AQUILAB

La Madeleine, France
La Madeleine, France

Time filter

Source Type

Dolz J.,AQUILAB | Dolz J.,Lille University Hospital Center | Massoptier L.,AQUILAB | Vermandel M.,Lille University Hospital Center
IRBM | Year: 2015

This work covers the current state of the art with regard to approaches to segment subcortical brain structures. A huge range of diverse methods have been presented in the literature during the last decade to segment not only one or a constrained number of structures, but also a complete set of these subcortical regions. Special attention has been paid to atlas based segmentation methods, statistical models and deformable models for this purpose. More recently, the introduction of machine learning techniques, such as artificial neural networks or support vector machines, has helped the researchers to optimize the classification problem. These methods are presented in this work, and their advantages and drawbacks are further discussed. Although these methods have proved to perform well, their use is often limited to those situations where either there are no lesions in the brain or the presence of lesions does not highly vary the brain anatomy. Consequently, the development of segmentation algorithms that can deal with such lesions in the brain and still provide a good performance when segmenting subcortical structures is highly required in practice by some clinical applications, such as radiotherapy or radiosurgery. © 2015 Elsevier Masson SAS.


Schaefer A.,Saarland University | Vermandel M.,Lille University Hospital Center | Baillet C.,Lille University Hospital Center | Dewalle-Vignion A.S.,Lille University Hospital Center | And 11 more authors.
European Journal of Nuclear Medicine and Molecular Imaging | Year: 2016

Purpose: The aim of this study was to evaluate the impact of consensus algorithms on segmentation results when applied to clinical PET images. In particular, whether the use of the majority vote or STAPLE algorithm could improve the accuracy and reproducibility of the segmentation provided by the combination of three semiautomatic segmentation algorithms was investigated. Methods: Three published segmentation methods (contrast-oriented, possibility theory and adaptive thresholding) and two consensus algorithms (majority vote and STAPLE) were implemented in a single software platform (Artiview®). Four clinical datasets including different locations (thorax, breast, abdomen) or pathologies (primary NSCLC tumours, metastasis, lymphoma) were used to evaluate accuracy and reproducibility of the consensus approach in comparison with pathology as the ground truth or CT as a ground truth surrogate. Results: Variability in the performance of the individual segmentation algorithms for lesions of different tumour entities reflected the variability in PET images in terms of resolution, contrast and noise. Independent of location and pathology of the lesion, however, the consensus method resulted in improved accuracy in volume segmentation compared with the worst-performing individual method in the majority of cases and was close to the best-performing method in many cases. In addition, the implementation revealed high reproducibility in the segmentation results with small changes in the respective starting conditions. There were no significant differences in the results with the STAPLE algorithm and the majority vote algorithm. Conclusion: This study showed that combining different PET segmentation methods by the use of a consensus algorithm offers robustness against the variable performance of individual segmentation methods and this approach would therefore be useful in radiation oncology. It might also be relevant for other scenarios such as the merging of expert recommendations in clinical routine and trials or the multiobserver generation of contours for standardization of automatic contouring. © 2015, Springer-Verlag Berlin Heidelberg.


PubMed | Center Henri Becquerel and 4108, AQUILAB, Albert Ludwigs University of Freiburg, Saarland University and Lille University Hospital Center
Type: Journal Article | Journal: European journal of nuclear medicine and molecular imaging | Year: 2016

The aim of this study was to evaluate the impact of consensus algorithms on segmentation results when applied to clinical PET images. In particular, whether the use of the majority vote or STAPLE algorithm could improve the accuracy and reproducibility of the segmentation provided by the combination of three semiautomatic segmentation algorithms was investigated.Three published segmentation methods (contrast-oriented, possibility theory and adaptive thresholding) and two consensus algorithms (majority vote and STAPLE) were implemented in a single software platform (Artiview). Four clinical datasets including different locations (thorax, breast, abdomen) or pathologies (primary NSCLC tumours, metastasis, lymphoma) were used to evaluate accuracy and reproducibility of the consensus approach in comparison with pathology as the ground truth or CT as a ground truth surrogate.Variability in the performance of the individual segmentation algorithms for lesions of different tumour entities reflected the variability in PET images in terms of resolution, contrast and noise. Independent of location and pathology of the lesion, however, the consensus method resulted in improved accuracy in volume segmentation compared with the worst-performing individual method in the majority of cases and was close to the best-performing method in many cases. In addition, the implementation revealed high reproducibility in the segmentation results with small changes in the respective starting conditions. There were no significant differences in the results with the STAPLE algorithm and the majority vote algorithm.This study showed that combining different PET segmentation methods by the use of a consensus algorithm offers robustness against the variable performance of individual segmentation methods and this approach would therefore be useful in radiation oncology. It might also be relevant for other scenarios such as the merging of expert recommendations in clinical routine and trials or the multiobserver generation of contours for standardization of automatic contouring.


PubMed | AQUILAB and Lille University Hospital Center
Type: | Journal: Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society | Year: 2016

Delineation of organs at risk (OARs) is a crucial step in surgical and treatment planning in brain cancer, where precise OARs volume delineation is required. However, this task is still often manually performed, which is time-consuming and prone to observer variability. To tackle these issues a deep learning approach based on stacking denoising auto-encoders has been proposed to segment the brainstem on magnetic resonance images in brain cancer context. Additionally to classical features used in machine learning to segment brain structures, two new features are suggested. Four experts participated in this study by segmenting the brainstem on 9 patients who underwent radiosurgery. Analysis of variance on shape and volume similarity metrics indicated that there were significant differences (p<0.05) between the groups of manual annotations and automatic segmentations. Experimental evaluation also showed an overlapping higher than 90% with respect to the ground truth. These results are comparable, and often higher, to those of the state of the art segmentation methods but with a considerably reduction of the segmentation time.


PubMed | AQUILAB, Lille 2 University of Health and Law, Albert Ludwigs University of Freiburg and Lille University Hospital Center
Type: Journal Article | Journal: Medical physics | Year: 2016

Accurate delineation of organs at risk (OARs) on computed tomography (CT) image is required for radiation treatment planning (RTP). Manual delineation of OARs being time consuming and prone to high interobserver variability, many (semi-) automatic methods have been proposed. However, most of them are specific to a particular OAR. Here, an interactive computer-assisted system able to segment various OARs required for thoracic radiation therapy is introduced.Segmentation information (foreground and background seeds) is interactively added by the user in any of the three main orthogonal views of the CT volume and is subsequently propagated within the whole volume. The proposed method is based on the combination of watershed transformation and graph-cuts algorithm, which is used as a powerful optimization technique to minimize the energy function. The OARs considered for thoracic radiation therapy are the lungs, spinal cord, trachea, proximal bronchus tree, heart, and esophagus. The method was evaluated on multivendor CT datasets of 30 patients. Two radiation oncologists participated in the study and manual delineations from the original RTP were used as ground truth for evaluation.Delineation of the OARs obtained with the minimally interactive approach was approved to be usable for RTP in nearly 90% of the cases, excluding the esophagus, which segmentation was mostly rejected, thus leading to a gain of time ranging from 50% to 80% in RTP. Considering exclusively accepted cases, overall OARs, a Dice similarity coefficient higher than 0.7 and a Hausdorff distance below 10 mm with respect to the ground truth were achieved. In addition, the interobserver analysis did not highlight any statistically significant difference, at the exception of the segmentation of the heart, in terms of Hausdorff distance and volume difference.An interactive, accurate, fast, and easy-to-use computer-assisted system able to segment various OARs required for thoracic radiation therapy has been presented and clinically evaluated. The introduction of the proposed system in clinical routine may offer valuable new option to radiation oncologists in performing RTP.


Servagi Vernat S.,Besancon University Hospital Center | Servagi Vernat S.,University of Franche Comte | Ali D.,University of Paris Descartes | Puyraveau M.,Besancon University Hospital Center | And 6 more authors.
Physica Medica | Year: 2014

Background: Intensity Modulated Arc Therapy (IMAT) can be planned and delivered via several techniques. Advanced Radiotherapy (ARTORL) is a prospective study that aims to evaluate the treatment costs and clinical aspects of implementing these IMAT techniques for head and neck cancers. In this context, we evaluated the potential dosimetric gain of Helical Tomotherapy (TomoTherapy, Accuray, HT) versus VMAT (Rapid'Arc®, Varian Medical System, RA) for oropharyngeal cancer (OC). Material and methods: Thirty patients were selected from our database in whom bilateral neck irradiation and treatment to the primary were indicated. Each patient was planned twice using both HT and RA planning systems using a simultaneous integrated boost approach. For the planning target volumes (PTV) and organs at risk, ICRU 83 reporting guidelines were followed. RA and HT plans were compared using paired Student's t-test. Results: RA and HT produced plans with a good coverage of PTVs and acceptable sparing of OARs. Although some dosimetric differences were statistically significant, they remained small. However, the near maximal dose to the PRV of spinal cord and brain stem was lower with HT. Regarding normal tissue, HT increased the volume irradiated at doses between 4 and 20Gy compared to RA. Conclusion: In OC, HT and RA showed similar dosimetric results. They represent the maximum gains obtained with photon beams. The medicoeconomic evaluation of our study is ongoing and may reveal differences between these techniques in terms of MU number, fraction time, and clinical evaluation. © 2013 Associazione Italiana di Fisica Medica.


Grant
Agency: European Commission | Branch: FP7 | Program: MC-ITN | Phase: FP7-PEOPLE-2011-ITN | Award Amount: 3.45M | Year: 2012

The greatest challenge for radiation therapy is to reach the highest probability of cure with the least morbidity. In practice, some difficulties remain to identify cancer cells, target them with radiation and minimize collateral damage. Over the last decades, remarkable progress has been made thanks to modern advances in computer and imaging technologies. Currently, the radiotherapy has reached a point where, besides 3D tumour morphology, time variations and biological variability within the tumour can also be taken into account. The SUMMER project is devised to produce unique software using several imaging sources (CT, MRI, PET, MR spectroscopy, fMRI, 4D PET-CT) for biological target volume delineation, based on spatial co-registration of multimodal morphologic and functional images. Furthermore, it will make additional biological information concerning tumour extension and tumour response available to radiotherapy, essential for patient treatment follow-up. Finally, radiation area will be more focused leading to less side-effect for the patient. Radiation oncology is now more dependent on medical imaging than it has ever been - and that dependence is only going to become greater. Therefore, convergence and collaboration of radiation oncology, nuclear medicine, diagnostic imaging and also computer science is the underlying driver to integrate efficiently and cost-effectively all information coming from various imaging technologies into the radiation therapy workflow. The design challenge is to combine the different level and kind of information into one interface, while currently doctors need to mentally do this operation. SUMMER will contribute to renew and strengthen this relationship through cross-disciplinary research, common workshops, and collaboration on training and education.


PubMed | Technical University of Delft, Erasmus University Rotterdam, Aquilab and Albert Ludwigs University of Freiburg
Type: Journal Article | Journal: Journal of digital imaging | Year: 2016

Accurate segmentation of organs at risk is an important step in radiotherapy planning. Manual segmentation being a tedious procedure and prone to inter- and intra-observer variability, there is a growing interest in automated segmentation methods. However, automatic methods frequently fail to provide satisfactory result, and post-processing corrections are often needed. Semi-automatic segmentation methods are designed to overcome these problems by combining physicians expertise and computers potential. This study evaluates two semi-automatic segmentation methods with different types of user interactions, named the strokes and the contour, to provide insights into the role and impact of human-computer interaction. Two physicians participated in the experiment. In total, 42 case studies were carried out on five different types of organs at risk. For each case study, both the human-computer interaction process and quality of the segmentation results were measured subjectively and objectively. Furthermore, different measures of the process and the results were correlated. A total of 36 quantifiable and ten non-quantifiable correlations were identified for each type of interaction. Among those pairs of measures, 20 of the contour method and 22 of the strokes method were strongly or moderately correlated, either directly or inversely. Based on those correlated measures, it is concluded that: (1) in the design of semi-automatic segmentation methods, user interactions need to be less cognitively challenging; (2) based on the observed workflows and preferences of physicians, there is a need for flexibility in the interface design; (3) the correlated measures provide insights that can be used in improving user interaction design.


Dolz J.,AQUILAB | Kirisli H.A.,AQUILAB | Viard R.,AQUILAB | Massoptier L.,AQUILAB
Progress in Biomedical Optics and Imaging - Proceedings of SPIE | Year: 2014

Computer-aided segmentation of anatomical structures in medical images is a valuable tool for efficient radiation therapy planning (RTP). As delineation errors highly affect the radiation oncology treatment, it is crucial to delineate geometric structures accurately. In this paper, a semi-automatic segmentation approach for computed tomography (CT) images, based on watershed and graph-cuts methods, is presented. The watershed pre-segmentation groups small areas of similar intensities in homogeneous labels, which are subsequently used as input for the graph-cuts algorithm. This methodology does not require of prior knowledge of the structure to be segmented; even so, it performs well with complex shapes and low intensity. The presented method also allows the user to add foreground and background strokes in any of the three standard orthogonal views - axial, sagittal or coronal - making the interaction with the algorithm easy and fast. Hence, the segmentation information is propagated within the whole volume, providing a spatially coherent result. The proposed algorithm has been evaluated using 9 CT volumes, by comparing its segmentation performance over several organs - lungs, liver, spleen, heart and aorta - to those of manual delineation from experts. A Dicés coefficient higher than 0.89 was achieved in every case. That demonstrates that the proposed approach works well for all the anatomical structures analyzed. Due to the quality of the results, the introduction of the proposed approach in the RTP process will be a helpful tool for organs at risk (OARs) segmentation. © 2014 SPIE.


Dolz J.,AQUILAB | Kirisli H.A.,AQUILAB | Viard R.,AQUILAB | Massoptier L.,AQUILAB
Progress in Biomedical Optics and Imaging - Proceedings of SPIE | Year: 2014

Accurate delineation of organs at risk (OAR) is required for radiation treatment planning (RTP). However, it is a very time consuming and tedious task. The use in clinic of image guided radiation therapy (IGRT) becomes more and more popular, thus increasing the need of (semi-)automatic methods for delineation of the OAR. In this work, an interactive segmentation approach to delineate OAR is proposed and validated. The method is based on the combination of watershed transformation, which groups small areas of similar intensities in homogeneous labels, and graph cuts approach, which uses these labels to create the graph. Segmentation information can be added in any view - axial, sagittal or coronal -, making the interaction with the algorithm easy and fast. Subsequently, this information is propagated within the whole volume, providing a spatially coherent result. Manual delineations made by experts of 6 OAR - lungs, kidneys, liver, spleen, heart and aorta - over a set of 9 computed tomography (CT) scans were used as reference standard to validate the proposed approach. With a maximum of 4 interactions, a Dice similarity coefficient (DSC) higher than 0.87 was obtained, which demonstrates that, with the proposed segmentation approach, only few interactions are required to achieve similar results as the ones obtained manually. The integration of this method in the RTP process may save a considerable amount of time, and reduce the annotation complexity. © 2014 SPIE.

Loading AQUILAB collaborators
Loading AQUILAB collaborators