Cybèle Ciofolo-Veit | Université de Rennes (original) (raw)

Uploads

Papers by Cybèle Ciofolo-Veit

Fetal, Infant and Ophthalmic Medical Image Analysis

3D ultrasound (US) is a promising technique to perform automatic extraction of standard planes fo... more 3D ultrasound (US) is a promising technique to perform automatic extraction of standard planes for fetal anatomy assessment. This requires prior organ localization, which is difficult to obtain with direct learning approaches because of the high variability in fetus size and orientation in US volumes. In this paper, we propose a methodology to overcome this spatial variability issue by scaling and automatically aligning volumes in a common 3D reference coordinate system. This preprocessing allows the organ detection algorithm to learn features that only encodes the anatomical variability while discarding the fetus pose. All steps of the approach are evaluated on 126 manually annotated volumes, with an overall mean localization error of 11.9 mm, showing the feasibility of multi-organ detection in 3D fetal US with machine learning.

Diagnostic and Interventional Imaging

Medical Imaging 2018: Image Processing

Ultrasound is increasingly becoming a 3D modality. Mechanical and matrix array transducers are ab... more Ultrasound is increasingly becoming a 3D modality. Mechanical and matrix array transducers are able to deliver 3D images with good spatial and temporal resolution. The 3D imaging facilitates the application of automated image analysis to enhance workflows, which has the potential to make ultrasound a less operator dependent modality. However, the analysis of the more complex 3D images and definition of all examination standards on 2D images pose barriers to the use of 3D in daily clinical practice. In this paper, we address a part of the canonical fetal screening program, namely the localization of the abdominal cross-sectional plane with the corresponding measurement of the abdominal circumference in this plane. For this purpose, a fully automated pipeline has been designed starting with a random forest based anatomical landmark detection. A feature trained shape model of the fetal torso including inner organs with the abdominal cross-sectional plane encoded into the model is then transformed into the patient space using the landmark localizations. In a free-form deformation step, the model is individualized to the image, using a torso probability map generated by a convolutional neural network as an additional feature image. After adaptation, the abdominal plane and the abdominal torso contour in that plane are directly obtained. This allows the measurement of the abdominal circumference as well as the rendering of the plane for visual assessment. The method has been trained on 126 and evaluated on 42 abdominal 3D US datasets. An average plane offset error of 5.8 mm and an average relative circumference error of 4.9 % in the evaluation set could be achieved.

… Image Computing and …, 2008

A typical Cardiac Magnetic Resonance (CMR) examination includes acquisition of a sequence of shor... more A typical Cardiac Magnetic Resonance (CMR) examination includes acquisition of a sequence of short-axis (SA) and long-axis (LA) images covering the cardiac cycle. Quantitative analysis of the heart function requires segmentation of the left ventricle (LV) SA images, while segmented LA views allow more accurate estimation of the basal slice and can be used for slice registration. Since manual segmentation of CMR images is very tedious and time-consuming, its automation is highly required. In this paper, we propose a fully automatic 2D method for segmenting LV consecutively in LA and SA images. The approach was validated on 35 patients giving mean segmentation error smaller than one pixel, both for LA and SA, and accurate LV volume measurements.

In the context of cardiac viability assessment, we propose a new fully automatic method to segmen... more In the context of cardiac viability assessment, we propose a new fully automatic method to segment and quantify myocardial pathological tissues in Late Enhancement Cardiac Magnetic Resonance images. Our two main contributions are a generic image intensity analysis and an original variational segmentation method, the Fast Region Competition. The obtained results are robust to anatomical variability and partial volume effects and false positives are avoided. To validate our results, we use representations that are independent of myocardium shape and size and compute clinically relevant indicators. The proposed method was tested on 100 slices and compared to other classical segmentation approaches, showing the best agreement with semi-automatic expert delineations.

Lecture Notes in Computer Science, 2004

We propose to segment volumetric structures with an atlasbased level set method. The classical fo... more We propose to segment volumetric structures with an atlasbased level set method. The classical formulation of the level set evolution force presents a stopping criterion, a directional term and a regularization term. Fuzzy labels registered from an atlas provide useful information allowing to automatically tune the respective influence of the different terms according to the desired application. This is done with a fuzzy decision system based on simple rules corresponding to an expert knowledge. Two applications are presented in details in the context of 3D brain MRI: the segmentation of white matter with the tuning of the regularization term, and the segmentation of the right hemisphere. Experimental results on the MNI Brainweb dataset and on a database of real MRI volumes are presented and discussed.

Nous nous proposons de segmenter des structures 3D avec des ensembles de niveau dont l'évolution ... more Nous nous proposons de segmenter des structures 3D avec des ensembles de niveau dont l'évolution est guidée par un modèle de forme et gérée par commande floue. Pour cela, plusieurs contours évoluent simultanément en direction de cibles anatomiques définies au préalable. Un système de décision floue combine l'information a priori fournie par un modèle de forme, utilisé comme atlas, avec la distribution d'intensité de l'image et les positions relatives des contours. Il donne en sortie le terme de direction de l'équation d'évolution de l'ensemble de niveau associé à chaque contour. Cela entraîne une expansion ou une contraction locale des contours, qui convergent finalement vers leurs cibles respectives. Le modèle de forme est construit par analyse en composantes principales. La forme moyenne et les variations obtenues permettent alors de localiser une cible et de déterminer les états flous caractérisant la distance du contour courant à cette cible. Cette méthode est appliquée à la segmentation des noyaux gris du cerveau et évaluée quantitativement sur une base de 18 sujets.

We propose a new method to segment 3D structures with competitive level sets driven by a shape mo... more We propose a new method to segment 3D structures with competitive level sets driven by a shape model and fuzzy control. To this end, several contours evolve simultaneously toward previously defined targets. The main contribution of this paper is the original introduction of prior information provided by a shape model, which is used as an anatomical atlas, into a fuzzy decision system. The shape information is combined with the intensity distribution of the image and the relative position of the contours. This combination automatically determines the directional term of the evolution equation of each level set. This leads to a local expansion or contraction of the contours, in order to match the borders of their respective targets. The shape model is produced with a principal component analysis, and the resulting mean shape and variations are used to estimate the target location and the fuzzy states corresponding to the distance between the current contour and the target. By combining shape analysis and fuzzy control, we take advantage of both approaches to improve the level set segmentation process with prior information. Experiments are shown for the 3D segmentation of deep brain structures from MRI and a quantitative evaluation is performed on a 18 volumes dataset.

2004 2nd IEEE International Symposium on Biomedical Imaging: Macro to Nano (IEEE Cat No. 04EX821), 2004

We propose to segment volumetric brain structures with a level set method including a fuzzy decis... more We propose to segment volumetric brain structures with a level set method including a fuzzy decision in the de-sign of the evolution force. The role of fuzzy logic is to fuse gradient-based and region-based information into a sin-gle force term, to take advantage of their ...

Medical Image Analysis, 2009

We propose a novel approach for the simultaneous segmentation of multiple structures with competi... more We propose a novel approach for the simultaneous segmentation of multiple structures with competitive level sets driven by fuzzy control. To this end, several contours evolve simultaneously toward previously defined anatomical targets. A fuzzy decision system combines the a priori knowledge provided by an anatomical atlas with the intensity distribution of the image and the relative position of the contours. This combination automatically determines the directional term of the evolution equation of each level set. This leads to a local expansion or contraction of the contours, in order to match the boundaries of their respective targets. Two applications are presented: the segmentation of the brain hemispheres and the cerebellum, and the segmentation of deep internal structures. Experimental results on real magnetic resonance (MR) images are presented, quantitatively assessed and discussed.

Lecture Notes in Computer Science, 2005

We propose to segment 3D structures with competitive level sets driven by fuzzy control. To this ... more We propose to segment 3D structures with competitive level sets driven by fuzzy control. To this end, several contours evolve simultaneously toward previously defined anatomical targets. A fuzzy decision system combines the a priori knowledge provided by an anatomical atlas with the intensity distribution of the image and the relative position of the contours. This combination automatically determines the directional term of the evolution equation of each level set. This leads to a local expansion or contraction of the contours, in order to match the borders of their respective targets. Two applications are presented: the segmentation of the brain hemispheres and the cerebellum, and the segmentation of deep internal structures. Experimental results on real MR images are presented, quantitatively assessed and discussed.

Page 1. ccsd-00001161 (version 1) : 19 Feb 2004 Proceeding of the Third International Alloy Confe... more Page 1. ccsd-00001161 (version 1) : 19 Feb 2004 Proceeding of the Third International Alloy Conference, Lisbon 2002 MONTE CARLO STUDY OF THE PRECIPITATION KINETICS OF Al3Zr IN Al-Zr Emmanuel Clouet1,2 and Maylise Nastar2 ...

2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, 2008

We propose a novel automatic method to segment the myocardium on late-enhancement cardiac MR (LE ... more We propose a novel automatic method to segment the myocardium on late-enhancement cardiac MR (LE CMR) images with a multi-step approach. First, in each slice of the LE CMR volume, a geometrical template is deformed so that its borders fit the myocardial contours. The second step consists in introducing a shape prior of the left ventricle. To do so, we use the cine MR sequence that is acquired along with the LE CMR volume. As the myocardial contours can be more easily automatically obtained on this data, they are used to build a 3D mesh representing the left ventricle geometry and the underlying myocardium thickness. This mesh is registered towards the contours obtained with the geometrical template, then locally adjusted to guarantee that scars are included inside the final segmentation. The quantitative evaluation on 27 volumes (272 slices) shows robust and accurate results.