Study evaluates stack mode viewing techniques

October 28, 2008
Douglas Page

In the absence of satisfactory 3D viewers, most volumetric medical data is currently viewed as stacks of 2D slices. Yet little research has been conducted on stack mode viewing techniques.

In the absence of satisfactory 3D viewers, most volumetric medical data is currently viewed as stacks of 2D slices. Yet little research has been conducted on stack mode viewing techniques.

A recent study from Canada evaluated three interaction techniques for scrolling stack mode displays of volumetric data:

  • Two techniques used a scroll-wheel mouse.

  • One used only the wheel, while another used a click-and-drag technique for fast scrolling, using the wheel only for fine-tuning adjustments.

"The paper addresses an underutilized research area of how new technology, in this case the contour shuttle device, can improve radiologist workflow during the typical task of viewing potentially hundreds of 2D CT and MR image slices using cine mode image navigation," said M. Stella Atkins, Ph.D., of the School of Computing Science, Simon Fraser University.

Atkins' experiment compared the three different interaction techniques in a group of nine radiologists and presented statistical results in the form of novel navigation charts, which illustrate image scrolling behavior over time (J Digit Imaging 2008 Jul 23 [Epub ahead of print]).

Findings show that the variation of the rate of scrolling was high within the group, but there was no statistical difference among the three image navigation interaction techniques.

"Subjective results suggest that radiologists prefer familiar mouse-based techniques for image navigation rather than the new contour shuttle," Atkins said.

Atkins did observe, however, that the fastest radiologist used the middle finger for mouse scrolling. The index and fourth fingers were used for conveniently pressing the mouse buttons.

"All other radiologists used the index finger for scrolling, which led to slower performance, because the index finger had to be moved to perform a right click," she said.

More radiologists also reported fatigue with this technique.

Atkins conducted a similar experiment with a user group of eight students.

"Most of the students preferred the new contour shuttle device," she said.

Atkins concludes that while having several techniques available on workstations for image navigation allows radiologists to choose their preferred method, it may also be advisable to teach radiologists to use the middle finger for scrolling.

"It may also be worthwhile to introduce new techniques to radiologists during training, such as the jog wheel," Atkins said.