A representation for gaining insight into clinical decision models
From Semantic Portal Wiki
{{#vardefine:category|Publication}}{{#vardefine:templatename|i.publication}}{{#vardefine:package|smwbp_instance_templates}}
| Edit |
Reference: {{#vardefine:pagename|a representation for gaining insight into clinical decision models }}
- [[]]
bibtex
{{#vardefine:pagename|A representation for gaining insight into clinical decision models }}{{#vardefine:key| }}
abstract: For many medical domains uncertainty and patient preferences are important components of decision making. Decision theory is useful as a representation for such medical models in computer decision aids, but the methodology has typically had poor performance in the areas of explanation and user interface. The additional representation of probabilities and utilities as random variables serves to provide a framework for graphical and text insight into complicated decision models. The approach allows for efficient customization of a generic model that describes the general patient population of interest to a patient-specific model. Monte Carlo simulation is used to calculate the expected value of information and sensitivity for each model variable, thus providing a metric for deciding what to emphasize in the graphics and text summary. The computer-generated explanation includes variables that are sensitive with respect to the decision or that deviate significantly from what is typically observed. These techniques serve to keep the assessment and explanation of the patient's decision model concise, allowing the user to focus on the most important aspects for that patient.
download:
- paper:
- slides:
| Abstract | For many medical domains uncertainty and p … For many medical domains uncertainty and patient preferences are important components of decision making. Decision theory is useful as a representation for such medical models in computer decision aids, but the methodology has typically had poor performance in the areas of explanation and user interface. The additional representation of probabilities and utilities as random variables serves to provide a framework for graphical and text insight into complicated decision models. The approach allows for efficient customization of a generic model that describes the general patient population of interest to a patient-specific model. Monte Carlo simulation is used to calculate the expected value of information and sensitivity for each model variable, thus providing a metric for deciding what to emphasize in the graphics and text summary. The computer-generated explanation includes variables that are sensitive with respect to the decision or that deviate significantly from what is typically observed. These techniques serve to keep the assessment and explanation of the patient's decision model concise, allowing the user to focus on the most important aspects for that patient. e most important aspects for that patient. |
| Author | Holly Brügge Jimison + |
| Bibtype | techreport + |
| Institution | Knowledge Systems, AI Laboratory + |
| Key | KSL-88-75 + |
| Month | November + |
| Note | 5 pages. + |
| Number | KSL-88-75 + |
| Tag | Computer science + |
| Title | A Representation for Gaining Insight into Clinical Decision Models + |
| Tr id | KSL-88-75 + |
| Year | 1988 + |
