A knowledge-based method for temporal abstraction of clinical data

From Semantic Portal Wiki

{{#vardefine:category|Publication}}{{#vardefine:templatename|i.publication}}{{#vardefine:package|smwbp_instance_templates}}

Edit

Reference: {{#vardefine:pagename|a knowledge-based method for temporal abstraction of clinical data }}

1. [[]]

bibtex

{{#vardefine:pagename|A knowledge-based method for temporal abstraction of clinical data }}{{#vardefine:key| }}

abstract: This dissertation describes a reasoning framework for knowledge-based systems,that is specific to the task of abstracting higher-level concepts fromtime-stamped data, but that is independent of any particular domain. I specifythe theory underlying the framework by a logical model of time, parameters,events, and contexts: a knowledge-based temporal-abstraction theory. Thedomain-specific knowledge requirements and the semantics of the inferencestructure that I propose are well defined and can be instantiated forparticular domains. I have applied my framework to the domain of clinicalmedicine.My goal is to create, from primary time-stamped patient data, interval-basedtemporal abstractions, such as "severe anemia for 3 weeks in the context ofadministering the drug AZT," and more complex patterns, involving several suchintervals. These intervals can be used for planning interventions fordiagnostic or therapeutic reasons, for monitoring plans during execution, andfor creating high-level summaries of electronic medical records. Temporalabstractions are also helpful for explanation purposes. Finally, temporalabstractions can be a useful representation for comparing a therapy planner'srecommendation with that of the human user, when the goals in both plans can bedescribed in terms of creation, maintenance, or avoidance of certain temporalpatterns.I define a knowledge-based temporal-abstraction method that decomposes the taskof abstracting higher-level, interval-based abstractions from input data intofive subtasks. These subtasks are then solved by five separate,domain-independent, temporal-abstraction mechanisms. The temporal-abstractionmechanisms depend on four domain-specific knowledge types. The semantics ofthe four knowledge types and the role they play in each mechanism are definedformally. The knowledge needed to instantiate the temporal-abstractionmechanisms in any particular domain can be parameterized and can be acquiredfrom domain experts manually or with automated tools. I present a computer program implementing the knowledge-basedtemporal-abstraction method: RESUME. The architecture of the RESUME systemdemonstrates several computational and organizational claims with respect tothe desired use and representation of temporal-reasoning knowledge. The RESUMEsystem accepts input and returns output at all levels of abstraction; generatescontext-sensitive and controlled output; accepts and uses data out of temporalorder, modifying a view of the past or of the present, as necessary; maintainsseveral possible concurrent interpretations of the data; represents uncertaintyin time and value; and facilitates its application to additional domains byediting only the domain-specific temporal-abstraction knowledge. The temporal-abstraction knowledge is organized in the RESUME system as threeontologies (domain-specific theories of relations and properties) ofparameters, events, and interpretation contexts, respectively, in each domain.I have evaluated the RESUME system in the domains of protocol-based care,monitoring of children's growth, and therapy of insulin-dependent diabeticpatients. I have demonstrated that the knowledge required for instantiatingthe temporal-abstraction mechanisms can be acquired in a reasonable amount oftime from domain experts, can be easily maintained, and can be used forcreating application systems that solve the temporal-abstraction task in thesedomains. Understanding the knowledge required for abstracting clinical data over time isa useful undertaking. A clear specification of that knowledge, and itsrepresentation in an ontology specific to the task of abstracting concepts overtime, as was done in the architecture of the RESUME system, supports designingnew medical and other knowledge-based systems that perform temporal-reasoningtasks. The formal specification of the temporal-abstraction knowledge alsosupports acquisition of that knowledge from domain experts, maintenance of thatknowledge once acquired, reusing the problem-solving knowledge for temporalabstraction in other domains, and sharing the domain-specific knowledge withother problem solvers that might need access to the domain's temporal-reasoningknowledge.

download:

• paper:
• slides: