Bayesian Network Fragments (BNFrags) provide a practical, computational methodology to encode a distributed library of computer-usable knowledge patterns for automated reasoning about aspects of homeland defense against terrorism. Multi-Entity Bayesian Networks provide a means of encoding repeated patterns and relationships in the form of BNFrags having variables that range over entities of a given type. New evidence either is matched to existing entities or triggers new entities to be hypothesized. BNFrag instances are created by replacing the variables by the names of entities in the situation. These BNFrags are combined to form situation-specific Bayesian networks (SSNs). We propose the use of MEBNs as the inferential cornerstone of a cumulative national, distributed knowledge base (KB) for homeland defense. In this paper we illustrate the use of MEBNs for these purposes with an example concerning a multi-city coordinated biowarfare attack. We show how current trends in the use of on-line reporting by health care and related facilities has the potential to enable opportunistic detection of and response to low probability, high consequence events for which it would otherwise be a practical impossibility to set up specifically directed monitoring capabilities.
KEYWORDS: Image segmentation, Radiography, Data modeling, 3D modeling, Image processing, 3D image processing, Tissues, Bone, Statistical modeling, Medical imaging
In this study we use a strong model of the center of the phalanx in hand radiographs to predict continuation of the phalanx boundary. The center phalanx is robustly segmented using conventional approaches. Estimating tangents allows us to find the minimum width of the phalanx by parallel tangent lines. This in turn predicts the phalanx 'center'. The phalanx boundaries are modeled as cubic splines, and the coronal cross-section is modeled as a hemi- ellipse. Using initial localization from the phalanx center and the above parametric models, the model of radiographic imaging is used to predict the continuation of the boundary. Results are shown for least squared error of model spline fits to continuation of the phalanx boundary anchored by initial match on the center of the projected phalanx.
KEYWORDS: Databases, Medical imaging, Human-machine interfaces, Picture Archiving and Communication System, Medical research, Radiology, Associative arrays, Information fusion, Software, Software development
Large scale feature searches of accumulated collections of medical imagery are required for multiple purposes, including clinical studies, administrative planning, epidemiology, teaching, quality improvement, and research. To perform a feature search of large collections of medical imagery, one can either search text descriptors of the imagery in the collection (usually the interpretation), or (if the imagery is in digital format) the imagery itself. At our institution, text interpretations of medical imagery are all available in our VA Hospital Information System. These are downloaded daily into an off-line computer. The text descriptors of most medical imagery are usually formatted as free text, and so require a user friendly database search tool to make searches quick and easy for any user to design and execute. We are tailoring such a database search tool (Liveview), developed by one of the authors (Karshat). To further facilitate search construction, we are constructing (from our accumulated interpretation data) a dictionary of medical and radiological terms and synonyms. If the imagery database is digital, the imagery which the search discovers is easily retrieved from the computer archive. We describe our database search user interface, with examples, and compare the efficacy of computer assisted imagery searches from a clinical text database with manual searches. Our initial work on direct feature searches of digital medical imagery is outlined.
Model-based reasoning techniques complement traditional image processing and pattern recognition to compensate for the ambiguities inherent in medical imagery data. In model- based reasoning, higher-order image features, such as inferred surface boundaries, are matched against parametric models of anatomy. The instantiated model can be used to predict approximate locations of other anatomical image features for further segmentation processing. The inferential power of this approach is limited by the accuracy with which anatomical models capture population statistics. One possible problem with a model-based approach is that it requires a large training samples to develop models. We show that selection of quasi- invariant features greatly reduces the population sample required to accurately model population distribution of features. Quasi-invariant features include certain ratios, angles and other functions of directly measurable image features that constrain each others values. When one observable is measured in the image, the likely values and locations of others are relatively constrained. For example, the ratios of lengths of phalanges has small variance across the population of all humans compared to the variance of each bone length individually. An experiment is presented on a population of 90 radiographs that supports this approach to segmentation for hand anatomy measures.
The transition from analog film-based to electronic imagery management in radiology departments and clinics requires accurate projection of storage and network capabilities. We previously developed a method of estimating from film usage, static storage requirements for the central archive systems. Planning requires us to project: (a) intermediate (e.g. magnetic disk) storage needs for local area networks and workstations serving clinical care areas, and (b) data transmission rates needed to deliver data to nodes on the network, given the expected dataflows empirically derived in our previous studies. The majority of medical imagery is currently stored on 14'' by 17'' film, each film representing about 6 Mbytes of storage at current standard digitization resolutions. With such applications, initial projections of data rates can be made using records of film usage available in most departments. However, it is also necessary to perform a survey of film users to determine usage of new imagery modalities and comparison imagery requirements in the areas to be served by the network.
KEYWORDS: Picture Archiving and Communication System, Computed tomography, Radiology, Medical imaging, Magnetic resonance imaging, Databases, Imaging systems, Surgery, Chest, Bone
PACS offers new capabilities, but presents some new limitations as well. We reviewed radiology practice
at four medical centers. For interpretation of new images, most radiologists felt that the most recent
examination for comparison would suffice, except for chest radiographs, for which three prior exams were
frequently needed. All of the centers used alternators to review images, with from 5-14 separate reading
areas. PACS capabilities should include organization and delivery of cases for interpretation and
consultation to the appropriate area, pnoritization of cases, and rapid availability of inpatient and clinic
patient data.
For PACS to be clinically desirable, transmiued imagery must be optimized for the various clinical and
research tasks. There are several processes that potentially degrade the diagnostic utility of the iransmitted
imagery. We examined our CommView PACS system at San Francisco VA Medical Center (SFVAMC)
to determine if imagery was altered from the original during passage into and through PACS, and whether
the alteration had any discernable effect on current or projected clinical utility of the imagery. Related
operational considerations were also examined. This limited review of our CommView system indicates that
the attenuation of the image data by frame-grabbing and inability of the workstation to work with processed
rather than formatted data from computed radiography limits the potential capabilities of the workstation.
The reversible compression algorithm for imagery archival functioned well, but serious high frequency
aliasing was introduced in imagery reduced for transmission to remote viewing stations.
Hand radiographs provide a valuable index of disease in arthritis and other generalized diseases such as secondary hyperparathyroidism and osteoporosis. Measures such as cortical volume intercortical width average and periarticular demineralization provide diagnostic indicators for these diseases. However visual analysis of hand radiographs is not quantitative and is compromised by both interobserver and intraobserver variation. Automation of these measures would provide repeatable comparable quantities to assist in diagnosis and disease and therapy monitoring. The computer calculations to perform these measures are straightforward. The key problem is automatic segmentation of the hand anatomy that is recognizing the pixels that correspond to specific imaged bones and joints. Our approach incorporates computer-represented hand models in addition to more traditional image processing algorithms. We describe our techniques for using a combination of predictive models and image processing evidence to automatically fmd bone and tissue boundaries and identify specific bone and joints. 2. COMPUTING ARTHRITIS MEASURES Digital scanners and radiograph digitizers make the radiograph available as a data source for computer algorithms that analyze medical imagery. This is significant because radiographs comprise more than 80 of all medical imagery at this time and they are considerably quicker and less costly than other digital modalities such as CT and Mill. Quantitative measures from digital radiographs can aid physicians in diagnosis tracking disease progress and in therapy planning and evaluation. We have begun studying diagnostic measures in arthritis
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.