Search

A method by which to register multimodality medical images accommodating soft tissue deformation is presented in the context of interventional therapy with a MR scanner. Accuracy testing with arbitrarily deformed MR images and application studies of a pig's brain were undertaken to evaluate the feasibility of the method. When Mutual Information is employed as the voxel similarity measure in the matching energy function, the algorithm can accommodate multimodality images. Coupled with rigid registration, the deformable registration of pre- and intraoperative multi-modality images enables surgeons to precisely define critical anatomical structures, such as vessels and functional areas, and to localize and optimize trajectories. The method directly and automatically works on volumetric multimodality images. Thus the algorithm is suitable for intraoperative registration, where stability and simplicity are desirable. [ABSTRACT FROM AUTHOR]

We present an approach which performs the automatic labeling of the main cortical sulci using a priori information for the 3D spatial distribution of these entities. We have developed a methodology to extract the 3D cortical topography of a particular subject from in vivo observations obtained through MRI. The cortical topography is encoded in a relational graph structure composed of two main features: arcs and vertices. Each vertex contains a parametric surface representing the buried part of a sulcus. Points on this parametric surface are expressed in stereotaxic coordinates (i.e., with respect to a standardized brain coordinate system). Arcs represent the connections between these entities. Manual sulcal labeling is performed by tagging a sulcal surface in the 3-D graph and selecting from a menu of candidate sulcus names. Automatic labeling is dependent on a probabilistic atlas of sulcal anatomy derived from a set of 51 graphs that were labeled by an anatomist. We show how these 3D sulcal spatial distribution maps can be used to perform the identification of the cortical sulci. We focus our attention on the peri-central area (including pre-central, post-central and central sulci). Results show that the use of spatial priors permit automatic identification of the main sulci with a good accuracy. [ABSTRACT FROM AUTHOR]

This paper describes the development of a needle insertion manipulator for stereotactic neurosurgery. This robot fulfils the requirements of having both a safe mechanical design and the capacity for being sterilized. Many kinds of robots are examined in neurosurgery. Their purpose is the precise positioning of surgical instruments such as biopsy needles, electrodes etc. Some are already available commercially and have been proven useful in the operating theatre. However, their clinical application is limited by specific problems including cost, safety, positioning requirements, maintenance requirements. The main problems have been with the safety of the mechanical design and difficulties with sterilization and disinfecting pre-and post operatively. The manipulator described in this report achieves mechanical safety and has the capacity for cover-sheet-free sterilization. The manipulator has three major components: the main mechanical component (with 6 degrees of freedom), the torque transmission component, and the electric motor, which cannot be sterilized. The electrical parts are detachable. Using this mechanism, we can clearly separate the surgical area from the mechatronics components. In this paper, the basic design and the prototype development and testing are described. [ABSTRACT FROM AUTHOR]

In this paper, we propose a robust fully non-supervised method dedicated to the segmentation of the brain in T1-weighted MR images. The first step consists in the analysis of the scale-space of the histogram first and second derivative. We show first that the crossings in scale-space of trajectories of extrema of different derivative orders follow regular topological properties. These properties allow us to design a new structural representation of a 1D signal. Then we propose an heuristics using this representation to infer statistics on grey and white matter grey level values from the histogram. These statistics are used by an improved morphological process combining two opening sizes to segment the brain. The method has been validated with 70 images coming from 3 different scanners and acquired with various MR sequences. [ABSTRACT FROM AUTHOR]

A new, practical, and efficient approach is proposed for 3D vascular segmentation and bifurcation structure extraction. The method uses a combination of mathematical morphology, region-growing schemes, and shape features in addition to greyscale information. By an extension of math-morphological operations within bounded space of vascular shape, smooth and natural region-growing and sensitivity-controllable bifurcation detection were realized. The algorithm was implemented in the interactive segmentation and visualization software package VOXEL-MAN and validated with clinical data of X-ray CT angiography and MRA. [ABSTRACT FROM AUTHOR]

A new method to segment MR volumes has been developed. The method matches elastically a 3D deformable prior model, describing the structures of interest, to the MR volume of a patient. The deformation is done using a deformation grid. Oriented distance maps are utilized to guide the deformation process. Two alternative restrictions are used to preserve the geometrical prior knowledge of the model. The method is applied to extract the body, the lungs and the heart. The segmentation is needed to build individualized boundary element models for bioelectromagnetic inverse problem. The method is fast, automatic and accurate. Good results have been achieved for four MR volumes tested so far. [ABSTRACT FROM AUTHOR]

This paper proposes a method for fully automatic segmentation of brain tissues and MR bias field correction using a digital brain atlas. We have extended the EM segmentation algorithm, including an explicit parametric model of the bias field. The algorithm interleaves classification with parameter estimation, yielding better results at every iteration. The method can handle multi-channel data and slice-per-slice constant offsets, and is fully automatic due to the use of a digital brain atlas. [ABSTRACT FROM AUTHOR]

We have used MRI-based three-dimensional (3D) reconstruction and a real-time, frameless, stereotactic navigation device to facilitate the removal of seizure foci in children suffering from intractable epilepsy. Using this system, the location of subdural grid and strip electrodes is recorded on the 3D model to facilitate focus localization and resection. Ten operations were performed — two girls and eight boys ranging in age from 3-17 — during which 3D reconstruction and surgical instrument tracking navigation was used. In all cases, the patients tolerated the procedure well and showed no post-operative neurological deficits. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of segmenting bone from Computed Tomography (CT) data. In clinical practice, identification of bone is done by thresholding, a method which is simple and fast. Unfortunately, thresholding alone has significant limitations. In particular, segmentation of thin bone structures and of joint spaces is problematic. This problem is particularly severe for thin bones such as in the skull (the paranasal sinus and around the orbit). Another area where current techniques often fail is automatic, reliable and robust identification of individual bones, which requires precise separation of the joint spaces. This paper presents a novel solution to these problems based on three-dimensional filtering techniques. Improvement of the segmentation results in more accurate 3D models for the purpose of surgical planning and intraoperative navigation. [ABSTRACT FROM AUTHOR]

This paper presents a method for automatic segmentation of the tibia and femur in clinical magnetic resonance images of knees. Texture information is incorporated into an active contours framework through the use of vector-valued geodesic snakes with local variance as a second value at each pixel, in addition to intensity. This additional information enables the system to better handle noise and the non-uniform intensities found within the structures to be segmented. It currently operates independently on 2D images (slices of a volumetric image) where the initial contour must be within the structure but not necessarily near the boundary. These separate segmentations are stacked to display the performance on the entire 3D structure. [ABSTRACT FROM AUTHOR]

The in vivo investigation of joint kinematics in normal and injured wrist requires the segmentation of carpal bones from 3D (CT) images and their registration over time. The non-uniformity of bone tissue, ranging from dense cortical bone to textured spongy bone, the irregular, small shape of closely packed carpal bones which move with respect to one another, and with respect to CT resolution, augmented with the presence of blood vessels, and the inherent blurring of CT imaging renders the segmentation of carpal bones a challenging task. Specifically, four characteristic difficulties are prominent: (i) gaps or weak edges in the carpal bone surfaces, (ii) diffused edges, (iii) textured regions, and, (iv) extremely narrow inter-bone regions. We review the performance of statistical classification, deformable models, region growing, and morphological operations for this application. We then propose a model which combines several of these approaches in a single framework. Specifically, initialized seeds grow in a curve evolution implementation of active contours, but where growth is modulated by a skeletally-mediated competition between neighboring regions, thus combining the advantages of local and global region growing methods, region competition and active contours. This approach effectively deals with many of the difficulties presented above as illustrated by numerous examples. [ABSTRACT FROM AUTHOR]

We present an approach to elastic registration of tomographic brain images which is based on thin-plate splines and takes into account landmark errors. The inclusion of error information is important in clinical applications since landmark extraction is always prone to error. In comparison to previous work, our approach can cope with anisotropic errors, which is significantly more realistic than dealing only with isotropic errors. In particular, it is now possible to include different types of landmarks, e.g., quasi-landmarks at the outer contour of the brain. Also, we introduce an approach to estimate landmark localization uncertainties directly from the image data. Experimental results are presented for the registration of 2D and 3D MR images. [ABSTRACT FROM AUTHOR]

This paper describes a new method of non-rigid registration using the combined power of elastic and statistical shape models. The transformations are constrained to be consistent with a physical model of elasticity to maintain smoothness and continuity. A Bayesian formulation, based on this model, on an intensity similarity measure, and on statistical shape information embedded in corresponding boundary points, is employed to find a more accurate and robust non-rigid registration. A dense set of forces arises from the intensity similarity measure to accommodate complex anatomical details. A sparse set of forces constrains consistency with statistical shape models derived from a training set. A number of experiments were performed on both synthetic and real medical images of the brain and heart to evaluate the approach. It is shown that statistical boundary shape information significantly augments and improves elastic model based non-rigid registration. [ABSTRACT FROM AUTHOR]

A comparison of six similarity measures for use in intensity based 2D-3D image registration is presented. The accuracy of the similarity measures are compared to a "gold-standard" registration which has been accurately calculated using fiducial markers. The similarity measures are used to register a CT scan to a fluoroscopy image of a spine phantom. The registration is carried out within a region of interest in the fluoroscopy image which is user defined to contain a single vertebra. Many of the problems involved in this type of registration are caused by features which were not modelled by a phantom image alone. More realistic "gold standard" data sets were simulated using the phantom image with clinical image features overlaid. Results show that the introduction of soft tissue structures and interventional instruments into the phantom image can have a large effect on the performance of some similarity measures previously applied to 2D-3D image registration. Two measures were able to register accurately and robustly even when soft tissue structures and interventional instruments were present as differences between the images. These measures are called pattern intensity and gradient difference. [ABSTRACT FROM AUTHOR]

We present a new approach for the non-rigid registration of contrast-enhanced breast MRI using normalised mutual information. A hierarchical transformation model of the motion of the breast has been developed: The global motion of the breast is modelled using affine transformation models while the local motion of the breast is modelled using spline-based free-form deformation (FFD) models. The algorithm has been applied to the fully automated registration of 3D breast MRI. In particular, we have compared the results of the proposed non-rigid registration algorithm to those obtained using rigid and affine registration techniques. The results clearly indicate that the non-rigid registration algorithm is much better able to recover the motion and deformation of the breast than rigid or affine registration algorithms. [ABSTRACT FROM AUTHOR]

We propose a new way of embedding shape distributions in a topological deformable template. These distributions rely on global shape descriptors corresponding to the 3D moment invariants. In opposition to usual Fourier-like descriptors, they can be updated during deformations at a relatively low cost. The moment-based distributions are included in a framework allowing the management of several simultaneously deforming objects. This framework is dedicated to the segmentation of brain deep nuclei in 3D MR images. The paper focuses on the learning of the shape distributions, on the initialization of the topological model and on the multi-resolution energy minimization process. Results are presented showing the segmentation of twelve brain deep structures. [ABSTRACT FROM AUTHOR]

A quick method to obtain the 3D transformation of a 3D free-form shape model from its 2D projection data is proposed. This method has been developed for the real-time registration of a 3D model of a cerebral vessel tree, obtained from pre-operative data (eg. MR Angiogram), to a X-ray image of the vessel (eg. Digital Subtraction Angiogram) taken during an operation. First, the skeleton of the vessel in a 2D image is automatically extracted in a model-based way using a 2D projection of a 3D model skeleton at the initial state (up to ±20 degree difference in rotation). Corresponding pairs of points on the 3D skeleton and points on the 2D skeleton are determined based on the 2D Euclidean distance between the projection of the model skeleton and the observed skeleton. In the process, an adaptive search region for each model point, which is determined according to the projected shape, effectively removes incorrect correspondences. Based on a good ratio of correct pairs, linearization of a rotation matrix can be used to rapidly calculate the 3D transformation of the model which produces the 2D observed projection. Experiments using real data show the practical usefulness of the method. [ABSTRACT FROM AUTHOR]

Over the last five years, new "voxel-based" approaches have allowed important progress in multimodal image registration, notably due to the increasing use of information-theoretic similarity measures. Their wide success has led to the progressive abandon of measures using standard image statistics (mean and variance). Until now, such measures have essentially been based on heuristics. In this paper, we address the determination of a new measure based on standard statistics from a theoretical point of view. We show that it naturally leads to a known concept of probability theory, the correlation ratio. In our derivation, we take as the hypothesis the functional dependence between the image intensities. Although such a hypothesis is not as general as possible, it enables us to model the image smoothness prior very easily. We also demonstrate results of multimodal rigid registration involving Magnetic Resonance (MR), Computed Tomography (CT), and Positron Emission Tomography (PET) images. These results suggest that the correlation ratio provides a good trade-off between accuracy and robustness. [ABSTRACT FROM AUTHOR]

We provide in this article a generic framework for pose estimation from geometric features. We propose more particularly two algorithms: a gradient descent on the Riemannian least squares distance and on the Mahalanobis distance. For each method, we provide a way to compute the uncertainty of the resulting transformation. The analysis and comparison of the algorithms show their advantages and drawbacks and point out the very good prediction on the transformation accuracy. An application in medical image analysis validates the uncertainty estimation on real data and demonstrates that, using adapted and rigorous tools, we can detect very small modifications in medical images. We believe that these algorithms could be easily embedded in many applications and provide a thorough basis for computing many image statistics. [ABSTRACT FROM AUTHOR]

A fully automated voxel-based algorithm for multimodal non-rigid image registration using a mutual information based similarity criterion is presented. The image deformation vector field is found by dividing images into overlapping neighborhood regions and translating these regions to locally increase the similarity criterion. The calculated displacement in each region is assigned to the centered pixel of each neighborhood region and propagated to its neighbors with a Gaussian window function assuring continuity and smoothness of the deformation field. Initial experiments on 2D test data qualitatively demonstrate the feasibility of this approach. [ABSTRACT FROM AUTHOR]

Peri-ictal SPECT is a potentially powerful tool to study changes in regional cerebral blood flow during a partial seizure and to localize the seizure focus for pre-surgical evaluation. Recently we have developed and validated a method for co-registering and normalizing 3D interictal SPECT images to 3D ictal SPECT images and deriving a difference image co-registered to the patient's 3D MRI. This method has been shown to significantly improve clinical outcomes in surgery patients. In order to improve the method's sensitivity and specificity a number of enhancements have been added to the basic technique, including using voxel-based registration to better align ictal and interictal images, dual isotope (Tc-99m and I-123) SPECT for effective tissue discrimination, and a custom software interface written to facilitate the entire process. We report results that show significant improvement over the original method, and we discuss future work which should continue to improve the method as a clinical tool. [ABSTRACT FROM AUTHOR]

In this paper we present a framework to simultaneously segment portal images and register them to 3D treatment planning CT data sets for the purpose of radiotherapy setup verification. Due to the low resolution and low contrast of the portal image, taken with a high energy treatment photon beam, registration to the 3D CT data is a difficult problem. However, if some structure can be segmented in the portal image, it can be used to help registration, and if there is an estimate of the registration parameters, it can help improve the segmention of the portal image. The minimax entropy algorithm proposed in this paper evaluates appropriate entropies in order to segment the portal image and to find the registration parameters iteratively. The proposed algorithm can be used, in general, for registering a high resolution image to a low resolution image. Finally, we show the proposed algorithm's relation to the mutual information [19] metric proposed in the literature for multi-modality image registration. [ABSTRACT FROM AUTHOR]

The registration of multimodal medical images is an important tool in surgical applications, since different scan modalities highlight complementary anatomical structures. We present a method of computing the best rigid registration of pairs of medical images of the same patient. The method uses prior information on the expected joint intensity distribution of the images when correctly aligned, given a priori registered training images. We discuss two methods of modeling the joint intensity distribution of the training data, mixture of Gaussians and Parzen windowing. The fitted Gaussians roughly correspond to various anatomical structures apparent in the images and provide a coarse anatomical segmentation of a registered image pair. Given a novel set of unregistered images, the algorithm computes the best registration by maximizing the log likelihood of the two images, given the transformation and the prior joint intensity model. Results aligning SPGR and dual-echo MR scans demonstrate that this algorithm is a fast registration method with a large region of convergence and sub-voxel registration accuracy. [ABSTRACT FROM AUTHOR]

Ultrasonic acquisition, by its simplicity, low cost and non invasiveness, emerged as an essential supplementary examination for cardiovascular diseases. The aim of this study is to demonstrate the feasibility of a global branching artery modelling. A geometrical approach is used for that purpose. A surface, based on a deformable skeleton and envelope, is defined. Including the shape as an a priori knowledge, this surface is fitted with 3D ultrasonic data by an iterative downhill gradient descent. Quality of the fit is illustrated by figures and numerical results. This tool could be used for objective spatial quantification and can be a useful instrument for blood flow modelling. [ABSTRACT FROM AUTHOR]

This paper presents an approach to semiautomatic segmentation over a time series without the use of a priori geometric anatomical knowledge, and demonstrates its applicability to pericardial effusion segmentation from ultrasound. The described technique solves the problem in two stages, first automatically calculating a set of exclusion zones, then leveraging the surgeon's anatomical knowledge, simply and interactively, to create a region which corresponds to the stable region within the target effusion. In preliminary testing, the system performs well versus manual segmentation, outperforming it both in terms of perceived quality, as measured in a blinded comparison by an expert, and in terms of time required for generation. [ABSTRACT FROM AUTHOR]

This paper describes an ultrasound imaging simulator dedicated to the training of physicians for the detection of deep venous thromboses of the lower limbs. Currently, a lot of pathologies of soft tissue are readily diagnosed using ultrasound imaging. This is the case of deep venous thromboses of the lower limbs. Because this examination is difficult and operatordependent, developing a simulator is very useful to give common databases of pathological samples on which physicians can both experiment image acquisition and evaluate their understanding of clinical cases. An ultrasound imaging simulator has been developed. An ultrasound volume is constructed from real images of any typical patient in an off-line pre-processing. Then, simulated images are generated from this volume. The image generation takes into account both the position of the virtual probe and the pressure applied by this probe on the body. Virtual ultrasonic images are generated using a particular interpolation technique and a deformable model of the significant anatomical structures. In most cases, the simulated images are indistinguishable from real ultrasound images. [ABSTRACT FROM AUTHOR]

An automatic image analysis method is presented, which finds the diagnostic landmarks for the determination of the femoral head coverage according to Graf's method in a 3-D ultrasound image volume. Some of the process steps depend on tuneable parameters. It is a typical experience that the quality of images differs between investigators and patients respectively. An image analysis algorithm therefore should produce results which are robust against such varying image quality. The sensitivity of intermediary and final image analysis process steps versus expert defined parameters is investigated. This analysis are performed in a pilot study on 3-D image data of 10 newborns. [ABSTRACT FROM AUTHOR]

One of the drawbacks of many 3D ultrasound systems is the two-stage nature of the process. The strict separation of data acquisition and visualisation disturbs the continuity of the ultrasound examination. Furthermore, some systems require the clinician to wait for an unacceptable amount of time while the voxel array is constructed. In this paper, we describe tools which allow arbitrary slices to be viewed through the 3D data set in realtime as the examination is taking place. This leads to a natural integration of the scanning and visualisation processes. [ABSTRACT FROM AUTHOR]

We present a multimodal volume-based pre-operative surgical planning system developed around the KRDL Virtual Workbench, a mirror-based Virtual Reality system for precise work with 3D data. The planning system is being developed in collaboration with the Singapore General Hospital (SGH), and is undergoing clinical evaluation at SGH and Tan Tock Seng Hospital where 16 cases have been planned. We give a systems perspective on the user interface, introduce the multimodal aspects of the volume rendering technique, and describe the stages involved in creating the final multimodal image for planning with a clinical case: registration, segmentation of tumors, bone removal, path planning and measurements. [ABSTRACT FROM AUTHOR]

This paper describes a low-cost computer system that takes CT images of the knee, and with three-dimensional models of knee prostheses allows a surgeon to position the prosthesis correctly pre-operatively in an interactive manner. Once in position the computer can process bone and prosthesis geometry to derive a set of constraint boundaries that constitute a safe cutting area for a force controlled robot (i.e. that avoids soft tissue such as ligaments), and provides the correct cutting planes for good prosthesis/bone alignment. This boundary information is used to program the robot, allowing a surgeon to move the robot within predefined regions to machine away bone accurately whilst preventing damage to soft tissue. [ABSTRACT FROM AUTHOR]

A procedure for visualization of radiolucent objects in CT scan images is proposed and applied to endosseous oral implant surgery planning. Many of these patients have a removable prosthesis, and visualisation of this prosthesis proves to be advantageous during planning. Such a prosthesis is usually quite radiolucent and thus not distinguishable on CT. The technique glues small markers on the prosthesis, and then scans it, giving a first image set. A second image set of the patient with the intraoral marked prosthesis is then acquired. From the first set, a surface rendered model of the prosthesis can be constructed. From the second set, a surface model of the bone can be made, and the markers indicate the position of the prosthesis in the mouth of the patient. The prosthesis model is then orthogonally transformed so that it fits the bone model. The views obtained are clinically very relevant since they indicate where the teeth of a later fixed prosthesis will come, and the planner can orient the axes of the implants towards the occlusal plane of the teeth. This double scanning procedure is a low-cost technique that nevertheless has significant clinical benefits. [ABSTRACT FROM AUTHOR]

In earlier work [1], we demonstrated that cortical registration could be improved on simulated data by using blurred, geometric, imagebased features (Lvv) or explicitly extracted and blurred sulcal traces on simulated data. Here, the technique is modified to incorporate sulcal ribbons in conjunction with a chamfer distance objective function to improve registration in real MRI data as well. Experiments with 10 simulated data sets demonstrate a 56% reduction in residual sulcal registration error (from 3.4 to 1.5mm, on average) when compared to automatic linear registration and an 28% improvement over our previously published non-linear technique (from 2.1 to 1.5mm). The simulation results are confirmed by experiments with real MRI data from young normal subjects, where sulcal misregistration is reduced by 20% (from 5.0mm to 4.0mm) and 11% (from 4.5 to 4.0mm) over the standard linear and nonlinear registration methods, respectively. [ABSTRACT FROM AUTHOR]

The evaluation of spatial correspondence between binary objects resulting from a segmentation step performed by two different observers or methods is a critical part of the validation of a segmentation criterion or technique. Several global measures of correspondence have been previously proposed, but all of them assume a one-to-one correspondence between objects, thus failing to address local problems such as the splitting of an object by one of the observers. Moreover, such global measures do not distinguish between the reference and the observed objects and most of them lack a solid theoretical foundation. In this paper, we introduce a set of spatial correspondence indices that can evaluate global (many-to-many), local (many-to-one) and individual (one-to-one) spatial correspondence between observed and reference objects and vice versa. The proposed measures, derived from applying information theory concepts to the problem of spatial correspondence, are shown to be well-behaved and suitable to be used in medical imaging applications. [ABSTRACT FROM AUTHOR]

We present a method of 3D/2D image registration. The algorithm is based on the property of near projective invariance in tubular objects. The skeletons of tubular anatomical structures (e.g., intracerebral blood vessels) are used as registration primitives. Experiments with Magnetic Resonance Angiogram (MRA) patient studies and both simulated and actual X-ray angiograms suggest that the algorithm is very accurate and robust. The algorithm requires only a small number of primitives. In addition, the algorithm is relatively insensitive to the choice of tubular structures used. Experimental results justifying these claims are included. [ABSTRACT FROM AUTHOR]

We have implemented automatic 3D thin-plate spline warping as a geometric interpolant to map one dataset volume onto another. Homologous control points in one space are iteratively moved by an optimizer to maximize the global mutual information between the two data volumes. Given two different poses between highly deformed objects we desire to compute the relative geometric deformation using a minimal set of control points as determined by number and placement. The general solution to this problem is not known. In this paper we assess retrospective control point selection for the case of significant patient motion during MRI breast imaging. [ABSTRACT FROM AUTHOR]

We present the design and a prototype implementation of a three-dimensional visualization system to assist with laparoscopic surgical procedures. The system uses 3D visualization, depth extraction from laparoscopic images, and six degree-of-freedom head and laparoscope tracking to display a merged real and synthetic image in the surgeon's video-see-through head-mounted display. We also introduce a custom design for this display. A digital light projector, a camera, and a conventional laparoscope create a prototype 3D laparoscope that can extract depth and video imagery. Such a system can restore the physician's natural point of view and head motion parallax that are used to understand the 3D structure during open surgery. These cues are not available in conventional laparoscopic surgery due to the displacement of the laparoscopic camera from the physician's viewpoint. The system can also display multiple laparoscopic range imaging data sets to widen the effective field of view of the device. These data sets can be displayed in true 3D and registered to the exterior anatomy of the patient. Much work remains to realize a clinically useful system, notably in the acquisition speed, reconstruction, and registration of the 3D imagery. [ABSTRACT FROM AUTHOR]

We have used computer based virtual endoscopy techniques as a novel approach to clarify the 3D surgical anatomy of the pancreas and improve preoperative surgical planning. 13 cases (18 lesions) with mucin-producing pancreas tumors were investigated by "virtual pancreatoscopy". All cystic tumors and the pancreatic ducts were displayed by virtual endoscopy. The surfaces of intraductal papillary adenocarcinomas were illustrated more irregularly than benign cystic lesions. Virtual pancreatoscopy was useful for surgical planning of minimally invasive resection of the pancreas. [ABSTRACT FROM AUTHOR]

In this paper we describe the current state of our surgical planning environment for a conservation method for unilateral glottic cancer with significant subglottic extension. The latter consists of a hemicricolaryngectomy and subsequent reconstruction of the laryngeal defect with a transferable patch of revascularized cervical trachea. In order to restore the three crucial functions of the larynx: airway patency, speech and swallowing, such a trachéal patch has to meet a typical surgical constraint (so-called paramedian position at glottic level). As such constraints are difficult to realise intraoperatively, we designed an image-based surgery simulator for the application. The planning environment visualizes medical image volumes (3D surfaces of the anatomy of interest together with multiplanar reslices) and takes specific surgical constraints into account. The guidance of the surgeon through the planning is menu-driven. To partly test our hypothesis that postoperative morphological results can be optimized by 3D planning, a cadaver study was set up. We report on both the planning environment and on the results of this study. [ABSTRACT FROM AUTHOR]

In this study, a virtual reality vitrectomy simulator is being developed to assist Ophthalmolgy residents in correcting retinal detachments. To simulate this type of surgery, a three dimensional computer eye model was constructed and coupled with a mass-spring system for elastic deformations. Five surgical instruments are simulated including: a pick, blade, suction cutter, laser, and drainage needle. The simulator will be evaluated by a group of fellows and retinal surgeons with a subjective Cooper-Harper survey commonly used for flight simulators. [ABSTRACT FROM AUTHOR]

We have developed an interactive virtual reality (VR) surgical simulator for the training and assessment of suturing technique in the context of end-to-end anastomosis. The surgical simulator is comprised of surgical tools with force feedback, a 3D visual display of the simulated surgical field, physics-based computer simulations of the tissues and tools, and software to measure and evaluate the trainee's performance. This study uses the simulator to compare the skills of experienced vascular surgeons to medical students. Eight parameters were measured to evaluate performance during VR suturing tasks. The data indicate significant differences between surgeon and non-surgeon performance, as well as improvement in performance with training. We believe that this study offers support for the use of virtual reality surgical simulators to augment surgical skill assessment and training. [ABSTRACT FROM AUTHOR]

This paper describes a method for creating object surfaces from binary-segmented data that are free from aliasing and terracing artifacts. In this method, a net of linked surface nodes is created over the surface of the binary object. The positions of the nodes are adjusted iteratively to reduce energy in the surface net while satisfying the constraint that each element in the surface net must remain within its original surface cube. This constraint ensures that fine detail such as cracks and thin protrusions that are present in the binary data are maintained. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of extrapolating very few range data to obtain a complete surface representation of an antomical structure. A new method that uses statistical shape models is proposed and its application to modeling a few points manually digitized on the femoral surface is detailed, in order to improve visualization of a system developped by TIMC laboratory for computer assisted anterior cruciate ligament (ACL) reconstruction. The model is built from a population of 11 femur specimen digitized manually. Data sets are registered together using an elastic registration method of Szeliski and Lavallée based on octree-splines. Principal Components Analysis (PCA) is performed on a field of surface deformation vectors. Fitting this statistical model to a few points is performed by non-linear optimisation. Results are presented for both simulated and real data. The method is very flexible and can be applied to any structures for which the shape is stable. [ABSTRACT FROM AUTHOR]

This paper proposes a method for automated labeling of bronchial branches in virtual bronchoscopy system. Anatomical names of bronchial branches are assigned basing upon a set of rules. If it is possible to assign the anatomical names to bronchial branches extracted from CT images by the computer and is possible to display the name of the currently observed branch on a virtual bronchoscopy image, it will help us to understand the current observing position. The proposed method constructs the knowledge base of anatomical names of bronchial branches. Tree structure of bronchus region is extracted from CT image. Labeling is performed by comparing tree structure information of the extracted bronchus with information of the branches described in the knowledge base. We implemented the proposed method with the virtual bronchoscopy system and obtained satisfactorily result. [ABSTRACT FROM AUTHOR]

We present a hierarchical object-based deformable atlas, a promising new approach for the automatic localization and quantitative analysis of neuroanatomy in MR images. The 3D finite element-based elastic atlas combines the advantages of both volumetric-and surface-based deformable atlases in one single unifying framework. This multiresolution framework is not only capable of deforming entire volumes or subvolumes but can deform individual atlas objects, allowing greater and more effective use of object shape and local image feature information. Object surface representations are embedded in the volumetric deformable atlas and image-feature-derived forces acting on these surfaces are automatically transferred to the containing 3D finite element lattice. Consequently, spatial relationship constraints of the atlas objects are maintained via the elastic lattice while an object is deformed to match a target boundary. Atlas objects are deformed in a hierarchical fashion, begining with objects exhibiting well-defined image features in the target scan and proceeding to objects with slightly less well-defined features. Experiments involving several subcortical atlas objects are presented. [ABSTRACT FROM AUTHOR]

So far, exploration of volume models is limited to cut planes or addition/removal of segmented objects. More capable exploration techniques are needed in order to allow a ‘look and feel' close to a real dissection. This is especially important for applications like the simulation of osteotomy surgery. Therefore, we have developed methods for free-form volume-sculpting operations which allow the interactive specification, representation and high-quality rendering of free form regions. The novelty of this approach is that these regions are represented within the generalized-voxel-model, together with a simulation of the partial-volume-effect, which allows a sub-voxel localization of cut surfaces. These techniques are implemented in our VOXEL-MAN visualization system, thus enhancing the exploration techniques for volume data. Furthermore, we developed an extended ray-casting algorithm for 3D-visualization of object motion with detection and visualization of interpenetrating volumes. These methods together provide a powerful tool for volume exploration and applications like the rehearsal of surgical interventions. [ABSTRACT FROM AUTHOR]

Several authors have recently compared the results of fMRI studies on neurosurgery patients with invasive electrophysiology. These studies aim to validate MRI against an accepted gold standard, and ascertain whether fMRI could replace invasive electrophysiology in neurosurgical patients. We have identified and quantified two characteristics of these data that make such comparisons problematic. Firstly, the epilepsy surgery patients (n=8) studied move significantly more during fMRI experiments than normal volunteers (n=6) performing the same task. This motion has a particularly large out-of-plane component, and is significantly more correlated with the stimulus than for the normal volunteers. This motion is especially large when performing a task on the side affected by the lesion. This additional motion is hard to correct and substantially degrades the quality of the resulting fMRI images, making it a much less reliable technique on these surgical patients than on other subjects. Secondly, we have found that, following electrode implantation, the brain surface can shift by more than 10 mm relative to the skull compared to its preoperative location, substantially degrading the accuracy of the comparison of electrophysiology measurements made on the deformed brain and fMRI studies carried out preoperatively. Taken together, these findings suggest that studies of this sort are currently of limited use for validating fMRI, and further image analysis research is necessary to solve the problems caused by subject motion and brain deformation. [ABSTRACT FROM AUTHOR]

Measuring motion in medical imagery becomes more and more important, in particular for object tracking, image registration, and local displacement measurements. Such measurements are especially difficult in megavoltage X-ray images (portal images), which are used to control the position of patients in high precision radiotherapy. Low contrast, blur, and noise render accurate measurements difficult. In this work we review the framework of a generic matching algorithm only based on the image signal and not on binary image features. Thus, the often unreliable step of feature extraction in such imagery is circumvented. Another major advantage is the possibility of self-diagnosis, which is used for restricting the transformation in motion measurements if the image quality is not sufficient. The method of digitally reconstructed radiographs (DRR) allow for the computation of error free reference images, avoiding the additional step of therapy simulation. The multi-modal match between such DRRs and portal images lead to an estimate of the patient position during radiotherapy treatment. Results of generated data with known ground truth as well as results of a multi-modal match are presented. [ABSTRACT FROM AUTHOR]

Nowadays, conventional X-ray radiographs are still the images of choice for evaluating spinal deformaties such as scoliosis. However, digital translation reconstruction gives easy access to high quality, digital overview images of the entire spine. This work aims at improving the description of the scoliotic deformity by developing semi-automated tools to assist the extraction of anatomical landmarks (on vertebral bodies and pedicles) and the calculation of deformity quantifying parameters. These tools are currently validated in a clinical setting. [ABSTRACT FROM AUTHOR]

The endovascular coil embolisation of intracranial saccular aneurysms requires a set of specific X-ray images with which to view the aneurysm during coiling. These two-dimensional images, known asworking projectionsshould be optimal for measuring the aneurysm sac diameter, inserting the first coil, and checking coil overhang into the surrounding vessels. At present the gantry angle that produces these images is found by the radiologist by trial and error. In this paper we present a method for automatically finding the angles that will produce the desired X-ray projections. Our method consists of four steps: (1) segmenting the vasculature from three-dimensional angiographic data; (2) locating the aneurysm neck; (3) labelling the aneurysm sac; and (4) determining the optimal angle for viewing the aneurysm during coiling. We discuss details of the algorithm steps and present the results of the algorithm applied to one synthetic and two pathological examples. [ABSTRACT FROM AUTHOR]

In this paper we address the problem of spatio-temporal acoustic boundary detection in echocardiography. We propose a phase-based feature detection method to be used as the front end to higher-level 2D+T/3D+T reconstruction algorithms. We develop a 2D+T version of this algorithm and illustrate its performance on some typical echocar-diogram sequences. We show how our temporal-based algorithm helps to reduce the number of spurious feature responses due to speckle and provides feature velocity estimates. Further, our approach is intensity-amplitude invariant. This makes it particularly attractive for echocardiographic segmentation, where choosing a single global intensity-based edge threshold is problematic. [ABSTRACT FROM AUTHOR]