TraumaSCAN-Web is a platform-independent decision support system that assists with the diagnosis of penetrating trauma to the chest and abdomen. The system assesses anatomic structure involvement in thoraco-abdominal gunshot and stab injuries and also evaluates medical conditions that may arise as a result of penetrating trauma, such as pneumothoraces, hemothoraces, and cardiac tamponades.
If your computer does not already have Java Web Start, click here to download it. To try out TraumaSCAN-Web, please read these instructions on how to evaluate a trauma case using the system. When you are done with the instructions, you have a number of options for running TraumaSCAN-Web:
Known problems: On the Mac G4 platform, certain ATI Rage graphics cards don't work well with the application. The torso location clicked on differs from the location at which a wound marker appears. This problem does not seem to appear on other platforms.
Penetrating trauma research has shown that early administration of expert care to patients would considerably reduce the number of deaths due to these injuries. Assessment of penetrating trauma requires integrating knowledge of human anatomy, physiology, and physical manifestations of injury/patient findings. Tools or knowledge aids that reinforce understanding of these factors and their relationships for a given patient presenting with penetrating injury, may be of value in enhancing patient care.
Earlier work on this problem (1995-1999) culminated in the development of TraumaSCAN, a computer system that combines 3D reasoning about anatomy/geometry with probabilistic reasoning based on Bayesian networks to predict the consequences of penetrating trauma to the chest and abdomen. TraumaSCAN makes use of the Silicon Graphics (SGI) platform's graphics libraries (GL). It utilizes proprietary 3D modeling software (Jack) developed at the University of Pennsylvania, and runs only on the SGI platform, limiting its potential for dissemination.
Advances in graphics hardware, etc., now make it possible to develop interactive applications that perform complex 3D computations on a desktop PC or laptop. This has made it possible to develop a web-based version of TraumaSCAN, TraumaSCAN-Web. TraumaSCAN-Web models and is able to evaluate more thoraco-abdominal penetrating trauma conditions than TraumaSCAN. It meets our goal of making a medical decision support system available to healthcare providers unused to treating penetrating trauma cases, and healthcare providers around the world that have internet and browser access, but may not have access to expensive technology such as CT-scanners.
We have developed TraumaSCAN-Web using Sun's Java3D scenegraph API; which allows us to create 3D applets that can be deployed on Windows, MacOS X, Unix, and Linux platforms. For probabilistic reasoning about injury consequences, TraumaSCAN-Web makes use of UCLA's SamIam Bayesian network API.
The TraumaSCAN-Web project is funded by the National Library of Medicine (NLM); grant number 1 RO1 LM07167.
For more information on TraumaSCAN-Web and TraumaSCAN, see the following publications:
Ogunyemi O. Methods for reasoning from geometry about anatomic structures injured by penetrating trauma. Journal of Biomedical Informatics, 39 (2006): 389-400.
Matheny M, Ogunyemi O, Rice P, Carke J. Evaluating the Discriminatory Power of a Computer-based System for Assessing Penetrating Trauma on Retrospective Multi-Center Data. Proc AMIA Annual Fall Symposium; 2005; pp. 500-504.
Ogunyemi O, Clarke JR, Ash N, Webber B. Combining Geometric and Probabilistic Reasoning for Computer-Based Penetrating Trauma Assessment. Journal of the American Medical Informatics Association. 2002 May-Jun;9(3):273-82.
Ogunyemi O, Clarke JR, Webber B, Badler N. TraumaSCAN: Assessing Penetrating Trauma with Geometric and Probabilistic Reasoning. Proceedings of the American Medical Informatics Association Annual Fall Symposium; 2000 November; Los Angeles, California; Philadelphia: Hanley & Belfus; 2000. pp. 620-624. [Full text]
Ogunyemi O, Clarke JR, and Webber B. Using Bayesian networks for diagnostic reasoning in penetrating injury assessment. Proceedings of the Thirteenth Annual IEEE Symposium on Computer-Based Medical Systems; Houston, Texas; IEEE Computer Society Press; 2000. pp. 115-120.
Ogunyemi OI. TraumaSCAN: Assessing Penetrating Trauma with Abductive and Geometric Reasoning. Ph.D. Thesis. Philadelphia (PA): University of Pennsylvania, 1999.
Ogunyemi O, Webber B, and Clarke JR. Probabilistically predicting penetrating injury for decision support. Proceedings of the Eleventh Annual IEEE Symposium on Computer-Based Medical Systems; 1998 Lubbock, Texas; IEEE Computer Society Press; 1998. pp. 44-49. [Full text]
Ogunyemi O, Webber B, and Clarke JR. Probabilistic predictions of penetrating injury to anatomic structures. In: Masys D, editor. Proceedings of the American Medical Informatics Association Annual Fall Symposium; 1997 October; Nashville, Tennessee; Philadelphia: Hanley & Belfus; 1997. pp. 714-718. [Full text]
Ogunyemi O, Kaye J, Webber B, Clarke JR. Generating penetration path hypotheses for decision support in multiple trauma. In: Gardner R, editor. Proceedings of the Nineteenth Annual Symposium on Computer Applications in Medical Care; 1995 October; New Orleans, Louisiana. Philadelphia: Hanley & Belfus; 1995. pp. 42-46. [Full text]