Intelligent processing of ambulatory monitoring data in gestationaldiabetes

摘要

This paper describes a system to analyze self-monitoring data ofgestational diabetic patients with the final goal of obtaining a dailyassessment of their metabolic control. Our approach is based on a causalprobabilistic network that represents qualitative medical knowledgeexpressed as relations between causes and effects and theirprobabilities. The system is able to manage incomplete data and to makereasoning under uncertainty, the two most important constraints whenanalyzing ambulatory monitoring data. The prototype works with bloodglucose and ketonuria data, as well as timing and quantity deviations ofinsulin and food intakes. The outcomes provided by the system areinformation on patient transgressions of the prescribed treatment, theneed of treatment adjustments and alterations on the patient metabolism