Christian Vasseur

A computer-assisted system for 3-D frameless localization in stereotaxic MRI

A low-cost PC-based system for 3-D localization of brain targets in stereotaxic imaging is presented. It relies on a method, using MR images, in which four markers are inserted in the fastenings of a Talairach stereotaxic frame during MRI examination. By locating these markers on the images with this system, the transformation matrixes can be computed to obtain the 3-D coordinates of the center of a tumour in the stereotaxic space or in the MRI space. The system calculates the frame and arc setting parameters of a probe trajectory to the target, either for an orthogonal or a double oblique approach if needed. Simulated probe trajectory intersections with consecutive slices can be viewed in order to validate the trajectory before and during the surgical procedure. The method presents no major constraints in routine examinations. Mathematical details on the calculation of the transformation matrices are given.

An Approximate Solution to Normal Mixture Identification with Application to Unsupervised Pattern Classification

In this paper, an approach to unsupervised pattern classifiation is discussed. The classification scheme is based on an approximation of the probability densities of each class under the assumption that the input patterns are of a normal mixture. The proposed technique for identifying the mixture does not require prior information. The description of the mixture in terms of convexity allows to determine, from a totally unlabeled set of samples, the number of components and, for each of them, approximate values of the mean vector, the covariance matrix, and the a priori probability. Discriminant functions can then be constructed. Computer simulations show that the procedure yields decision rules whose performances remain close to the optimum Bayes minimum error-rate, while involving only a small amount of computation.

Treatment planning optimization by conjugate gradients and simulated annealing methods in stereotactic radiosurgery

PURPOSEnThis paper presents a new optimization method of treatment planning in linac stereotactic radiosurgery.nnnMETHODS AND MATERIALSnOn a workstation integrating x-rays, computed tomography (CT), magnetic resonance imaging (MRI), and digital subtracted angiography (DSA) images, we first determine the outlines of the target volume and surrounding healthy tissues to spare. To achieve complete optimization of the treatment plans, this method decomposes the optimization process in two steps. The position of the isocenters and the diameter of the collimators are first deduced by a conjugate gradients method, from the position and size of ellipsoids or spheres modeling the target volume. The other irradiation parameters, such as the isocenter dose, the aperture, and the weight of each irradiation plane and of their irradiation sectors are finally deduced by a simulated annealing optimization algorithm.nnnRESULTSnThe system can perform multitarget/multisector treatment plans that are automatically obtained in a satisfactory time (as a rule, 20 min for a two-target irradiation), much faster than the time needed for a manual treatment planning. We present the results in two cases: the simulation of a single-target treatment and a two-target real treatment with constraints. In these two cases, we can control the dose received by target and sensitive volumes.nnnCONCLUSIONnThis method achieves an excellent conformation of the estimated isodose curves with the outlines of the target volume, which allows us to avoid the surrounding healthy tissues, thanks to the different weighting factors given on each volume concerned according to the importance we grant to each of them.

A Fast Algorithm for Nonparametric Probability Density Estimation

A fast algorithm for the well-known Parzen window method to estimate density functions from the samples is described. The computational efforts required by the conventional and straightforward implementation of this estimation procedure limit its practical application to data of low dimensionality. The proposed algorithm makes the computation of the same density estimates with a substantial reduction of computer time possible, especially for data of high dimensionality. Some simulation experiments are presented which demonstrate the efficiency of the method. They indicate the computational savings that may be achieved through the use of this fast algorithm for artificially generated sets of data.

Volume delineation by fusion of fuzzy sets obtained from multiplanar tomographic images

Techniques of three-dimensional (3-D) volume delineation from tomographic medical imaging are usually based on 2-D contour definition. For a given structure, several different contours can be obtained depending on the segmentation method used or the users choice. The goal of this work is to develop a new method that reduces the inaccuracies generally observed. A minimum volume that is certain to be included in the volume concerned (membership degree /spl mu/=1), and a maximum volume outside which no part of the volume is expected to be found (membership degree /spl mu/=0), are defined semi-automatically. The intermediate fuzziness region (0</spl mu/<1) is processed using the theory of possibility. The resulting fuzzy volume is obtained after data fusion from multiplanar slices. The influence of the contrast-to-noise ratio was tested on simulated images. The influence of slice thickness as well as the accuracy of the method were studied on phantoms. The absolute volume error was less than 2% for phantom volumes of 2-8 cm/sup 3/, whereas the values obtained with conventional methods were much larger than the actual volumes. Clinical experiments were conducted, and the fuzzy logic method gave a volume lower than that obtained with the conventional method. Our fuzzy logic method allows volumes to be determined with better accuracy and reproducibility.

Filtering by optimal projection and application to automatic artifact removal from EEG

A new approach to filter multi-channel signals is presented, called filtering by optimal projection (FOP) in this paper. This approach is based on common spatial subspace decomposition (CSSD) theory. Moreover, an evolution of this method for non-stationary signals is also introduced which is called adaptative FOP (AFOP). As ICA, a filtering matrix is set up in the best way to remove artifacts with linear combination of channels. This filtering matrix is characterized by two subspaces. The first one is determined during a learning phase, by finding components maximizing the ratio signal over noise. The second one will be determined during a filtering phase, by reconstructing signals of a sliding window, by a least square method. These methods are completely automated and enable to filter independently numerous artifact types. Moreover, this filtering can be improved by applying this process on frequency band decomposed signals. Various tests have been made on electroencephalogram (EEG) signals in order to remove ocular and muscular activity while conserving pathological activity (slow waves, paroxysms). The results are compared with ICA filtering and medical inspection has been carried out to prove that this approach yields very good performance.

Registration, Matching, and Data Fusion in 2D/3D Medical Imaging: Application to DSA and MRA

This paper deals with a new approach of registration in multimodal imaging. Modalities involved are Digital Subtracted Angiography (DSA, 2D) and Magnetic Resonance Angiography (MRA, 3D). Our approach is an hybrid one, mixing feature and intensity based approaches. This approach is based on the extraction of a anatomical referential common to both MRA and DSA. This step appears to be the “geometric-like” aspect. Then, a high level optimization scheme gives the best registration, using an iconic similarity measure. Several ways of matching planar and tomographic imaging are proposed through superimposition, point to point matching or 3D data fusion. The results obtained prove the methods efficiency in a clinical context.

Validation of a new method for stereotactic localization using MR imaging

Magnetic resonance is recognized as potentially the best imaging procedure for localization in stereotactic neurosurgery. However, special difficulties necessitate specific adaptation to localize targets in the stereotactic frame. We developed a new method for stereotactic localization. The MR studies were performed using a 0.5 T imager. Four small boxes filled with CuSO4 solution were inserted into the intracranial holders of a Talairach frame. Using fast sequences, thirty 7-mm thick contiguous sagittal slices and twenty 5-mm thick axial slices enabled us to image the entire brain. The image data were transferred for analysis to an image processing station, including special software to handle stereotactic calculations. The accuracy of the origin of the trihedron and systematic geometrical errors were carefully evaluated using a cubic phantom, and corrective algorithms were applied when needed. Moreover, checks have been designed to detect geometrical distortion due to ferromagnetic artifacts, alterations in gradient calibration, or movements made by the patient. This localization method does not necessitate the use of stereotactic frames and appears to be precise enough for clinical use. Duration of MR examination is not a restricting factor, mainly because the patient can be positioned easily.

Hybrid control for vision based Cart-Inverted Pendulum system

This paper presents a vision based cart-inverted pendulum (CIP) system under a hybrid feedback configuration: the continuous carts position measured by encoder and the delayed & sampled inverted pendulums upper coordinates, obtained from a visual sensor. The challenge here is to stabilize the CIP from a big inclined initial angle by using a low cost CCD camera. Under this scheme, we propose a hybrid control which consists in a jumping-up (bang-bang) control and a two causal stabilization loops control: the first one (inner loop) realizes a linearization and the stabilization control of the pendulum based on an innovative piecewise continuous reduced order Luenberger observer (PCROLO) coupled with a linearization module, the second one (the outer loop) realizes a Lyapunov based control for the unstable internal system with lower dynamics than that of the pendulum. This hybrid control method is capable of balancing the CIP system within small carts displacement. Performances issues of the proposed method are illustrated by the experimental figures and videos.

Piecewise Continuous Systems Used in Trajectory Tracking of a Vision Based X-Y Robot

This paper deals with a new approach of trajectory tracking for a vision based x-y robot which delivers a delayed and sampled output. A control theory of using a class of piecewise continuous systems named as piecewise continuous controllers and a specific observer enabling sampled tracking is developed. The experimental results show the effectiveness and robustness aspects of the method.