| Abstract
|
Objective: Automating the task of planning … Objective: Automating the task of planning protocol-directed therapy requiresa computer program to take as input clinical data stored in an electronicpatient-record system, and to generate as output recommendations fortherapeutic interventions and laboratory testing that are defined bypredefined protocols. The output must be tailored for the current patientsituation and stage of protocol execution. Our goal has been to model thefunctional requirements of the therapy-planning task.Design: We constructed a computational model that includes components that(1) interpret abstract protocol specifications to construct appropriatepatient-specific treatment plans, (2) infer from time-stamped patient datahigher-level, interval-based, abstract concepts, (3) perform time-orientedqueries on a time-oriented patient database, and (4) allow for acquisition andmaintenance of protocol knowledge in a manner that facilitates efficientprocessing both by humans and computers. We have implemented these componentsin a computer system known as EON.Results: The EON architecture brings together (1) a therapy planner based ona reusable problem-solving method known as episodic skeletal-plan refinement,(2) the RESUME temporal-abstraction system, which implements theknowledge-based temporal-abstraction problem-solving method, (3) the Chronusdatabase system, which processes complex temporal queries referred to aclinical, time-oriented relational database, and (4) special-purposeknowledge-acquisition tools that are generated automatically from descriptionsof the relevant clinical domains by the PROTEGE-II system. We have evaluatedthe capabilities of the EON components by implementing T-Helper, acomputer-based patient record system that uses EON to offer advice regardingthe management of patients who have AIDS and HIV infection.Conclusion: Each of the modules that comprise the EON architecture has beendeveloped as a self-contained, reusable component. The integration of theseelements, however, leads to an aggregate component that is itself reusable forautomating the task of planning protocol-based therapy in a variety ofclinical domains. EON provides a comprehensive, yet flexible approach thatfacilitates development, maintenance, and execution of electronic knowledgebases that encode clinical protocols of substantial complexity.Objective: Automating the task of planning protocol-directed therapy requiresa computer program to take as input clinical data stored in an electronicpatient-record system, and to generate as output recommendations fortherapeutic interventions and laboratory testing that are defined bypredefined protocols. The output must be tailored for the current patientsituation and stage of protocol execution. Our goal has been to model thefunctional requirements of the therapy-planning task.Design: We constructed a computational model that includes components that(1) interpret abstract protocol specifications to construct appropriatepatient-specific treatment plans, (2) infer from time-stamped patient datahigher-level, interval-based, abstract concepts, (3) perform time-orientedqueries on a time-oriented patient database, and (4) allow for acquisition andmaintenance of protocol knowledge in a manner that facilitates efficientprocessing both by humans and computers. We have implemented these componentsin a computer system known as EON.Results: The EON architecture brings together (1) a therapy planner based ona reusable problem-solving method known as episodic skeletal-plan refinement,(2) the RESUME temporal-abstraction system, which implements theknowledge-based temporal-abstraction problem-solving method, (3) the Chronusdatabase system, which processes complex temporal queries referred to aclinical, time-oriented relational database, and (4) special-purposeknowledge-acquisition tools that are generated automatically from descriptionsof the relevant clinical domains by the PROTEGE-II system. We have evaluatedthe capabilities of the EON components by implementing T-Helper, acomputer-based patient record system that uses EON to offer advice regardingthe management of patients who have AIDS and HIV infection.Conclusion: Each of the modules that comprise the EON architecture has beendeveloped as a self-contained, reusable component. The integration of theseelements, however, leads to an aggregate component that is itself reusable forautomating the task of planning protocol-based therapy in a variety ofclinical domains. EON provides a comprehensive, yet flexible approach thatfacilitates development, maintenance, and execution of electronic knowledgebases that encode clinical protocols of substantial complexity. nical protocols of substantial complexity.
|
