The following are selected research projects which demonstrate the depth and breadth of our collaborative partnerships. Click on the title of the project for the full description.
The objective of this research is to develop the methodology and tools necessary to implement adaptive, multimodal human-computer interfaces that are personalized for individual users representing a full spectrum of visual capabilities. Due to the growing popularity of the Internet and other information technology systems, there is a critical need for all citizens to be empowered to access information electronically. As a result, the concept of universal accessibility has emerged in the fields of human-computer interaction and interface design. To achieve universal access to electronic information technologies, designs must overcome barriers that have been perpetuated by traditional "one-size-fits-all" philosophies in order to accommodate the disabled, the elderly, and technologically unsophisticated individuals. The specific modes to be considered for display and control interactions are: (1) speech, vocalizations, and other aural information, (2) vision, including text, images, video and computer graphics for iconic displays, and (3) haptics, including tactile information.
This three-year initiative involves research and development activities that will significantly support and drive modernization of the Medicaid program for children in the State of Georgia. The research protocol is comprised of a comprehensive ten-step process of interrelated and coordinated activities and will be pursued and carried out jointly by a team of interdisciplinary researchers and clinicians at Children's Healthcare of Atlanta and Georgia Tech. This project will also serve as a model for other states to adopt.
This project aims to improve understanding of the effectiveness of resources and guidelines available to designers and developers of "universally designed" products. Universal design is intended to enable the full utilization of mainstream products by those with disabilities In this project, it is not products themselves that are being evaluated but rather the effectiveness of print and web resources utilized by those attempting to incorporate universal design provisions into a product design. In-field and laboratory experiments will be conducted.
Security is an important and challenging issue in multi-institutional environments, because resources may be managed by a variety of systems with local security components. These components do not readily interoperate with each other, resulting in problems with managing user accounts and access control policies across systems. BIP is developing a standards-based security framework in a collaborative effort between GT (Drs. Blough and Ahamad) and Emory (Drs. Saltz, Kurc, and Post) to enhance security capabilities in federated environments. The goal is to provide a common set of tools for the specification and maintenance of user accounts and authorization policies across multiple platforms in a federated environment. This work leverages standards such as XACML for authorization policy management, Shibboleth for user attribute federation and single sign on, and SAML for authentication. The first phase of implementation is focused on instance and attribute level access control support. It supports the management and enforcement of policies defined in XACML and the development of interfaces to the access control components of external applications such as i2b2.
Systems engineers from the Health Systems Institute's Center for Pediatric Outcomes and Quality are working with Children's Healthcare of Atlanta (Children's) and Children's private practice physician community to build a Community Health Record (CHR). Through the CHR, providers will electronically share information to create a complete, integrated, longitudinal health record for each child. Having access to more complete medical information will enable providers to improve the quality of care they provide to patients. At the same time, it will streamline the information sharing process, enabling providers to achieve operational efficiencies. In addition to these benefits, the longitudinal health records in the CHR will enable researchers from Children's, Georgia Tech, Emory, and other groups to conduct innovative prevention, diagnosis, and treatment research that can dramatically effect the health of children not only in Georgia, but nationwide.
In 2004, Children's Healthcare of Atlanta (Children's), the Health Systems Institute's (HSI) Center for Interactive Systems Engineering (CISE), and the Emory University Center on Health Outcomes and Quality were awarded a grant to study health information technologies (HIT) implemented to improve patient safety and quality. This collaborative research examines the effect that key components of an electronic medical record have on operations at Children's. These components include: an inpatient Pharmacy system, electronic Medication Administration Record (eMAR), and Computerized Provider Order Entry (CPOE). Because Children's is implementing these systems using a phased approach, the research will measure the incremental impact each component has at Children's.
This research assesses and validates the use of pupillary response as a reliable and robust mechanism for assessing mental workload in the performance of tasks involving visual displays under conditions of visual impairment due to disability-induced impairment and situationally-induced impairments (SII), through the development of fast, inexpensive, and reliable statistical classification procedures that are able to discern and discriminate pupillary response measurements from healthy individuals and to detect and classify various pathologies of the eye in people with visual impairment. Many complex tasks, including those in military command and control settings, involve the use of visual displays, especially those requiring real-time interaction between a human and a computer. Individuals use the visual displays to search and locate task-relevant information in a timely and accurate manner. In order to assess a person's performance using the visual displays, and/or to assess the effectiveness of the displays, it is often useful to know whether or not cognitive activity occurs when the individual focuses on certain features of the display.
Errors due to human mistakes and software faults are unavoidable in complex, dynamic federated environments. The Biomedical Informatics Program (BIP) of the Atlanta Clinical and Translational Science Institute (ACTSI) is addressing this issue by developing a framework of tools and techniques designed to detect errors by combining domain knowledge, modeling, and software testing techniques. The framework implements support for test models, which represent constraints and dependencies associated with federated databases, and a family of testing techniques, which leverage these models to evaluate data integrity and correctness in the federated environment. A test model is a set of rules derived from (1) constraints within and among databases, (2) business processes (e.g., study protocols), (3) user-defined rules, and (4) rules based on domain knowledge. The testing techniques are model-based and will identify relevant test scenarios for the environment. The framework is presently under development using a testbed consisting of several of the ACTSI data repositories. This project is funded by BIP with additional funds from GT and Emory and led by GT (Drs. Orso and Harrold) in collaboration with Emory (Drs. Saltz and Kurc).
This innovative research is particularly compelling due to the rapid adoption of mobile computing. As mobile computing becomes more common, SII will become a significant barrier to the effective use of these systems. At the same time, researchers have yet to systematically investigate the factors that lead to SII, the effects SII have on user interactions, and the similarities between SII and DII. With the continuing expansion of mobile computing, users enjoy increased flexibility in terms of where and when they record, retrieve, and transmit information. As a result, the conditions under which computing devices are used are becoming more variable, less predictable, and in many situations less hospitable. When the working conditions, in combination with the user's tasks, result in demands that exceed the user's capabilities, the user experiences what we refer to as situationally-induced impairments (SII). Unlike disability-induced impairments (DII), SII are both dynamic and transient.
The objective of this research is to identify, develop, and test a new nurse-sensitive measure of care coordination. The research team takes the perspectives and tools of nursing, health systems engineering, and the social sciences and incorporates them into the challenge of measuring care coordination in a hospital setting. Specific aims of this research are to identify key care coordination activities carried out by nurses in the hospital and to develop indicators to measure these care coordination activities. An additional aspect will focus on testing the performance of new care coordination measure using National Quality Forum (NQF) criteria for quality performance measures.
This project explores the possibility of having the computer itself bridge the gap that exists between computer technology (all the benefits that it can provide) and a significant number of individuals suffering from severe eye conditions. This research introduces a new approach to visual compensation that is, in principle, more general than current optical methods of visual correction, since it naturally accommodates the higher-order components of the visual limitations, not addressed by current solutions. These higher-order effects have an impact on 7,000,000 Americans who have excessively irregular refractive errors, and may partially account for the 14,000,000 Americans who have severe visual conditions not correctable by glasses.
Predictive health by inferring the health condition of individuals through the analysis of their activity patterns at home
This interdisciplinary collaboration between Emory’s Center for Comprehensive Informatics and Georgia Tech’s School of Interactive Computing investigates new approaches for Infrastructure-mediated sensing (IMS) as an approach to human activity recognition, based on the idea that human activities (e.g. running a dishwasher, turning on a reading light, or walking through a doorway) can be sensed via their manifestations in an environment’s existing infrastructures (e.g. a home’s water, electrical, and HVAC infrastructures). Because of its practical, low-cost, and unobtrusive nature, infrastructure-mediated sensing offers significant promise as a general method for activity recognition. This initiative proposes to apply, e.g., water-based IMS to infer medically-meaningful activities of individuals who suffer from HIV/Cancer. We believe that by accurately tracking and logging human behavior in this context, and integrating the behavioral data into patient's clinical profile, physicians will have access to an additional layer of information from which to prescribe medication, examine behavior, modify recovery programs, etc. We leverage knowledge and facilities from physicians from the Emory medical community, and determine types of activities of daily living (ADL) information which would lead to improved decision support for physicians caring for patients, and deploy our sensor in patients home and integrate all the acquired data into the grid computing/database framework.
This study aims at developing and testing a new method that can better capture preference for multistate health profiles. The need for the study arose from the failure of the QALY (Quality-Adjusted Life Year) model in capturing preferences for multistate health profiles. As past literature shows, the conventional QALY model violates one of its required assumptions, the additive independence. These findings imply that preference between individual health states are not independent to each other. This study proposes a novel approach designed to measure preferences for multistate health profiles by looking at two consecutive health states at a time. It hypothesizes that the evaluation of a future health state is dependent or "conditioned" on the current health state.
Visualized Decision Making: Development and Application of Information Visualization Techniques to Improve Decision Quality of Nursing Home Choice
The objective of this research is to develop and test a visualization tool that will enable more effective decision making using an approach we term "visualized decision making." This approach enables efficient management of the multitudinous attributes and complex data present in many decision processes. For our research, we apply the approach to choosing a nursing home.
The primary objective of this project is to develop a new workforce health assessment model to provide employers the ability to estimate workforce health risk and assist in the determination and selection of cost-effective strategies to reduce those health risks and the associated long-term workforce health costs. To educate employers and assist them in making optimal decisions, the project proposes to develop, in partnership with Comprehensive Health Services, Inc., a predictive model. This model will support the identification and quantification of health risks associated with the workforce and will provide the basis for a proactive analysis of the cost-effectiveness of various strategies and programs, in particular, programs designed to prevent and/or reduce those risks.