Automated Trend Detection with Alternate Temporal Hypotheses

We have written a prototype computer program called TrenDx for automated trend detection during process monitoring. The program uses a representation called trend templates that define disorders as typical patterns of relevant variables. These patterns consist of a partially ordered set of temporal intervals with uncertain endpoints. Bound to each temporal interval arc value constraints on real-valued functions of measurable parameters. TrenDx has been used to diagnose trends in growth patterns from examining heights, weights and other parameters of pediatric patients. As TrenDx analyzes successive data points, the program updates its hypotheses about which stage of the growth process each data point belongs to. We present an example of TrenDx reaching temporally plausible diagnoses for an actual patient with delayed growth currently being seen at Boston Children's Hospital.