Depth Recognition in 3D Translucent Stereoscopic Imaging of Medical Volumes by Means of a Glasses-Free 3D Display

In many scientific fields including the medical field, we often utilize translucent viewing, that is, see-through viewing, to investigate the inner 3D structures of complex objects. On the other hand, it is known that stereoscopic vision is effective in allowing us to intuitively understand 3D shapes and to realize natural depth feel of visualized scenes. It is expected that the combination of these two visualization techniques, that is, translucent viewing and the stereoscopic vision, namely translucent stereoscopic visualization, should be effective for our easier and intuitive understanding of inner structures of 3D objects. However, the cognitive effects that arise when combining these two techniques have not been fully understood for us until now. In this paper, we investigate the cognitive effects that arise when combining these two techniques of computer visualization. We specially focus on medical volume visualization to investigate influences of the luminance gradient, which is inherent in the stochastic point-based rendering (SPBR) that we proposed recently. We conducted psychophysical experiments in which observers analysed the perceived 3D structure based on translucent stereoscopic visualization. The experiments are executed under the conditions of monocular, binocular viewing and motion parallax. We found that the luminance gradient is effective in the perceived depth magnitude in the translucent stereoscopic viewing of medical volume data.