Research Focus

Research Area #1: Facilitating clinical data capture and integration

While clinical data have great potential to support decision making as well as clinical and translational research, the current methods for capturing such data have been suboptimal. Clinical data are often captured in an inefficient and inaccurate manner due to poor computer system design, such as unsatisfactory usability of computer user interfaces and failure to consider clinical workflow. In this research area, we aim to address these issues by designing and deploying innovative solutions derived from scientific methods in the field of human-computer interaction, computer-supported cooperative work, data mining, and visual analytics. For example, in an Electronic Health Record (EHR) system, we study users’ system use behaviors and working patterns based on system logs to generate insights of working performance and find potential bottlenecks, which can bring beneficial impact on healthcare efficiency, cost, and patient experience.

Research Area #2: Enabling the retrieval of clinical notes

Although structured clinical data are generally more preferred as they are standardized and easy to compute, unstructured data are still pervasively used because of their flexibility and expressivity. Unfortunately, unstructured clinical data such as progress notes and discharge summaries are frequently unsearchable in data repositories and cannot be readily utilized in research. A medical IR (Information Retrieval) system, called Electronic Medical Search Engine (EMERSE), can serve as a core tool to support cohort discovery, direct clinical work as well as secondary use of free-text data. Therefore, we design research studies to explore and evaluate its usefulness and effectiveness, and further implement new features derived from these insights in laboratory settings and in the.

Research Area #3: Improving the readability and the comprehensibility of clinical notes for patient communication

There has been evidence that patients can benefit from having direct, electronic access to their medical records including clinical notes. For example, patients can review their medical history, engage in effective self-management, and make decisions jointly with their providers. Unfortunately, a majority of patients have insufficient knowledge to read and understand clinical notes written by healthcare professionals. This issue not only diminishes the expected benefits, but may also cause unintended adverse consequences, i.e. misunderstanding of acronyms such as “OBS” and even unnecessary lawsuits. In this area of research, our goals are enhancements of the readability and comprehensibility of clinical notes shared to patients and provide some transformation tools that allow clinicians to semi-automatically translate clinical notes to a version suitable for laypeople.