zation, interpretation, visualization, and exploration of large sets of heterogeneous time-oriented clinical data that originate from clinical-trial patients. The software tools to be developed bridge the gap between the large numbers of data that need to be analyzed and summarized, and the capabilities of health-are professionals to process effectively these data. We will develop an innovative distributed service for Domain-based Abstraction, Visualization and Exploration (DAVE) of time-oriented clinical data. The service will include tools that will (l) interpret the data using terms and properties specific to the particular clinical domain and protocol (e.g., context-sensitive toxicity level definitions and predefined temporal patterns), (2) enable visualization and exploration of the data and its abstractions, and (3) support a distributed architecture combining patient databases, domain-specific knowledge bases, and local visualization and exploration modules. The tools will support several important clinical tasks, including monitoring of clinical-trial patients, patient-specific application of clinical protocols, assessment of quality of the application, and clinical research.
The technology developed has significant potential for reducing the costs in time and money, while improving the efficiency and the quality, of analysis of clinical data to support clinical therapy and quality assessment of protocol-based care. Thus, potential applications are tools for reducing the cost and duration of clinical trials in the oncology domain. Potential immediate applications exist in other medical domains in which data of chronic patients needs to be gathered and summarized over time, such as in care of diabetes and AIDS patients.
