Computational Narrative

We are currently studying how data analysts use Jupyter Notebooks to track their work and share it with others, with a focus on how they craft computational narratives. We are also developing extensions to Jupyter that make it easier to track and share analytical process.

Clinical Summaries

Clinicians often need to get up-to-speed on a patient's case by reviewing their medical record, but relevant information is fragmented across the record. This project is looking at ways to reorganize and summarize records so that clinicians can quickly understand to patient's current condition and past medical care.



Activity Resumption

Interruptions - by colleagues, computer alerts, and even self interruption - are a staple of knowledge work. They keep us up to date with relevant information and help us prioritize. However, they can also make us forget what we were doing before being interrupted. Images have been shown to evoke rich episodic memory about past events, bringing to mind not only what was happening, but also why and how it was occuring. We are studying how people resume interrupted work and crafting visualizations of past computer activity that help them recall the mental context of their suspended activities. Keep track of the project at the project website.

[j1] [c2] [w2] [w1] [p2] [p1]

Medication Order Entry

Electronic Health Records (EHRs) have greatly increased the usability and portability of health information by storing it as structured data. However, today's EHRs rely on an aging paradigm of windows, tables, and drop-down menus to collect and display this information. This paradigm separates structured data from the valauble unstructured commentary of clinical notes. The goal of ActiveNotes, a Veterans Medical Research Foundation project, was to develop interactive progress notes that unify entry, access, and retrieval of structured and unstructured health information. My contributions to the project were the design and evaluation of an interface for free text entry of medication orders within a note that would pre-populate a set of clinical orders.