Temporal Summaries: Supporting Temporal Categorical Searching, Aggregation and Comparison

Abstract

Abstract-When analyzing thousands of event histories, analysts often want to see the events as an aggregate to detect insights and generate new hypotheses about the data. An analysis tool must emphasize both the prevalence and the temporal ordering of these events. Additionally, the analysis tool must also support flexible comparisons to allow analysts to gather visual evidence. In a previsous work, we introduced align, rank, and filter (ARF) to accentuate temporal ordering. In this paper, we present temporal summaries, an interactive visualization technique that highlights the prevalence of event occurrences. Temporal summaries dynamically aggregate events in multiple granularities (year, month, week, day, hour, etc.) for the purpose of spotting trends over time and comparing several groups of records. They provide affordances for analysts to perform temporal range filters. We demonstrate the applicability of this approach in two extensive case studies with analysts who applied temporal summaries to search, filter, and look for patterns in electronic health records and academic records.

Extracting Insights from Electronic Health Records: Case Studies, a Visual Analytics Process Model, and Design Recommendations

“…In fact, in cases where heavy exploration is needed, as in the early adoption heparin-induced thrombocytopenia case study, we notice that the physicians keep their eyes fixed on the temporal summary as a variety of Filters are applied. When we work on the mature heparin-induced thrombocytopenia study (25) , a different set of physicians also show the same tendency to focus on the temporal summaries as the data is being manipulated. The physicians often ask questions out loud on how a particular manipulation (e.g.…”

“…The physicians often ask questions out loud on how a particular manipulation (e.g. Align, (25) . Over all, temporal summaries provide indispensable guidance to the physicians.…”

“…Temporal summaries (25) are histograms of events over time. Additionally, analysts can change number of events to number of EHRs or number of events per EHR over time.…”

“…In this phase, Lifelines2 was used for a number of different analysis tasks: (1) Replicating a study (26) that investigates the relationship between day light savings time change and heart attack incidents using clinical data. (2) Performing a follow-up study on heparin-induced thrombocytopenia in ICU patients (25) , (3) studying hospital room transfer patterns as a measure for quality assurance (two case studies, one for the Bounce-Back patterns, and the other for the Step-Up patterns), and (4) studying the impact on patient care due to a change in protocol that governs when Bi-level Positive Airway Pressure (BiPAP) is applied. Table 1 shows the statistics of selected datasets All of our case studies were selected and initiated by our collaborators (regardless of phase).…”

Visual Exploration of Time-Oriented Patient Data for Chronic Diseases: Design Study and Evaluation

“…However, these designs cater more to different tasks than open-ended exploration. In the surroundings of this work we also find the problem domains of querying patient databases (e.g., LifeLines2 [34], Similan [35]) and exploring patient cohorts (e.g., CareGiver [5], Dare [7], Gravi++ [14], TimeRider [27], VISITORS [16]). …”