Comparing visual-spatial task ability using monoscopic and stereoscopic views in medical data

Marisol Martinez Escobar, Bethany Junke, Joseph Holub, Eliot Winer, PhD. Iowa State University.

The dramatic rise in the use of digital medical imaging technologies like Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) allow medical professionals to see inside their patients like never before. Software products are available that allow this data to be viewed in various 2D and 3D representations. Basic research questions on human spatial perception are raised as the use of these representations grow, and the devices on which they are viewed. Understanding spatial ability in medical tasks is crucial as it is required in numerous everyday activities from surgery to radiology. As new image modalities and visualization technologies are introduced in the medical field, understanding the advantages and shortcomings with spatial tasks becomes a necessary research issue to address.

The work presented here builds on a previous study attempting to isolate the benefits between monoscopic 3D and stereoscopic 3D displays in relative position tasks. The results of this previous study showed stereoscopic displays yielded a higher score than monoscopic displays, but the results were not always statistically significant. There was also a question on the use of color in the visualization and its impact on spatial performance.

A new study was conducted to further explore these issues, specifically: 1) investigate the role of color in spatial tasks in a medical context, and 2) gain a better understanding of the best practices for using stereoscopic versus monoscopic displays for different medical spatial tasks. 46 participants with medical experience ranging from 1 to 20 years were shown a 3D representation of a chest cavity CT with three cylinders positioned within. The participants were given 40 tasks to perform. Each task consisted of a pre-determined colored and windowed view of the CT volume. Participants were asked to determine which of two outer cylinders were closer to the central cylinder. As the tasks progress, the outer cylinders change positions within predetermined ranges, while the central cylinder remains in the same position. Thus, each task requires a spatial judgement to be made. Each participant performed all 40 tasks once using colored cylinders and once using gray cylinders with identifying text labels. Each participant used either a monoscopic or a stereoscopic display for the entirety of their study. Participants were randomized and counter balanced between the different testing modes.

Results of the study showed on average participants using the stereoscopic display correctly selected the closer cylinder 76.1% of the time versus 69.2% of the time for participants using monoscopic displays. This difference is statistically significant to a 99% confidence level. While overall the participants using stereoscopic displays performed better, there were certain tasks that were consistently performed better on the monoscopic displays. Ten of the forty tasks were performed better on the monoscopic displays with two of them being statistically significant. Interestingly, the tasks performed better in monoscopic were consistent between the previous study and this study. The results also determined that color had no impact on task performance.

View Poster

« Return to SAGES 2014 abstract archive