The research will develop a standard definition and specifications for a clinician-oriented knowledge server (KS), to automatically assist providers in formulating context-sensitive queries of external knowledge bases. We will focus specifically upon an initial definition of a universal standard for interfacing providers'patient-care systems with any of several external knowledge bases through the KS mechanism. The standard will address modes of inquiry, types of data searches typically encountered, and the means to enhance the interaction between provider and data sources. PDQ and CANCERLIT knowledge bases will be used to develop this prototype. The research will also demonstrate the potential for using a voice-activated interface between the provider and the computer. The work will illustrate the application of a patient care system based upon voice recognition, such as the Kurzweil VoiceTM product, and the Knowledge-Server specification, for accessing data within the PDQ or CANCERLIT knowledge bases. Potential advantages and constraints of this type of interface will be identified. The work will promote and facilitate the integration of knowledge-servers and voice activated computer technologies into commercial patient care software, thereby increasing the use of medical knowledge bases such as PDQ and CANCERLIT.Awardee's statement of the potential commercial applications of the research:Potential applications include licensing standards for knowledge-server (KS) applications, preparing publications for KS standard usage and interfacing different technological platforms, and providing support to vendors of CD-ROM/on-line databases and vendors of clinical workstations and health information systems in improving design of commercial information systems (e. g. seminars, consulting services).National Cancer Institute (NCI)
