David G. Heath

Skeletal 3-D CT: advantages of volume rendering over surface rendering.

Abstract Both surface rendering and volume rendering have been extensively applied to CT data for 3-D visualization of skeletal pathology. This review illustrates potential limitations of each technique by directly comparing 3-D images of bone pathology created using volume rendering and surface rendering. Surface renderings show gross 3-D relationships most effectively, but suffer from more stairstep artifacts and fail to effectively display lesions hidden behind overlying bone or located beneath the bone cortex. Volume-rendering algorithms effectively show subcortical lesions, minimally displaced fractures, and hidden areas of interest with few artifacts. Volume algorithms show 3-D relationships with varying degrees of success depending on the degree of surface shading and opacity. While surface rendering creates more three-dimensionally realistic images of the bone surface, it may be of limited clinical utility due to numerous artifacts and the inability to show subcortical pathology. Volume rendering is a flexible 3-D technique that effectively displays a variety of skeletal pathology with few artifacts.

Surgical planning for liver resection

Surgical resection is the cornerstone of curative therapy for primary and metastatic liver tumors. For best results, the surgeon must know the location of all hepatic tumor nodules relative to the major vessels that define the livers surgical anatomy. Computed tomography is very sensitive for detecting liver tumors, but its planar slices do not fully address the three-dimensional nature of this surgical problem. We have developed a technique using volume rendering of computed tomography data that provides a preoperative 3D map of the liver showing tumor location relative to key blood vessels. This technique also has important implications for emerging, minimally invasive therapies.

Dual-phase spiral CT angiography with volumetric 3D rendering for preoperative liver transplant evaluation: preliminary observations.

PURPOSE The goal of our study was to determine whether dual-phase spiral CT angiography with 3D volume rendering could be used for preoperative evaluation and patient selection for orthotopic liver transplantation candidates. METHOD Fifty consecutive potential candidates for liver transplantation were evaluated with dual-phase spiral CT with 3D volume rendering. Intravenous contrast medium was administered as bolus peripheral injection at 3 ml/s. The protocol consisted of a contrast-enhanced dual-phase spiral CT (arterial phase acquisition at 30 s after initiation of contrast medium injection followed by portal venous phase beginning at 60 s) with scan parameters of 0.75 s gantry rotation speed, 3 mm collimation, 5 to 6 mm/s table speed, and reconstruction at 1 mm intervals for arterial-phase images and 3 mm collimation for portal venous-phase studies (Siemens Plus 4 scanner; Siemens Medical Systems, Iselin, NJ, U.S.A.). All scan information was sent to a free-standing workstation (Silicon Graphics Onyx or Infinite Reality, Mountain View, CA, U.S.A.) for interactive real-time 3D volume rendering using a customized version of the Volren volume renderer (Silicon Graphics; Advanced Imaging Laboratory, Johns Hopkins Medical Institutions, Baltimore, MD, U.S.A.). The arterial phase was used to create vascular maps of the celiac axis including the origin(s) of the hepatic artery and origin of the superior mesenteric artery. The portal phase was used to define portal venous patency as well as the hepatic venous anatomy. All images were analyzed for vascular patency, shunting, or collateralization as well as the status of the underlying liver (i.e., liver size, cirrhosis, tumor, etc.). RESULTS All 50 studies were successfully completed without complication. The 3D CT angiograms defined key arterial and venous structures including origin(s) of the hepatic artery, portal vein and/or superior mesenteric vein thrombosis, cavernous transformation of the portal vein, and/or other collateral vasculature. Ten patients (20%) demonstrated anomalous anatomy at the origin(s) of the hepatic artery. Portal vein thrombosis with cavernous transformation of the portal vein was shown in six patients, and there were three cases of partial venous thrombosis. Underlying liver tumors as well as parenchymal liver disease were well defined. Hepatic masses were found in five patients. Masses were pathologically proven as hepatocellular carcinoma (n = 1), giant cavernous hemangioma (n = 1), hepatic adenoma (n = 1), and focal nodular hyperplasia (n = 2). CONCLUSION Preliminary results suggest that dual-phase spiral CT with CT angiography can provide a comprehensive preoperative liver transplant evaluation, supplying the necessary information for patient selection and surgical planning. As a single, minimally invasive examination, this should significantly impact patient care by minimizing procedures and avoiding potential complications.

Abdominal image segmentation using three-dimensional deformable models

RATIONALE AND OBJECTIVES The authors develop a three-dimensional (3-D) deformable surface model-based segmentation scheme for abdominal computed tomography (CT) image segmentation. METHODS A parameterized 3-D surface model was developed to represent the human abdominal organs. An energy function defined on the direction of the image gradient and the surface normal of the deformable model was introduced to measure the match between the model and image data. A conjugate gradient algorithm was adapted to the minimization of the energy function. RESULTS Test results for synthetic images showed that the incorporation of surface directional information improved the results over those using only the magnitude of the image gradient. The algorithm was tested on 21 CT datasets. Of the 21 cases tested, 11 were evaluated visually by a radiologist and the results were judged to be without noticeable error. The other 10 were evaluated over a distance function. The average distance was less than 1 voxel. CONCLUSIONS The deformable model-based segmentation scheme produces robust and acceptable outputs on abdominal CT images.

Best-case results for nearest-neighbor learning

Proposes a theoretical model for analysis of classification methods, in which the teacher knows the classification algorithm and chooses examples in the best way possible. The authors apply this model using the nearest-neighbor learning algorithm, and develop upper and lower bounds on sample complexity for several different concept classes. For some concept classes, the sample complexity turns out to be exponential even using this best-case model, which implies that the concept class is inherently difficult for the NN algorithm. The authors identify several geometric properties that make learning certain concepts relatively easy. Finally the authors discuss the relation of their work to helpful teacher models, its application to decision tree learning algorithms, and some of its implications for experimental work. >

Multidetector-row computed tomography with three-dimensional volume rendering of pancreatic cancer: a complete preoperative staging tool using computed tomography angiography and volume-rendered cholangiopancreatography.

&NA; Volume rendering, a postprocessing computer algorithm that creates three‐dimensional (3D) displays from computed tomography (CT) datasets, can create 3D cholangiographic images (volume‐rendered cholangiopancreatography, or VRCP) from intravenous contrast‐enhanced abdominal CT datasets without the use of a biliary contrast agent. This article illustrates the utility of VRCP in the setting of biliary obstruction due to pancreatic cancer. The 3D renderings of the intra‐ and extrahepatic biliary tree provide valuable information for planning biliary drainage, including the location and length of the obstruction as well as the relationship of intrahepatic ducts to liver metastases. Index Terms: cholangiopancreatography, multidetector‐row computed tomography, pancreatic cancer, three‐dimensional computed tomography (3D CT), volume rendering

Automated Multidetector Row CT Dataset Segmentation with an Interactive Watershed Transform (IWT) Algorithm: Part 2—Body CT Angiographic and Orthopedic Applications

The preceding manuscript describes the principles behind the Interactive Watershed Transform (IWT) segmentation tool. The purpose of this manuscript is to illustrate the clinical utility of this editing technique for body multidetector row computed tomography (MDCT) imaging. A series of cases demonstrates clinical applications where automated segmentation of skeletal structures with IWT is most useful. Both CT angiography and orthopedic applications are presented.

Chapter 18 Committees of decision trees

Abstract Many intelligent systems are designed to sift through a mass of evidence and arrive at a decision. Certain pieces of evidence may be given more weight than others, and this may affect the final decision significantly. When more than one intelligent agent is available to make a decision, we can form a committee of experts. By combining the different opinions of these experts, the committee approach can sometimes outperform any individual expert. In this paper, we show how to exploit randomized learning algorithms in order to develop committees of experts. By using the majority vote of these experts to make decisions, we are able to improve the performance of the original learning algorithm. More precisely, we have developed a randomized decision tree induction algorithm, which generates different decision trees every time it is run. Each tree represents a different expert decision-maker. We combine these trees using a majority voting scheme in order to overcome small errors that appear in individual trees. We have tested our idea with several real data sets, and found that accuracy consistently improved when compared to the decision made by a single expert. We have developed some analytical results that explain why this effect occurs. Our experiments also show that the majority voting technique outperforms at least some alternative strategies for exploiting randomization.

Automated Multidetector Row CT Dataset Segmentation with an Interactive Watershed Transform (IWT) Algorithm: Part 1. Understanding the IWT Technique

Segmentation of volumetric computed tomography (CT) datasets facilitates evaluation of 3D CT angiography renderings, particularly with maximum intensity projection displays. This manuscript describes a novel automated bone editing program that uses an interactive watershed transform (IWT) technique to rapidly extract the skeletal structures from the volume. Advantages of this tool include efficient segmentation of large datasets with minimal need for correction. In the first of this two-part series, the principles of the IWT technique are reviewed, followed by a discussion of clinical utility based on our experience.

The complexity of finding minimal Voronoi covers with applications to machine learning

Abstract Our goal in this paper is to examine the application of Voronoi diagrams, a fundamental concept of computational geometry, to the nearest neighbor algorithm used in machine learning. We consider the question “Given a planar polygonal tessellation T and an integer k, is there a set of k points whose Voronoi diagram contains every edge in T?” We show that this question is NP-hard. We encountered this problem while studying a learning model in which we seek the minimum sized set of training examples needed to teach a given geometric concept to a nearest neighbor learning program. That is, given a concept which can be described by a planar tessellation, we are seeking to construct the smallest set of points whose Voronoi diagram is consistent with the given tessellation. In a sense, this question captures the difficulty of teaching the nearest neighbor algorithm a simple structure, using a minimal number of examples. In addition, we consider the natural inverse to the problem of computing Voronoi diagrams. Given a planar polygonal tessellation T, we describe an algorithm to find a set of points whose Voronoi diagram is T, if such a set exists.