Serge Miguet

Simulation of four-dimensional CT images from deformable registration between inhale and exhale breath-hold CT scans.

PURPOSE We propose to simulate an artificial four-dimensional (4-D) CT image of the thorax during breathing. It is performed by deformable registration of two CT scans acquired at inhale and exhale breath-hold. MATERIALS AND METHODS Breath-hold images were acquired with the ABC (Active Breathing Coordinator) system. Dense deformable registrations were performed. The method was a minimization of the sum of squared differences (SSD) using an approximated second-order gradient. Gaussian and linear-elastic vector field regularizations were compared. A new preprocessing step, called a priori lung density modification (APLDM), was proposed to take into account lung density changes due to inspiration. It consisted of modulating the lung densities in one image according to the densities in the other, in order to make them comparable. Simulated 4-D images were then built by vector field interpolation and image resampling of the two initial CT images. A variation in the lung density was taken into account to generate intermediate artificial CT images. The Jacobian of the deformation was used to compute voxel values in Hounsfield units. The accuracy of the deformable registration was assessed by the spatial correspondence of anatomic landmarks located by experts. RESULTS APLDM produced statistically significantly better results than the reference method (registration without APLDM preprocessing). The mean (and standard deviation) of distances between automatically found landmark positions and landmarks set by experts were 2.7(1.1) mm with APLDM, and 6.3(3.8) mm without. Interexpert variability was 2.3(1.2) mm. The differences between Gaussian and linear elastic regularizations were not statistically significant. In the second experiment using 4-D images, the mean difference between automatic and manual landmark positions for intermediate CT images was 2.6(2.0) mm. CONCLUSION The generation of 4-D CT images by deformable registration of inhale and exhale CT images is feasible. This can lower the dose needed for 4-D CT acquisitions or can help to correct 4-D acquisition artifacts. The 4-D CT model can be used to propagate contours, to compute a 4-D dose map, or to simulate CT acquisitions with an irregular breathing signal. It could serve as a basis for 4-D radiation therapy planning. Further work is needed to make the simulation more realistic by taking into account hysteresis and more complex voxel trajectories.

A Real-Time, Multiview Fall Detection System: A LHMM-Based Approach

Automatic detection of a falling person in video sequences has interesting applications in video-surveillance and is an important part of future pervasive home monitoring systems. In this paper, we propose a multiview approach to achieve this goal, where motion is modeled using a layered hidden Markov model (LHMM). The posture classification is performed by a fusion unit, merging the decision provided by the independently processing cameras in a fuzzy logic context. In each view, the fall detection is optimized in a given plane by performing a metric image rectification, making it possible to extract simple and robust features, and being convenient for real-time purpose. A theoretical analysis of the chosen descriptor enables us to define the optimal camera placement for detecting people falling in unspecified situations, and we prove that two cameras are sufficient in practice. Regarding event detection, the LHMM offers a principle way for solving the inference problem. Moreover, the hierarchical architecture decouples the motion analysis into different temporal granularity levels, making the algorithm able to detect very sudden changes, and robust to low-level steps errors.

Patient setup error measurement using 3D intensity-based image registration techniques

PURPOSE Conformal radiotherapy requires accurate patient positioning with reference to the initial three-dimensional (3D) CT image. Patient setup is controlled by comparison with portal images acquired immediately before patient treatment. Several automatic methods have been proposed, generally based on segmentation procedures. However, portal images are of very low contrast, leading to segmentation inaccuracies. In this study, we propose an intensity-based (with no segmentation), fully automatic, 3D method, associating two portal images and a 3D CT scan to estimate patient setup. MATERIALS AND METHODS Images of an anthropomorphic phantom were used. A CT scan of the pelvic area was first acquired, then the phantom was installed in seven positions. The process is a 3D optimization of a similarity measure in the space of rigid transformations. To avoid time-consuming digitally reconstructed radiograph generation at each iteration, we used two-dimensional transformations and two sets of specific and pregenerated digitally reconstructed radiographs. We also propose a technique for computing intensity-based similarity measures between several couples of images. A correlation coefficient, chi-square, mutual information, and correlation ratio were used. RESULTS The best results were obtained with the correlation ratio. The median root mean square error was 2.0 mm for the seven positions tested and was, respectively, 3.6, 4.4, and 5.1 for correlation coefficient, chi-square, and mutual information. CONCLUSIONS Full 3D analysis of setup errors is feasible without any segmentation step. It is fast and accurate and could therefore be used before each treatment session. The method presents three main advantages for clinical implementation-it is fully automatic, applicable to all tumor sites, and requires no additional device.

Medical Images Simulation, Storage, and Processing on the European DataGrid Testbed

The European 1ST DataGrid project was a pioneer in identifying the medical imaging field as an application domain that can benefit from Grid technologies. This paper describes how and for which purposes medical imaging applications can be Grid-enabled. Applications that have been deployed on the DataGrid testbed and middleware are described. They relate to medical image manipulation, including image production, secured image storage, and image processing. Results show that Grid technologies are still in their youth to address all issues related to complex medical imaging applications. If the benefit of Grid enabling for some medical applications is clear, there remain opened research and technical issues to develop and integrate all necessary services.

Discrete Curvature Based on Osculating Circle Estimation

In this paper, we make an overview of the existing algorithms concerning the discrete curvature estimation. We extend the Worring and Smeulders [WS93] classification to new algorithms and we present a new and purely discrete algorithm based on discrete osculating circle estimation.

Real Time Foreground-Background Segmentation Using a Modified Codebook Model

Real time segmentation of scene into objects and background is really important and represents an initial step of object tracking. Starting from the codebook method [4] we propose some modifications which show significant improvements in most of the normal and also difficult conditions. We include parameter of frequency for accessing, deleting, matching and adding codewords in codebook or to move cache codewords into codebook. We also propose an evaluation method in order to objectively compare several segmentation techniques, based on receiver operating characteristic (ROC) analysis and on precision and recall method. We propose to summarize the quality factor of a method by a single value based on a weighted Euclidean distance or on a harmonic mean between two related characteristics.

A cognitive and video-based approach for multinational License Plate Recognition

License Plate Recognition (LPR) is mainly regarded as a solved problem. However, robust solutions able to face real-world scenarios still need to be proposed. Country-specific systems are mostly, designed, which can (artificially) reach high-level recognition rates. This option, however, strictly limits their applicability. In this paper, we propose an approach that can deal with various national plates. There are three main areas of novelty. First, the Optical Character Recognition (OCR) is managed by a hybrid strategy, combining statistical and structural algorithms. Secondly, an efficient probabilistic edit distance is proposed for providing an explicit video-based LPR. Last but not least, cognitive loops are introduced at critical stages of the algorithm. These feedback steps take advantage of the context modeling to increase the overall system performances, and overcome the inextricable parameter settings of the low-level processing. The system performances have been tested in more than 1200 static images with difficult illumination conditions and complex backgrounds, as well as in six different videos containing 525 moving vehicles. The evaluations prove our system to be very competitive among the non-country specific approaches.

2D and 3D visibility in discrete geometry: an application to discrete geodesic paths

In this article, we present a discrete definition of the classical visibility in computational geometry based on digital straight fines. We present efficient algorithms to compute the set of pixels in a non-convex domain that are visible from a source pixel. Based on these definitions, we define discrete geodesic paths in discrete domain with obstacles. This allows us to introduce a new geodesic metric in discrete geometry.

A robust appearance model for tracking human motions

We propose an original method for tracking people based on the construction of a 2-D human appearance model. The general framework, which is a region-based tracking approach, is applicable to any type of object. We show how to specialize the method for taking advantage of the structural properties of the human body. We segment its visible parts, construct and update the appearance model. This latter one provides a discriminative feature capturing both color and shape properties of the different limbs, making it possible to recognize people after they have temporarily disappeared. The method does not make use of skin color detection, which allows us to perform tracking under any viewpoint. The only assumption for the recognition is the approximate viewpoint correspondence during the matching process between the different models. Several results in complex situations prove the efficiency of the algorithm, which runs in near real time. Finally, the model provides an important clue for further human motion analysis process.

Scattering on a ring of processors

Abstract In this note, we prove that the complexity of scattering in an oriented ring of p processors is ( p - 1) (β + Lτ) where L is the length of the messages, β the communication startup, and τ the elemental propagation time.