The purpose of this chapter is to present a survey of recent publications concerning medical image registration techniques. These publications will be classified according to a model based on nine salient criteria, the main dichotomy of which is extrinsic versus intrinsic methods The statistics of the classification show definite trends in the evolving registration techniques, which will be discussed. At this moment, the bulk of interesting intrinsic methods is either based on segmented points or surfaces, or on techniques endeavoring to use the full information content of the images involved. Keywords: registration, matching Received May 25, 1997

Abstract. 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.

. Augmented reality #AR# is a technology in which a user's view of the real world is enhanced or augmented with additional information generated from a computer model. Using AR technology, users can interact with a combination of real and virtual objects in a natural way. This paradigm constitutes the core of a very promising new technology for many applications. However, before it can be applied successfully, AR has to ful#ll very strong requirements including precise calibration, registration and tracking of sensors and objects in the scene, as well as a detailed overall understanding of the scene. At ECRCwe see computer vision and image processing technology play an increasing role in acquiring appropriate sensor and scene models. To balance robustness with automation, weintegrate automatic image analysis with both interactive user assistance and input from magnetic trackers and CAD-models. Also, in order to meet the requirements of the emerging global information society...

While the role and utility of Magnetic Resonance Images as a diagnostic tool is well established in current clinical practice, there are a number of emerging medical arenas in which MRI can play an equally important role. In this article, we consider the problem of image-guided surgery, and provide an overview of a series of techniques that we have recently developed in order to automatically utilize MRI-based anatomical reconstructions for surgical guidance and navigation. 1 Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge MA 2 Department of Radiology, Brigham and Womens Hospital, Harvard Medical School, Boston MA 1 Introduction In recent years, Magnetic Resonance Imaging (MRI) has become a commonplace medical diagnostic tool [2], especially for cases involving soft tissue, such as in the brain. Two factors combine to make MRI a very valuable clinical tool: fine scale spatial resolution allows for the detection and delineation of detailed stru...

Fast, simple and effective registration methods are needed in a wide variety of computer-assisted surgical procedures in which readily locatable anatomical landmarks are not available. Orthopedic procedures about the knee, in particular, are adversely affected if the registration accuracy exceeds about 1 mm in translation or about ¤ ¥ in rotation, and the proximal tibia and distal femur are typically devoid of small distinct features. Surface-based least-squares registration methods can be used, but are susceptible to poor initial pose estimates and to error contamination during intraoperative data collection. We have developed a fast, statistically robust method for surface-based registration. The method is based on the iterative closest point (ICP) algorithm, and fits a set of sparsely measured data points to a planar facet model. An initial registration estimate is obtained by having the user contact the anatomy in a set of general anatomical regions (rather than contacting distinctive features). A small number of additional data points are acquired to refine the registration. Starting from the initial estimate, a robust scored perturbation method is used to find an initial registration. This is followed by an M-estimate ICP registration which is taken as the final registration. Simulation results show that this method is robust for data sets containing up to 25 % gross outliers. The method has been tested in vitro on plastic bone models, where it robustly outperformed the least-squares estimate and maintained the required 1mm/¤ ¥ accuracy. The in vivo use of spotlights in computer-enhanced osteotomies of the knee and wrist have confirmed that the method is easy to use and sufficiently accurate. 1.

Validating the accuracy of image-guided surgical systems is a challenging and important problem which has received little attention in the literature. Potential sources of inaccuracy include: CT imaging noise, model generation errors, errors introduced by fixturing, intra-operative data noise, registration errors, and inaccuracies in surgical tools and actions. In this paper, we discuss our experience in validating an example image-guided application in orthopaedic surgery. Various sources of inaccuracy are identified and approaches for mitigating their effects are suggested. The difficult problem of generating a reliable ground-truth for evaluating the accuracy of surgical registration is discussed, and surface-based registration accuracy results are presented. A fiducial marker design which can be used to establish highly accurate

We describe a method for mapping the functional regions of the brain using a transcranial magnetic stimulation (TMS) device. This device, when placed on a subject's scalp, stimulates the underlying neurons by generating focused magnetic field pulses. A brain mapping is then generated by measuring responses of different motor and sensory functions to this stimulation. The key process in generating this mapping is the association of the 3D positions and orientations of the TMS probe on the scalp to a 3D brain reconstruction such as is feasible with a magnetic resonance image (MRI). We perform this matching process by (1) registering the subject's head position to an a priori MRI scan, (2) tracking the 3D position/orientation of the TMS probe, (3) transforming the TMS probe position/orientation to the MRI coordinate frame, and (4) tracking movements in the subject's head position to factor out any head motion. The resultant process generates a high resolution, accurate brain mapping which...

Medical imagery, in particular three-dimensional imagery, provides highly accurate and informative insight into internal anatomical structure and function. While such data is becoming readily available for the clinician, very few tools have been developed to properly exploit the structural and geometric information inherent in the images. A key problem in medical image analysis is registration--the process of aligning datasets to the same coordinate frame. Accurate and automated image registration facilitates powerful solutions for such applications as surgical guidance, disease diagnosis, and treatment monitoring. A hierarchical 3D surface-based optimization approach is proposed to address these registration problems. The goals are to register same-subject imagery taken over time in a wide range of situations ranging from rigid to deformable tissues and to evaluate registration accuracy and reliability. Resultant prototype registration systems have been applied to image-guided surgery...

INTRODUCTION Intra-operative position sensing and tracking devices are fundamental building blocks which are often used in Computer-Assisted Orthopaedic Surgery (CAOS). These devices are used during surgery to precisely localize conventional surgical tools, rigid anatomical structures, other medical imaging equipment (e.g., X-Ray or ultrasound scanners), surgical implants, etc. For example, in computer-assisted spine surgery, it may be desirable to determine the position and orientation of a drill relative to a vertebral body. Intra-operative sensors can be used to independently measure the locations of the vertebral body and the drill, and then the relative position and orientation can be inferred. In this abstract, the term position sensor refers to a type of sensor which can localize one or more discrete points in 3-D space. Such sensors can also compute object orientation when 3 or more measured points are attached to a single rigid object. The term

Augmented reality (AR) is a technology by which a user's view of the real world is augmented with additional information from a computer model. It constitutes a very promising new user interface concept for many applications. Yet, AR applications require fast and accurate solutions to several very complex problems, such as user and real object tracking, occlusion and reflection handling, as well as virtual user motion. Currently, computer vision based solutions are considered to be among the most promising approaches towards solving these issues. This paper discusses several such AR issues and potential solutions.