Quantitative assessment of regional heart motion has significant potential to provide more specific diagnosis of cardiac disease and cardiac malfunction than currently possible. Local heart motion may be captured from various medical imaging scanners. In this study, 3-D reconstructions of pre-infarct and post-infarct hearts were obtained from the Dynamic Spatial Reconstructor[1,2,3] (DSR). Using functional parametric mapping of disturbances in regional contractility and relaxation, regional myocardial motion during a cardiac cycle is color mapped onto a deformable heart model to facilitate appreciation of the structure-to-function relationships in the myocardium, such as occurs in regional patterns of akinesis or dyskinesis associated with myocardial ischemia or infarction resulting from coronary artery occlusion.

Regional heart wall dynamics has been shown to be a sensitive indicator of LV wall ischemia. Rates of local LV wall thickening during a cardiac cycle can be measured and illustrated using functional parametric mappings. This display conveys the spatial distribution of dynamic strain in the myocardium and thereby provides a rapid qualitative appreciation of the severity and extent of the ischemic region.3D reconstructions were obtained in an anesthetized pig from 8 adjacent, shortaxis, slices of the left ventricle imaged with an Electron Beam Computer Tomograph at 11 time points through one complete cardiac cycle. The 3D reconstructions were obtained before and after injection of 100 mm microspheres into the Left Anterior Descending (LAD) coronary artery. This injection causes microembolization of LAD artery branches within the heart wall. The image processing involved radially dividing the tomographic images of the myocardium into small subdivisions with color encoding of the local magnitude of regional thickness or regional velocities of LV wall thickening throughout the cardiac cycle. We compared the effectiveness of animation of wall thickness encoded in color versus a static image of computed rate of wall thickness change in color. The location, extent and severity of regional wall akinesis or dyskinesis, as determined from these displays, can then be compared to the region of embolization as indicated by the distribution of altered LV wall perfusion.

3D Trans-Urethral Ultrasound (TUUS) imaging is a new imaging technique for the diagnosis and treatment of prostate disease. Our current research focuses on the potential of TUUS in therapy guidance during transperineal interstitial permanent prostate brachytherapy (TIPPB). TUUS may complement or potentially replace x-ray fluoroscopy and TRUS in providing data for determining the prostate boundary and radiation source locations. Prostate boundary detection and source localization using TUUS were tested on an ultrasound-equivalent prostate phantom and in a patient during TIPPB. Data collection was conducted with a 10 French, 10 MHz ultrasound catheter controlled by an Acuson Sequoia workstation. 2D and 3D TUUS scans were acquired after radioactive seeds were placed in the phantom and in the patient. Data was reconstructed, processed, and analyzed using Analyze software. Segmentation of the prostate boundary was performed semi-automatically, and seed segmentation was performed manually. Image artifacts in TUUS data resulted in incorrect reconstruction of the seeds. Intelligent processing of the seed data improved reconstruction. Comparison to the CT data suggests that TUUS data provides: 1) greater spatial resolution, 2) greater temporal resolution and 3) better contrast for soft tissue differentiation. The reconstructed source sizes and locations were measured and found accurate. Placement of the TUUS catheter into the urethra provides excellent 2D sections which can be used to acquire volumetric data for 3D analysis of the prostate and radioactive sources. Preliminary results suggest that TUUS will be useful for guidance of seed placement, post-implant seed localization, and intra-operative dosimetry.

Quantitative assessment of 3-D regional heart motion has significant potential to provide more specific diagnosis of cardiac malfunction than currently possible. Using functional parametric mapping, regional myocardial motion during a cardiac cycle can be color-mapped onto a deformable heart model to provide better understanding of the structure-to-function relationships in the myocardium, including regional patterns of akinesis or dyskinesis associated with ischemia or infarction. In this study, 3-D reconstructions of human hearts were obtained from Electron-Beam Computed Tomography (EB-CT), comparing stages of treatment after myocardial infarction.

Visualizable objects in biology and medicine extend across a vast range of scale, from individual molecules and cells, through the varieties of tissue and interstitial interfaces, to complete organs, organ systems and body parts, and include functional attributes of these systems, such as biophysical, biomechanical and physiological properties. Visualization in three dimensions of such objects and their functions is now possible with the advent of high resolution tomographic scanners and imaging systems. Medical applications include accurate anatomy and function mapping, enhanced diagnosis, accurate treatment planning and rehearsal, and education/training. Biologic applications include study and analysis of structure to function relationships in individual cells and organelles. The potential for revolutionary innovation in the practice of medicine and in biologic investigations lies in direct, fully immersive, real-time multisensory fusion of real and virtual information data streams into online, real-time visualizations available during actual clinical procedures or biological experiments. Current high-performance computing, advanced image processing and high fidelity rendering capabilities have facilitated major progress toward realization of these goals. With these advances in hand, there are several important applications of 3-D visualization that will have a significant impact on the practice of medicine and on biological research.

Stereotactic navigational systems have demonstrated signi?cant clinical in?uence, and are being incorporated into an increasing number of neurosurgical procedures. Preoperatively acquired 3D images are used for planning the procedure, and are also employed in intraoperative navigations to help localize and resect brain lesions. However, as the operation progresses, multiple factors contribute to changes that limit the accuracy of the navigation based on pre-operative images alone. The opening of the dura with the associated loss of cerebrospinal ?uid and cortical swelling, craniotomy location, tumor decompression, and collapse of neural tissue around the operative site are some of the clinical factors that contribute to brain shift and consequent errors in navigation, particularly when the navigation is based solely on pre-operatively acquired images. Our method to correct for brain shift involves the use of ultrasound intraoperatively to update patient specifc pre-operative MRI scans using a physics based dynamic model. To validate the imaging and modeling process, a phantom was designed that simulates the brain and its shifting patterns resulting from several of the clinical factors present during a brain operation. MRI and Ultrasound datasets were acquired for several permutations of phantom parameters. Deformation algorithms were then applied to the phantom data to demonstrate the efficacy of this approach as a method to effectively update the pre-operatively acquired MRI data during an operation.

Study of cardiac dynamics requires analysis of multi-dimensional spatial and temporal variables and complex electromechanical coupling mechanisms. Therefore it requires careful attention to these factors in order to develop a model of the heart that realistically simulates cardiac function. Motion information can be extracted by examining and tracking two- to four-dimensional image data. Electro-physiological recordings can provide data on the conduction system of the heart. By effectively combining and representing such data, both qualitative and quantitative analyses can be carried out to reveal important dynamic changes during the heartbeat. To acquire an entire heart cycle from a patient is often not possible in clinical practice. Also, the parameters derived from the patient’s pre-operative image data cannot reflect changes in the patient’s cardiac morphology and physiology during or after operations. Normal functioning of the heart depends on healthy and synchronous conduction and contraction systems, which are influenced heavily by the anisotropic fiber microstructure of the heart. We have developed a method that incorporates muscle fiber track information, a physics-based deformable model and a classic electrical conduction system to realistically simulate cardiac dynamics. The simulation aims to accurately reproduce myocardial motion during the heartbeat. The simulation provides a multi-dimensional visualization of cardiac dynamics. In complex clinical procedures, such as intra-cardiac catheter ablation, a real-time dynamic heart model that displays the patient's updated heart anatomy and physiology with low latency during the procedure provides an effective image-guided tool for optimal therapeutic results.

Quantitative assessment of regional heart motion has significant potential for more accurate diagnosis of heart disease and / or cardiac irregularities. Local heart motion may be studied from medical imaging sequences. Using functional parametric mapping, regional myocardial motion during a cardiac cycle can be color mapped onto a deformable heart model to obtain better understanding of the structure-to-function relationships in the myocardium, including regional patterns of akinesis or diskinesis associated with ischemia or infarction. In this study, 3D reconstructions were obtained from the Dynamic Spatial Reconstructor (DSR) at 15 time points throughout one cardiac cycle of pre-infarct and post-infarct hearts. Deformable models were created from the 3-D images for each time point of the cardiac cycles. From these polygonal models, regional excursions and velocities of each vertex representing a unit of myocardium were calculated for successive time intervals. The calculated results were visualized through model animations and / or specially formatted static images. The time point of regional maximum velocity and excursion of myocardium through the cardiac cycle was displayed using color mapping. The absolute value of regional maximum velocity and maximum excursion were displayed in a similar manner. Using animations, the local myocardial velocity changes were visualized as color changes on the cardiac surface during the cardiac cycle. Moreover, the magnitude and direction of motion for individual segments of myocardium could be displayed. Comparisons of these dynamic parametric displays suggest that the ability to encode quantitative functional information on dynamic cardiac anatomy enhances the diagnostic value of 4D images of the heart. Myocardial mechanics quantified this way adds a new dimension to the analysis of cardiac functional disease, including regional patterns of akinesis and diskinesis associated with ischemia and infarction. Similarly, disturbances in regional contractility and filling may be detected and evaluated using such measurements and displays.

Modeling of moving anatomic structures is complicated by the complexity of motion intrinsic and extrinsic to the structures. However when motion is cyclical, such as in heart, effective dynamic modeling can be approached using modern fast imaging techniques which provide three-dimensional structural data. Data may be acquired as a sequence of 3-D volume images throughout the cardiac cycle. To model the intricate non-linear motion of the heart, we created a physics-based surface model which can realistically deform between successive time points in the cardiac cycle, yielding a dynamic four-dimensional model of cardiac motion. Sequences of fifteen 3-D volume images of intact canine beating hearts were acquired during complete cardiac cycles using the Dynamic Spatial Reconstructor (DSR) and the Electron Beam CT (EBCT). The chambers of the heart were segmented at successive time points, typically at 1/15-second intervals. The left ventricle of the first time point (near enddiastole) was reconstructed as an initial triangular mesh. A mass-spring physics-based deformable model, which can expand and shrink with local contraction and stretching forces distributed in an anatomically accurate simulation of cardiac motion, was applied to the initial mesh and allowed the initial mesh to deform to fit the left ventricle in successive time increments of the sequence. The resultant 4-D model can be interactively transformed and displayed with associated regional electrical activity mapped onto the anatomic surfaces, producing a 5-D model, which faithfully exhibits regional cardiac contraction and relaxation patterns over the entire heart. The beating heart model can be interactively transformed and viewed from different angles, showing regional cardiac contraction and relaxation over the entire heart. For acquisition systems that may provide only limited 4-D data, (e.g., only images at end-diastole and end-systole) the model can provide interpolated anatomic shapes between time points. This physics-based deformable model accurately represents dynamic cardiac structural changes throughout the cardiac cycle. Such models provide the framework for minimizing the number of time points required to usefully depict regional motion of myocardium and allowing quantitative assessment of regional myocardial dynamics. The electrical activation mapping provides spatial and temporal correlation within the cardiac cycle. In procedures such as intra-cardiac catheter ablation, visualization of the dynamic model can be used to accurately localize the foci of myocardial arrhythmias and guide positioning of catheters for effective ablation.

Quantitative assessment of regional heart motion has significant potential for more accurate diagnosis of heart disease and / or cardiac irregularities. Local heart motion may be studied from medical imaging sequences. Using functional parametric mapping, regional myocardial motion during a cardiac cycle can be color mapped onto a deformable heart model to obtain better understanding of the structure-to-function relationships in the myocardium. In this study, 3D reconstructions were obtained from the Dynamic Spatial Reconstructor (DSR) at 15 time points throughout one cardiac cycle. Deformable models were created from the 3-D images for each time point of the cardiac cycle. From these polygonal models, regional excursions and velocities of each vertex representing a unit of myocardium were calculated for successive time intervals. The calculated results were visualized through model animations and / or specially formatted static images. The time point of regional maximum velocity and excursion of myocardium through the cardiac cycle was displayed using color mapping. The absolute value of regional maximum velocity and maximum excursion were displayed in a similar manner. Using animations, the local myocardial velocity changes were visualized as color changes on the cardiac surface during the cardiac cycle. Moreover, the magnitude and direction of motion for individual segments of myocardium could be displayed. These results suggest that the ability to encode quantitative functional information on dynamic cardiac anatomy enhances the diagnostic value of 4D images of the heart. Myocardial mechanics quantified this way adds a new dimension to the analysis of cardiac functional disease, including diastolic filling deficits and / or disturbances in regional electrophysiology and contraction patterns.

There is a growing consensus that mutual voxel information based measures hold great promise for fully automated multimodal image registration. We have found that image greyscale binning using a specific variation of contrast-limited histogram equalization (which we call histogram preservation) provides significant reduction of noise and spurious local maxima in the normalized mutual information function without causing significant displacement or smoothing of the global maximum. These effects are also relatively robust in the presence of image subsampling, so that accurate subpixel coregistration of typical medical volume images may be achieved in a few seconds by a very simple optima search algorithm based on a few thousand sampled voxels. In this paper, we illustrate these effects by presenting the results of random tests on patient data. Intramodal performance is evaluated by image self-reregistration using a variety of patient image volumes. Reregistration error is measured as the mean of the residual Euclidean displacement of the eight corner points of the image volumes after reregistration. The performance of histogram preservation prebinning is compared to linear prebinning, and the effect of image subsampling and number of bins on algorithm speed and accuracy is also assessed.

Visualizable objects in biology and medicine extend across a vast range of scale, from individual molecules and cells, through the varieties of tissue and interstitial interfaces, to complete organs, organ systems and body parts. These objects include functional attributes of these systems, such as biophysical, biomechanical and physiological properties. Visualization in three dimensions of such objects and their functions is now possible with the advent of high-resolution tomographic scanners and imaging systems. Medical applications include accurate anatomy and function mapping, enhanced diagnosis, accurate treatment planning and rehearsal, and education/training. Biologic applications include study and analysis of structure-to-function relationships in individual cells and organelles. The potential for revolutionary innovation in the practice of medicine and in biologic investigations lies in direct, fully immersive, real-time multisensory fusion of real and virtual information data streams into online, real-time visualizations available during actual clinical procedures or biological experiments. Current high-performance computing, advanced image processing and high-fidelity rendering capabilities have facilitated major progress toward realization of these goals. With these advances in hand, there are several important applications of three-dimensional visualization that will have a significant impact on the practice of medicine and on biological research.

For computer assisted surgical simulation to be effective, objects in the simulated environment should respond to the user’s actions dynamically with correct visual information. This includes dragging and cutting that cause changes in geometry, topology and appearance. Geometric object representation can be manipulated intuitively in real-time but does not preserve interior information. Volumetric data representation, on the other hand, preserves volume content but direct manipulation is compute-intensive. 3-D texture mapping provides an alternative in representing volumetric information. We present a surgical simulation system based on geometric models that allows interactive deformation and incision of objects while displaying correct volumetric information corresponding to these changes. This is accomplished by dynamic 3-D texture mapping. This method can be applied to anatomical data and patient CT and MR images to facilitate data/patient specific surgical simulations.

Virtual endoscopy (or computed endoscopy) is a new method of diagnosis using computer processing of 3-D image datasets (such as CT or MRI scans) to provide simulated visualizations of patient specific organs similar or equivalent to those produced by standard endoscopic procedures. Conventional CT and MRI scans produce cross section "slices" of the body that are viewed sequentially by radiologists who must imagine or extrapolate from these views what the actual 3 dimensional anatomy should be. By using sophisticated algorithms and high performance computing, these cross sections may be rendered as direct 3-D representations of human anatomy. Specific anatomic data appropriate for realistic endoscopic simulations can be obtained from 3-D MRI digital imaging examinations or 3D acquired spiral CT data.

Thousands of endoscopic procedures are performed each year. They are invasive and often uncomfortable for patients. They sometimes have serious side effects such as perforation, infection and hemorrhage. Virtual endoscopic visualization avoids the risks associated with real endoscopy, and when used prior to performing an actual endoscopic exam can minimize procedural difficulties and decrease the rate of morbidity, especially for endoscopists in training. Additionally, there are many body regions not accessible to or compatible with real endoscopy that can be explored with virtual endoscopy. Eventually, when refined, virtual endoscopy may replace many forms of real endoscopy.

Grayscale inhomogeneities in magnetic resonance (MR) images cause significant problems in automated quantitative image analysis. Removal of such inhomogeneities is a difficult task, but it has been investigated by a number of different authors recently. The most common methods used involve some type of homomorphic filtering to create a smoothed version of the original image, which is then used as an estimate of the bias field to be removed from the image. Many investigators have implemented variations of this technique and have demonstrated their usefulness for a wide range of applications, but no investigator has yet attempted a systematic, quantitative study to describe the effects these algorithms have on images. This study introduces a quantitative paradigm for evaluating inhomogeneity correction algorithms by their performance on a constructed simulation image with different bias fields applied. We find that mean filter algorithms are more successful than median filter algorithms, and that larger kernel sizes than what are currently reported in the literature offer significant improvements in post-correction image quality.

A comprehensive software system for interactive visualization, manipulation and measurement of multi-modality 3-D medical images has been developed, used and evaluated at the Mayo Clinic for Computer Aided Surgery (CAS) and Radiation Treatment Planning (RTP) for more than a decade (1,2). This software system, called ANALYZE , has provided surgeons and physicians with powerful and flexible computational support both for pre-operative surgical and treatment planning and for post-operative evaluation. The software has been applied to a variety of surgical and medical problems, and used on significant numbers of patients at the Mayo Clinic and at many other institutions (3,4). This scope and depth of clinical experience has fostered continual refinement of approaches and techniques, and has provided significant information and insights related to the practical clinical usefulness of CAS and RTP and their impact on medical treatment outcome and cost. This experience and information has led to design of a new approach to CAS using Virtual Reality technology which holds significant promise for optimizing certain surgical procedures, minimizing patient risk and morbidity, and reducing health care costs. Practicing Mayo surgeons, physicians and therapists are committed to assisting with development, evaluation and routine clinical implementation of this system.

To be successful, a medical virtual reality (VR) system needs to accurately segment a desired object from volumetric data, detect its surface and generate a good polygonal representation of this surface from a fixed polygonal "budget" (defined as the constraint on the number of polygons for effective real time display). Currently available hardware is able to render and manipulate approximately 20,000 complex polygons (a complex polygon includes shading, texture mapping and anti-aliasing) per frame at real time rates. Polygonization algorithms such as Marching Cubes, Spiderweb and the Wrapper produce high resolution surfaces using 40,000 to several million polygons. In this paper, we present a method for the production of a polygonal surface containing a selected number of polygons from volumetric data; based on extracting a set of curvature weights from the volumetric data and using these weights as the input vectors to a 2-D Kohonen network. The adaptation of the network to the input vectors results in a polygonal suurface of 1,000 to 20,000 polygons that preserves useful detail and produces realistic geometric models from patient-specific volumetric data.

We are developing a system called Virtual Reality Assisted Surgery Program (VRASP) for implementation into the hospital operating room. VRASP will give the surgeon flexible computational support intraoperatively. It will permit modification and control of very large scan datasets in real time. It will render and transmit virtual imagery in response to the surgeon's commands without interfering with normal surgical activities. And it will register the displayed imagery simultaneously with respect to the surgeon and the patient, without computing or display lag. The project is designed in three phases: 1) surgery planning, 2) surgery rehearsal and 3) surgery delivery. The first phase has been implemented, and significant experience has been gained to facilitate effective design and implementation of the second and third phases.

VRASP is being developed at Mayo to specifically assist surgeons during craniofacial, orthopedic, brain and prostate surgery. VRASP will enable surgeons to interactively visualize 3-D renderings of CT and MRI data with hands-free manipulation of the virtual display. The surgeon will be able to scale, orient and position prescanned body imagery on-line in real time from any desired perspective. The clinical goal is dynamic fusing of 3-D body scan data with the actual patient in the operating room. The customized interface will permit ready, on-line access to the preoperative plan and to update measurement and analysis based on the real-time operating room data.

Practicing Mayo surgeons are committed to assisting with development, evaluation and deployment of the VRASP system. VRASP will bring to the OR all of the pre-surgical planning data and rehearsal information in synchrony with the actual patient and operation in order to optimize the effectiveness of the procedure, minimize patient morbidity, and reduce health care costs

Thousands of radical prostatectomies for prostate cancer are performed each year . Radical prostatectomy is a challenging procedure due to anatomical variability and the adjacency of critical structures, including the external urinary sphincter and neurovascular bundles that subserve erectile function. Because of this, there are significant risks of urinary incontinence and impotence following this procedure. Preoperative interaction with three-dimensional visualization of the important anatomical structures might allow the surgeon to understand important individual anatomical relationships of patients. Such understanding might decrease the rate of morbidities, especially for surgeons in training. Patient specific anatomic data can be obtained from preoperative 3D MRI diagnostic imaging examinations of the prostate gland utilizing endorectal coils and phased array multicoils. The volumes of the important structures can then be segmented using interactive image editing tools and then displayed using 3-D surface rendering algorithms on standard work stations. Anatomic relationships can be visualized using surface displays and 3-D colorwash and transparency to allow internal visualization of hidden structures. Preoperatively a surgeon and radiologist can interactively manipulate the 3-D visualizations. Important anatomical relationships can better be visualized and used to plan the surgery. Postoperatively the 3-D displays can be compared to actual surgical experience and pathologic data. Patients can then be followed to assess the incidence of morbidities. More advanced approaches to visualize these anatomical structures in support of surgical planning will be implemented on virtual reality (VR) display systems. Such realistic displays are "immersive", and will allow surgeons to simultaneously see and manipulate the anatomy, to plan the procedure and to rehearse it in a realistic way. Ultimately the VR systems will be implemented in the operating room (OR) to assist the surgeon in conducting the surgery. Such an implementation will bring to the OR all of the pre-surgical planning data and rehearsal experience in synchrony with the actual patient and operation to optimize the effectiveness and outcome of the procedure.

A significant increase in diagnostic incidence of prostate cancer underscores the need to accurately stratify and quantify the cancers to facilitate appropriate therapy. Currently there is no reliable method to preoperatively predict pathological stage and thus malignant potential of prostate cancer. Tumor volume and microvessel density have been shown postoperatively to be accurate predictors of cancer metastatic potential. Three-dimensional visualization and analysis of image volumes produced from series of immunocytochemically stained pathological sections may improve our understanding of the relationships of the tumor to angiogenesis, i.e., to the microvessel density of the tumor. Sequential thinly sliced (~7 microns) pathological sections of the prostate will be differentially stained with fluorescent antibodies to clotting factor VIII-related antigen, which labels the endothelial cells of the vessels, facilitating automated color separation for visualization of the microvessels. Digitized images of the sections can be synthesized into 3-D volumes and measured to quantify vessel quantity and density. Using 3-D colorwash and transparency display techniques, anatomic and densitrometric relationships between the tumor and microvessels can be visualized. The microvessel density can be measured using image processing algorithms and compared to measurements made by pathologists. Advanced approaches to imaging the prostate in vivo include dynamic MRI techniques using contrast agents to accurately detect and quantify the region of prostate cancer. The cancerous region can be correlated with histologic specimens using the same methods described for measurement of microvessel density. This detailed information could lead to improved methods to properly stratify patients with diagnosed prostate cancer

Multimodality images obtained from different medical imaging systems such as Magnetic Resonance (MR), Computed Tomography (CT), Ultrasound (US), Positron Emission Tomography (PET), Single Photo Emission Computed Tomography (SPECT), provide largely complementary characteristic or diagnostic information. Therefore, it is an important research objective to "fuse" or combine this complementary data into a composite form which would provide synergistic information about the objects under examination. An important first step in the use of complementary fused images is 3D image registration, where multi-modality images are brought into spatial alignment so that the point-to-point correspondence between image data sets is known. Current research in the field of multimodality image registration has resulted in the development and implementation of several different registration algorithms, each with its own set of requirements and parameters. Our research has focused on the development of a general paradigm for measuring, evaluating and comparing the performance of different registration algorithms. Rather than evaluating the results of one algorithm under a specific set of conditions, we suggest a general approach to validation using simulation experiments, where the exact spatial relationship between data sets is known, along with phantom data to characterize the behaviour of a algorithm via a set of quantitative image measurements. This behaviour may then be related to the algorithms performance with real patient data, where the exact spatial relationship between multimodality images is unknown. Current results indicate that our approach is general enough to apply to several different registration algorithms. Our methods are useful for understanding the different sources of registration error and for comparing the results between different algorithms.

Multispectral classification uses registered 3-D image volumes from more than one imaging modality or from different sequences within a modality to classify tissues within those volumes. The complementary information contained within the different image volumes may allow for the seperation of tissue class types in multidimensional feature space when the same tissue classes would be indistinct using just one image volume. When segmentation is complete, attributes of these classes may be determined (e.g., volumes), or the classes may be visualized as objects in 3-D. There are two main types of classification algorithms: supervised and unsupervised. Unsupervised classifiers offer the promise of totally automated classification of tissue types and calculation of tissue volumes and other tissue properties in medical images,. This would have two benefits: (1) elimination of the time-consuming process of manual segmentation by medical experts, and (2) ensuring reproducable results. While accurate performance by unsupervised classifiers is, in general, still impossible, an intermediate step is the development of tools to allow used to obtain useful results in a relatively short period of time. This paper describes such a tool which allows users to quickly and easily experiment with various choices of unsupervised classification algorithms and their input parameters and evaluate the results.

We are developing a system called Virtual Reality Assisted Surgery Program (VRASP) for implementation into the hospital operating room. VRASP will give the surgeon flexible computational support intraoperatively. It will permit modification and control of very large scan datasets in real time. It will render and transmit virtual imagery in response to the surgeon's commands without interfering with normal surgical activities. And it will register the displayed imagery simultaneously with respect to the surgeon and the patient, without computing or display lag. The project is designed in three phases: 1) surgery planning, 2) surgery rehearsal and 3) surgery delivery. The first phase has been implemented, and significant experience has been gained to facilitate effective design and implementation of the second and third phases.

VRASP is being developed at Mayo to specifically assist surgeons during craniofacial, orthopedic, brain and prostate surgery. VRASP will enable surgeons to interactively visualize 3-D renderings of CT and MRI data with hands-free manipulation of the virtual display. The surgeon will be able to scale, orient and position prescanned body imagery on-line in real time from any desired perspective. The clinical goal is dynamic fusing of 3-D body scan data with the actual patient in the operating room. The customized interface will permit ready, on-line access to the preoperative plan and to update measurement and analysis based on the real-time operating room data.

Practicing Mayo surgeons are committed to assisting with development, evaluation and deployment of the VRASP system. VRASP will bring to the OR all of the pre-surgical planning data and rehearsal information in synchrony with the actual patient and operation in order to optimize the effectiveness of the procedure, minimize patient morbidity, and reduce health care costs

A comprehensive software package called ANALYZE has been developed by the Biomedical Imaging Resource at Mayo Clinic which provides an integrated set of display, manipulation and measurement tools for detailed investigation and evaluation of 3-D biomedical images. This software can be used with many 3-D imaging modalities used in functional neuro-imaging, such as x-ray computed tomography, radionuclide emission tomography, and magnetic resonance imaging. The package is unique in its synergistic integration of fully interactive modules for direct display, manipulation and measurement of multi-dimensional, multi-modality image data. The inclusion of a variety of interactive display, editing and quantitative mensuration tools significantly enhances the usefulness of the software for analysis of structure-to-function relationships. Any arbitrary region/volume-of-interest can be manually specified and/or automatically segmented for numerical determination and statistical analyses of distances, areas, volumes, densities, textures, and shapes. Several original algorithms are included which improve image display efficiency and quality. Among these are versatile and powerful interactive volume rendering algorithms. These are optimized to be fast and flexible, without compromising image quality, and permitting interactive definition, manipulation and measurement of multiple rendered objects. Advanced algorithms are also included for automated segmentation using math morphology, image registration using surface matching, and tissue classification using multispectral analysis. These advanced tools can be used in combination, even recursively, and in conjunction with the rendering display methods to provide comprehensive capabilties for multi-modality image fusion - useful in many neuro-imaging applications. Interactive simulation of neurosurgery, 3-D radiotherapy planning, tumor volume response to treatment, 3-D cerebral blood flow, and anatomic mapping of cerebral function are examples of such applications. The software architecture permits systematic enhancements and upgrades which has fostered development of a readily expandable package. ANALYZE runs on standard UNIX computers without special-purpose hardware, which has facilitated implementation on a variety of popular workstation systems

We have developed a software module for general purpose voxel classification of multi-spectral data. The present version includes (1) manual tools for exploring and classifying two channel images, and (2) automated classification algorithms which oerate on images of any number of channels. The emphasis throughout the module is on speed and interactivity. We present examples of the module's operation on MRI data and discuss its capabilities and limitations.