate. However, in the last 3 years, the... more
International audienceThe detection of intracranial aneurysms from Magnetic Resonance Angiography images is a problem of rapidly growing clinical importance, but also extremely challenging to automate. However, in the last 3 years, the raise of deep convolutional neural networks has instigated a streak ofmethods that have convincingly removed the technological deadlock and show promising performance. The major issue to address is the very severe class imbalance. Previous authors have focused their efforts on the network architecture and loss function. This paper tackles the data. A rough but fast annotation is considered : each aneurysm is approximated by a sphere defined by two points. Second, a small patch approach is taken so as to increase the number of samples. Third, samples are generated by a combination of data selection (negative patches are centered half on blood vessels and half on parenchyma) and data synthesis (patches containing an aneurysm are duplicated and deformed ...
Research Interests: CNN and Data Sampling
Research Interests:
Research Interests:
electromagnetic sensors and ultrasound images for
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
L’apport de la simulation numerique a la radiologie interventionnelle ? Une formation medicale enrichie et un temps de traitement plus court au benefice du patient.
Research Interests: Philosophy and Simulation
Research Interests: Computer Science, Computer Graphics, Surgery, Computational Modeling, Biomechanics, and 12 moreAugmented Reality, Medicine, Computer assisted orthopaedic surgery, Humans, Computer Simulation, Liver, Minimally Invasive Surgery, Computer User Interface Design, Soft Tissue, Computer assisted surgery, Liver neoplasms, and liver tissue
Research Interests:
Minimally invasive fluoroscopy-based procedures are the gold standard for diagnosis and treatment of various pathologies of the cardiovascular system. This kind of procedures imply for the clinicians to infer the 3D shape of the device... more
Minimally invasive fluoroscopy-based procedures are the gold standard for diagnosis and treatment of various pathologies of the cardiovascular system. This kind of procedures imply for the clinicians to infer the 3D shape of the device from 2D images, which is known to be an ill-posed problem. In this paper we present a method to reconstruct the 3D shape of the interventional device, with the aim of improving the navigation. The method combines a physics-based simulation with non-linear Bayesian filter. Whereas the physics-based model provides a prediction of the shape of the device navigating within the blood vessels (taking into account non-linear interactions between the catheter and the surrounding anatomy), an Unscented Kalman Filter is used to correct the navigation model using 2D image features as external observations. The proposed framework has been evaluated on both synthetic and real data, under different model parameterization, filter parameters tuning and external obser...
Research Interests:
In this paper1, a framework to acquire and process dynamic data of the tongue during speech processing is presented. First, a setup to acquire data of the tongue shape combining ultrasound images, electromagnetic localization sensors and... more
In this paper1, a framework to acquire and process dynamic data of the tongue during speech processing is presented. First, a setup to acquire data of the tongue shape combining ultrasound images, electromagnetic localization sensors and sound is presented. Techniques to automatically calibrate and synchronize the data are described. A method to extract the tongue shape is then proposed, by combining a preprocessing of the ultrasound images with an image-based tracking method that integrates adapted constraints.
Research Interests:
Les travaux de recherche repris de maniere synthetique dans ce memoire, abordent les questions liees a l'imagerie medicale multimodale et interventionnelle dans deux contextes applicatifs. Le premier traite de l'acquisition de... more
Les travaux de recherche repris de maniere synthetique dans ce memoire, abordent les questions liees a l'imagerie medicale multimodale et interventionnelle dans deux contextes applicatifs. Le premier traite de l'acquisition de donnees multimodales dynamiques (son, images echographiques, capteurs de mouvement electromagnetiques, . . . ) chez un locuteur. Des proce- des de recalage, de synchronisation et de segmentation de ces donnees sont decrits pour bâtir des modeles du conduit vocal associes a une variabilite statistique. Le deuxieme s'interesse a la neuroradiologie interventionnelle, therapie guidee par l'image radiologique. En vue d'ameliorer l'experience visuelle du praticien, un premier ensemble de travaux traite de l'integration de l'angiographie rotationnelle 3D dans la salle d'operation d'une part en la superposant aux images per-operatoires, et d'autre part en y replacant le micro-outil chirurgical reconstruit en temps interactif...
Research Interests: Computer Science and Art
The radiotherapic treatment of cerebral arterious malformations (AVM) requires an accurate estimation of the AVM shape. This estimation is classically obtained from the delineation of the AVM in several 2D angiographic views. In this... more
The radiotherapic treatment of cerebral arterious malformations (AVM) requires an accurate estimation of the AVM shape. This estimation is classically obtained from the delineation of the AVM in several 2D angiographic views. In this paper, a clinical study of the inter-observer variability in the AVM detection is first performed. It proves that the estimated volume varies a lot between observers. For these reasons, we propose a framework for AVM delineation which makes use of 2D and 3D angiographic images: the initial estimate obtained with 2D angiographic images is then refined within the 3D volume using deformable models. Results are presented demonstrating shape delineation on various AVMs.
Research Interests:
The restriction of CBCT to micro-coil imaging allows for integrating key a priori knowledge that a coil is a high-density spatially sparse curvilinear structure. In this paper, we investigate acquisition patterns specifically designed for... more
The restriction of CBCT to micro-coil imaging allows for integrating key a priori knowledge that a coil is a high-density spatially sparse curvilinear structure. In this paper, we investigate acquisition patterns specifically designed for biplane systems allowing a faster workflow and reduced dose. Each pattern is a subsampling of a standard tomographic acquisition reconstructed with an 1-constrained algorithm to promote sparsity together with diffusion filters that promote the curvilinear nature of the coil. Three tensor-based 3D diffusion filters are investigated. Quantitative and qualitative results are provided for one coil and four patients datasets. They show how the reconstruction performs according to the selected acquisition pattern (uniform versus non-uniform subsampling), the quantity of missing information and the selected diffusion filter. We observed a systematically better recovery of the coil in reconstructions obtained using a uniform subsampling pattern but at the ...
Research Interests:
The radiotherapic treatment of cerebral arterious malformations (AVM) requires an accurate estimation of the AVM shape. This estimation is classically obtained from the delineation of the AVM in several 2D angiographic views. In this... more
The radiotherapic treatment of cerebral arterious malformations (AVM) requires an accurate estimation of the AVM shape. This estimation is classically obtained from the delineation of the AVM in several 2D angiographic views. In this paper, a clinical study of the inter-observer variability in the AVM detection is first performed. It proves that the estimated volume varies a lot between observers. For these reasons, we propose a framework for AVM delineation which makes use of 2D and 3D angiographic images: the initial estimate obtained with 2D angiographic images is then refined within the 3D volume using deformable models. Results are presented demonstrating shape delineation on various AVMs.
A method of calibrating an imaging system, wherein calibrating the intrinsic parameters for the different types of images can be taken in predetermined angular positions of the transmitting means and the receiving means, the intrinsic... more
A method of calibrating an imaging system, wherein calibrating the intrinsic parameters for the different types of images can be taken in predetermined angular positions of the transmitting means and the receiving means, the intrinsic parameters being estimated using a two-dimensional ghost to establish a distortion field for a particular angle, the distortion field for any angulation being calculated by polynomial interpolation of the distortion fields established for each control point, a control point being defined by a particular angulation axis of rotation of the imaging system to be calibrated.
Research Interests:
During an interventional neuroradiology exam, knowing the exact location of the catheter tip with respect to the patient can dramatically help the physician. An image registration between digital subtracted angiography (DSA) images and a... more
During an interventional neuroradiology exam, knowing the exact location of the catheter tip with respect to the patient can dramatically help the physician. An image registration between digital subtracted angiography (DSA) images and a volumic pre-operative image (magnetic resonance or computed tomography volumes) is a way to infer this important information. This mono-patient multimodality matching can be reduced to finding the projection matrix that transforms any voxel of the volume onto the DSA image plane. This modelization is unfortunately not valid in the case of distorted images, which is the case for DSA images. A classical angiography room can now generate 3D X-ray angiography volumes (3DXA). Since the DSA images are obtained with the same machine, it should be possible to deduce the projection matrix from the sensor data indicating the current machine position. We propose an interpolation scheme, associated to a pre-operative calibration of the machine that allows us to...
Research Interests:
This paper describes a new method for coupling ultrasound im - ages with tree-dimensional electromagnetic data, to recov er larger parts of the tongue during speech production. The electromagnetic datais superimposed on ultrasound images... more
This paper describes a new method for coupling ultrasound im - ages with tree-dimensional electromagnetic data, to recov er larger parts of the tongue during speech production. The electromagnetic datais superimposed on ultrasound images after spatial and temporal calibratio n. Successful fusion results are presented on various speech sequences. A comple te setup for evalu- ation of the electromagnetic system is further presented.
Research Interests:
Research Interests:
Research Interests:
Research Interests: Acoustics, Magnetic Resonance Imaging, Speech Acoustics, Phonation, Multidisciplinary, and 13 moreHumans, Automation, Transducers, Ultrasonography, Calibration, Three Dimensional Imaging, Multimodal imaging, Voice Quality, Electromagnetic phenomena, Tongue, Time Factors, Biomechanical Phenomena, and Larynx
Résumé : Ce chapitre décrit le système d'acquisition multimodale de données articulatoires élaboré dans le cadre du projet européen ASPI. Ce système intègre un échographe pour obtenir les contours de la langue, un dispositif de... more
Résumé : Ce chapitre décrit le système d'acquisition multimodale de données articulatoires élaboré dans le cadre du projet européen ASPI. Ce système intègre un échographe pour obtenir les contours de la langue, un dispositif de localisation électromagnétique pour localiser la ...
Coil embolization offers a new approach to treat aneurysms. This medical procedure is namely less invasive than an open-surgery as it relies on the deployment of very thin platinum-based wires within the aneurysm through the arteries.... more
Coil embolization offers a new approach to treat aneurysms. This medical procedure is namely less invasive than an open-surgery as it relies on the deployment of very thin platinum-based wires within the aneurysm through the arteries. When performed intracranially, this procedure must be particularly accurate and therefore carefully planned and performed by experienced radiologists. A simulator of the coil deployment represents an interesting and helpful tool for the physician by providing information on the coil behavior. In this paper, an original modeling is proposed to obtain interactive and accurate simulations of coil deployment. The model takes into account geometric nonlinearities and uses a shape memory formulation to describe its complex geometry. An experimental validation is performed in a contact-free environment to identify the mechanical properties of the coil and to quantitatively compare the simulation with real data. Computational performances are also measured to ...
Research Interests: Computer Aided Design, Humans, Computer Simulation, Three Dimensional Imaging, Computer User Interface Design, and 8 moreShape Memory, Experimental Validation, Interactive Simulation, Equipment Design, Prosthesis Implantation, Equipment Failure Analysis, Mechanical Property, and Intracranial aneurysm
... 22 A Corpus ASPI 25 A.1 Combescure sentences . . . . . ... Indeed some speakers have recorded articulatory data before the ASPI project. 1 Page 7. ... The main issue with MRI is to design a protocol that does not disturb articulation... more
... 22 A Corpus ASPI 25 A.1 Combescure sentences . . . . . ... Indeed some speakers have recorded articulatory data before the ASPI project. 1 Page 7. ... The main issue with MRI is to design a protocol that does not disturb articulation of speech sounds. ...
Research Interests:
Research Interests: Tomography and Kinematics
Research Interests:
This paper reports work in progress on X-ray angiography acquired under stereotactic conditions. The objective is to be able to match multimodality images (typically MRI and X-ray) without a stereotactic frame but with stereotactic... more
This paper reports work in progress on X-ray angiography acquired under stereotactic conditions. The objective is to be able to match multimodality images (typically MRI and X-ray) without a stereotactic frame but with stereotactic precision. We have identified potential problems and have studied them in detail. We conclude that, although the overall application is feasible, much work remains to be done on the estimation of the X-ray system conic projection and on automatic matching based on vascular structures.