For Fusion '98 Conference Proceedings
Use of Biometrics and Biomedical Imaging in Support of Battlefield Diagnosis Joyce D. Williams Lockheed Martin Advanced Technology Laboratories 1 Federal Street, A&E 3W Camden, NJ 08102 Abstract Infrared imaging has been proven both reliable and useful in support of many facets of military operations. Systems employing this technology range from target detection, identification, and classification to medical diagnosis for chemical and gaseous exposure. There are few systems that support both biometrics and biomedical imaging in a convenient, centralized location (i.e., on the battlefield). On the battlefield, rapid medical diagnosis and treatment is paramount. The concept behind the design of this portable sensor is to develop a system capable of incorporating both biometrics and biomedical imaging to provide a complete medical profile of a warfigher on the battlefield. A system of this type would help ensure that timely and accurate medical diagnosis and attention was being provided to a warfigher for his unique profile. It would also serve as a mechanism for identification of injured. Keywords: biometrics, biomedical sensors, biomedical imaging, battlefield diagnosis 1. Introduction On the battlefield, rapid medical diagnosis and treatment is paramount. Approximately 90 percent of combat deaths occur in the zone of close combat before medical or surgical intervention is received. In addition, fratricide continues at casualty rates as high as 20 to 30 percent [1]. Casualty location is a continuing battlefield problem. Realistic, peacetime combat medical and surgical training is minimal. Medical Theater-of-War communications are archaic and non-functional [1]. This system is an attempt to provide a complete medical profile and casualty assessment of warfighers on the battlefield. This profile and assessment would include information such as, prior medical history, patient ID, and current patient status. This type of system would help ensure that timely and accurate medical diagnosis and treatment was being provided to a warfigher for his unique profile and medical history. 2. Design Concept The goal of the system design is to have a system capable of supporting biometrics and biomedical data interpretation and analysis while on the battlefield. Figure 1 depicts the possible interaction of system components. The warfighters biometrics would be input (off-line) into a biometrics/ biomedical database(s) before the warfighters were deployed to the battlefield. The database(s) would contain the patient ID, medical history, retina scan, vein ID, and fingerprints, as well as traditional measurements such as blood pressure, temperature, etc. The sensor suite would be comprised of CCD cameras, infrared imagers, biosensors, gas detectors, and traditional equipment used to measure blood pressure, temperature, heart rate, etc. If a warfighers' identification information were found in the database(s), that information would be processed by a series of algorithms in the fusion and medical processing center. If a
Biometrics Data (off-line before combat) Database(s) 3. Fusion of Biometrics and Biomedical Imaging Data Sensor Suite Display & Transmitter Medical Fusion & Processing Center Medic Additional Sources Figure 1. Biometrics/Biomedical System Design warfighters' identification information were not found in the database(s), the medic would have the capability of inputting that data into the system directly while the warfighters are on the battlefield. This information would be collected, assessed, and input into the fusion and processing center via a display and keyboard. The medic would interface with additional data sources (i.e., other medics, medical references) as necessary. In our experiences processing hyperspectral and multispectral imaging data [2], we encountered issues that will be relevant in the fusion of medical images such as image alignment, where the images are taken with different sensors and at different angles. We were able to use commercially available software to do preliminary analysis but found the algorithms to be lacking when it came to the more detailed identification and classification. The fusion and processing center would contain a series of algorithms [2] capable of interpreting and analyzing parameters for known adversities that warfighters are likely to encounter on a battlefield (i.e., gunshot wounds, chemical burns, exposure to poisonous gases, etc.). These parameters would come from either the database(s) and/or the medic. Medical interpretation, diagnosis, and advice would be displayed to the medic in the field as well as alerting the command post of the type of injury or casualty in-route. This is analogous to medical emergency response teams contacting hospitals while in-route today. Collected data would be transmitted to awaiting receivers for further interpretations and medical advice. This information transfer would serve as a vehicle for collecting real-world data to be used to better prepare warfighters and medics for future training activities. Fusion of medical data from a variety of sources is not a new concept. Fusion is beneficial because it can provide additional insight for complicated diagnosis. As an example, consider the use of Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) which are used daily in procedures around the world. The drawback of using these systems separately is that the CT is good for dense objects such as bones, where the MRI's usefulness is in imagery of soft tissues. There are many complex injuries where a biomedical system that incorporated both of these system-specific attributes would be useful (e.g., spinal cord injuries). The concept and reality of being able to accomplish this feat on the battlefield is a new and challenging one. From a biometrics perspective, the proposed system would include but not be limited to the following biometrics: medical history, retina scan, vein ID, and fingerprints. The system should have the flexibility to support any or all of the following: trauma injuries, burns, and exposure to halogens, chemicals, and/or diseases, etc. These parameters could be modified to support a variety of other domain specific areas. Medical diagnosis on the battlefield has to take in to account certain conditions that are usually not present or experienced in controlled environments such as hospitals, laboratories, and/or clinics. Conditions such as: Actual image fusion for real-time diagnosis, assessment and/or treatment on the battlefield, Adverse atmospheric conditions (e.g., night, dense cloud coverage, exposure to toxic chemicals and vapors), and Limited accessibility and/or availability of medical references and resources (e.g, power, space) 4. Other Areas to Benefit This technology is not limited to military use, and with modifications could be used to support other specific domains, such as law enforcement, public services, hospitals, and airports. Examples of viable areas of this technology insertion into
law enforcement and public services are represented in Table 1. Table 1 - Areas that would benefit from biometrics and biomedical sensor use Law enforcement Police departments Crime scenes (e.g., sex offenders, violent criminals, missing and exploited children, etc.) Forensic science Public services Department of Social Services Schools Fire department Security at highrisk terrorist areas Law enforcement is already using highly advanced equipment for analyzing data at the crime scene. Using a biometrics and biomedical sensors in conjunction with their other equipment would provide officials on the scene with a timely method of data processing, thus reducing data and crime scene contamination. Other non-military applications include mine detection, search and valuable commercial applications include predictive and preventive maintenance of roofs, electrical distribution systems, production processes and other equipment conditions monitoring are discussed in [3,6]. 5. Advantages and Disadvantages As with most advance technologies, a system of this type would not come without advantages and disadvantages. Table 2 lists some of the most notable obstacles that would have to be overcome. However, vast majorities of the system components are already in mature stages of development and use today. The challenge would be in the integration and testing of these components in a centralized system. 6. Summary Advanced tools for medical diagnosis are changing the way physicians work with much more successful payoffs. There have been remarkable advances made in imaging technologies, procedures, and equipment. As Table 2 - Advantages and Disadvantages of Biometrics/Biomedical Sensors Advantages Positive identification of warfighter if "dogtag" is lost or missing Introduces additional medical data and imagery not normally found on the battlefield Increases the odds of fast, reliable treatment to warfighter Useable day, night, or in adverse weather conditions Portable, lightweight, batteryoperated Disadvantages Confidentiality issue of the information in the database Accuracy of the information in the database Requires highly trained technologist to interpret medical image datasets [4] What to do if trauma or exposure characteristics are not in the database System maintenance impressive as these payoffs are, doctors and researchers expect to reap even greater rewards as imaging technologies improve [6]. Hardware, software, communications, and training, as they are today, may need to be modified to support the integration of these technologies. The biggest rewards will come in a reduction in the time required to process data and better image quality. Issues such as whether the biomedical imaging equipment would be hand-held or helmet-mounted also need to be addressed. Medical personnel using this technology on the battlefield would need to be trained on the system what to do if data is not available or there is a system malfunction. Incorporating biometrics and biomedical data into a single, portable sensor will open new and additional avenues for research and development not only for military but other broad areas that require any type of human diagnosis and treatment is required.
7. References [1] Defense Advanced Research Project Agency (DARPA) Defense Sciences Office (DSO), Combat Casualty Care, http://www. sainc.com/arpa/combat/index.htm [2] J. Williams, Hyperspectral Imagining Data Analysis Report, Lockheed Martin Internal Work Authorization, 31-32, Jan. 1998. [3] J.E. Davidson, Sr., B.J. Anderesen, M Strojnik, Infrared Imaging-based Combat Casualty Care System, Proceedings of SPIE, Infrared Technology and Applications XXIII, April 20-25, 1997, Orlando, FL. [4] E.A. Hoffman, R. Stahlberg, J. Cook- Granrith, S. Chang, N. D'Souza, J. Reinhardt, Expanding the Availability of Advanced Image Display and Quantitative Medical Image Analysis Through Web-based Tutorials and Interactive On-line Consultation, American Medical Informatics Association (AMIA) Proceedings, May 28-31, 1997. [5] The Whitaker Foundation, http://www. whitaker.org/94_annual_report/over.html. [6] J.L. Miller, H. Duvoisin III, Wiltsey, B.J. Anderesen, M. Strojnik, Helmet-mounted Uncooled FPA Camera for Buried Object Detection, Proceedings of SPIE, Infrared Technology and Applications XXIII, April 20-25, 1997, Orlando, FL.