Intelligent Analysis of Clinical Time Series by Combining Structural Filtering and Temporal Abstractions

This paper describes the application of Intelligent Data Analysis techniques for extracting information on trends and cycles of time series coming from home monitoring of diabetic patients. In particular, we propose the combination of structural Time Series analysis and Temporal Abstractions for the interpretation of longitudinal Blood Glucose measurements. First, the measured time series is analyzed by using a novel Bayesian technique for structural filtering; second, the results obtained are post-processed using Temporal Abstractions, in order to extract knowledge that can be exploited "at the point of use" from physicians. The proposed data analysis procedure can be viewed as a typical Intelligent Data Analysis process applied to time-varying data: Background Knowledge is exploited in each step of the analysis, and the final result is a meaningful, abstract description of the complex process at hand. The work here described is part of a web-based telemedicine system for the management of Insulin Dependent Diabetes Mellitus patients, developed within the EU-funded project called T-IDDM.