Using Virtual Reality Effectively in Scientific Data Exploration - Perception, Usability and Design in Immersive Displays

In recent years, the rising popularity and availability of Virtual Reality (VR) systems has enabled the use of immersive display systems for scientific data exploration. This opened up a large design space for novel visualization techniques that needs to be thoroughly investigated in order to build practical software. This dissertation investigates aspects of creating effective VR visualizations from multiple angles: First, by analyzing how display fidelity affects user performance immersive data exploration situations. Here, results from a large scale study evaluating the factors of display resolution, field of view and display artifacts show their varying impact on user effectiveness depending on the exploration tasks presented.

A second line of studies analyzed the limits of human perception within high-resolution VR environments, by evaluating the ability of users to identify minute details in immersive medical scan visualizations and their ability to read text within 3D environments containing visual obstructions. In an effort to obtain generalizable results, the experiments above were performed across multiple VR hardware systems, leading to recommendations about their effective uses.

Finally, this dissertation provides insights into best practices for VR application development through a large-scale case study of the “Scientific Sketching” design methodology. A managed collaboration of biologists, art students and computer science students over the course of three years, resulted in a novel fluid-dynamics visualization application tailored to the specific needs of our biology collaborators. The information gathered alongside the development process built a strong basis for future VR application development. Together, the contributions of this work form a set of guidelines for the effective use of immersive visualizations in scientific settings.

Brown Digital Repository