SWEETInfo

How can clinical investigators discover scientifically relevant temporal patterns, such as correlations, trends, and outcomes, hidden among vast amounts of time-oriented health data? There are few, if any, widely available, ready-to-use tools for the temporal exploration of individual- and population-level data. As a result, investigators can miss crucial findings within data collected from electronic health records and patient registries.

To address this need, we have developed a tool called Semantic Web-Enabled Exploration of Temporal Information (SWEETInfo) that allows investigators to interactively import sets of patient data, view their temporal dimensions on a timeline, and perform data transformations. SWEETInfo is built as an open-source, Web-based architecture that combines Semantic Web standards, such as OWL and SWRL, with advanced Web development software, such as the Google Web Toolkit. Our NIH-supported work has resulted in a novel extendable knowledge-based infrastructure for the representation, visualization, querying, and aggregation of time-oriented patient data.

User interaction with SWEETInfo occurs through the creation of (1) data mappings that transform time-stamped data in relational databases into ontology-based temporal information and (2) data pipelines based on operations for filtering, grouping, and visualization. Both of these specifications are saved within an OWL ontology and can be shared with and adapted by other users via the Web. We have initially designed SWEETInfo to meet the needs of our collaborators in HIV drug resistance and breast cancer care. We are undertaking extensive testing to evaluate the usability of our novel intelligent user interface design and its ability to support time-oriented data exploration and transformations for cohort identification, trend analysis, and outcomes studies.