Virtual Tomographic Reflection for Psychophysical Analysis of the Sonic Flashlight

Abstract

Augmented reality systems show great promise for improving the ability of medical personnel to incorporate image guidance into a wider variety of procedures than they are currently able. In its present form, medical image guidance often requires that patients be moved to a specialized interventional radiology suite and that specially trained personnel conduct the image guided therapy. Much of the difficulty in medical image guidance centers on mentally fusing the medical image with the patient? body. Augmented reality systems can alleviate this difficulty be visually fusing the medical image with the patient? body, allowing much more natural guidance. Among the questions that can be asked about augmented reality systems in medicine are A) ?o these systems actually improve the ability of medical personnel to treat patient??and B) ?ow do these systems affect perception of the medical image information?? This dissertation focuses on the Sonic Flashlight, an augmented reality device that provides real-time in situ visualization of ultrasound images by reflecting calibrated images displayed on a flat-panel monitor from a partially transparent half silvered mirror. Quantitative evaluation of the efficacy of the Sonic Flashlight versus conventional ultrasound guidance is difficult because of the lack of a real-world ?old-standard?reference frame in which to report results. In addition, psychophysics experimentation with the Sonic Flashlight is difficult due to the complexity of manipulating the image characteristics of conventional ultrasound phantoms. In this dissertation I explore the use of a virtual environment for conducting accuracy and psychophysical experiments on both the Sonic Flashlight and conventional ultrasound. In contrast to conventional phantom-based analysis of augmented reality systems, the virtual environment provides quantitative performance measures within a high-precision global frame of reference. In addition, the ease with which the virtual environment may be manipulated allows for a much more rapid exploration of psychophysical phenomena than would be possible with physical phantoms. The long term goals for this project are to support an ongoing effort to examine the impact of augmented reality technologies on medical procedures, and to provide a test environment for future augmented reality technology. i