I began my training as a cognitive psychologist interested in how different ways of presenting information influence peoples' ability to make judgments and decisions. In my dissertation I showed that the more complex the presentation of information, the more judgment errors people made based on that information. However, when information was integrated in presentation, thus reducing the cognitive ‘work’ that people had to do, their judgments improved. Because our visual system is fast and powerful, I believe that complex information can best be integrated for people via visual presentation. My post-doctoral position (since March 2000) has allowed me to pursue this notion in the domain of anesthesiology.

Research

In anesthesiology information about the status of the patient is presented on computer displays as a combination of digital and chart-based information. The design of traditional displays has been driven by the single sensor single indicator approach: Information about the values of variables measured by a single sensor is displayed with a single indicator on the monitor. Given that a typical display shows over thirty variables in real-time, the information-processing capacity of the average human is clearly exceeded. One reason for the high number of anesthesiology related errors is that information is presented in sub-optimal ways. Existing displays do not support the anesthesiologist in detecting problems, making diagnoses and treating the patient.

An alternative approach is to develop an integrative display in which information from different sensors is integrated into a single graphical object (e.g. a sphere) (Figure 2). The basic assumption is that integrated information reduces the cognitive workload of the anesthesiologist, because she need not integrate the information herself. Moreover, using a graphical object permits exploitation of the human visual system's sensitivity to change in objects. Thus, deviations from perfect geometric shapes can be used to indicate changes. Our empirical evaluation supports the notion that visual integration is a good strategy. For example, in an interdisciplinary collaboration with Dwayne Westenskow (Bioengineering, Anesthesiology), Noah Syroid (Bioengineering), Julio Bermudez (Architecture), I ran a study where we compared anesthesiologists' performance in simulated surgeries using an integrated display and a standard display. We were especially interested in the time needed to detect changes in variables and to diagnose the cause for changes. Using integrated objects to provide information reduced detection time and improved diagnosis.

Figure 2: Integration of variables (heart object)

In another study we examined the effect of a new and innovative way of displaying drug information to anesthesiologists. In addition to other tasks anesthesiologists must constantly supervise patients' drug dosages to maintain appropriate levels of analgesia, anesthesia, and neuromuscular blockade. We wanted to see if integration of information assists them in doing so. We provided anesthesiologists with integrated information about effects of drugs on simulated patients. Here we were interested in the effect of visually displaying the pharmacodynamic effects of drugs on anesthesiologists cognitive workload and drug dosage management. Providing anesthesiologists with the drug display reduced workload, and improved drug dosage management.

In both studies, providing physicians with information that is already integrated improved their performance. In the applied context of the operating room, the result would be faster detection and diagnosis of problems, more effective drug administration, and an overall reduction of human error.

In addition to the studies reported above we ran several experiments where we psychophysically scaled the objects, which will be part of a new anesthesia display. Additionally, we did some experiments where we tested the controllability of these objects. Currently, we are running subjects in using eye tracking (ASL 501) to determine precisely which information anesthesiologist’s extract from the newly designed displays. The participants in these experimental studies are performing anesthesia in a high fidelity simulator in the Patient Simulation Center at the University of Utah. They have to treat two evolving critical scenarios. By recording fixations, pupil diameter and other performance data like detection times, diagnosis time and quality of diagnosis, we will be able to determine the improvement in performance when using the integrated display. Further it will be possible to determine precisely which parts of the display are used to extract relevant information, and which parts are not relevant for the anesthesiologist.

All reported studies are part of a larger interdisciplinary research program (Grant National Institute of Health Grant (IRO1 HL64590). The goal of this project is to develop and to evaluate a new, integrated display for anesthesiologists, which uses principles of graphically integrating information. The project is an interdisciplinary collaboration between Bioengineers, Anesthesiologists, Cognitive Psychologists, Computer Scientists and Architects.

Interests

The topics I am interested in, are focussed around experimental issues and issues of perception and visualization.

With regard to perception and visualization I am interested in questions about ways of presenting integrated information more efficiently. This includes questions about using space, texture, and color on a display to provide information about differences and similarities of elements. Further, I am interested in using variables like eye-movements and fixations to determine extraction of information for decision making.

With regard to experimental issues I am interested in evaluation techniques of displays and methods of validating the effectiveness of new experimental techniques.