A comparison of head-tracked and non-head-tracked steering modes in the targeting of radiotherapy treatment beams

Proceedings of the 1992 symposium on Interactive 3D graphics - SI3D '92 (1992), pp. 193-196, doi: 10.1145/147156.147204
Experiments
Tasks
Findings

Do head-tracked steering interactions result in better performance than non-head-tracked steering interactions for beam-targeting tasks?

Variables and Constants

Independent Variables
  • head tracking - four head-tracked steering interactions and 3 non-head-tracked steering modes were evaluated, all conducted on an hmd.
  • steering mode - steering modes were the direct independent variable, with each interaction mode either using or disabling head tracking. the experiment evaluated different head-tracked and non-head-tracked steering interactions. here are the 3 head-tracked steering modes
Constants

Image quality: The experiment ran all on an HMD, enabling and disabling head tracking for the different tasks to keep image quality and size constant. All other components of immersion were constant besides on/off head tracking including display size, weight, brightness, frame rate, refresh rate, jitter, resolution, etc.

System Info

Displays
Software
None

Participant Info

Not reported, but reported that participants came from graduate students and staff from their Computer Science, Radiation Oncology, and Radiology department.

Total # Age Range Gender Balance
14 -

Beam-targeting is a task in radiotherapy, where the doctor must position and rotate a beam such that the entire tumor is covered in the beam, with the least amount of beam touching the surrounding healthy tissue. The task is to move a beam towards a tumor section of the model, while minimizing hitting the healthy parts of the model, which we consider here as performance. Participants were also timed as well, and asked how confident were they in their results. User Preference in this case was based on a rating on ease of use and preference, where participants ranked each interaction technique.

Environment

The environment was a clustered set of approximately 20 spheres in different colors. The tumor in this was was a multi-colored ball in the near center of the cluster.

Dimensionality Scale Density Visual Realism
3D Medium Not reported Not reported
Metrics

  • errors - percentage of healthy tumors hit by the beam
  • time - completion time
  • user preference - preference in interaction technique
  • user task perception - in this case, this was ease of use for each of the interaction techniques
  1. There was a significant interaction between head tracking and steering mode on errors for a travel - maneuvering and object manipulation task.

    The Orbital, a head-tracking technique, did significantly better most head and non-head-tracking steering interaction techniques.

    Specificity: Highly specific
    Orbital did significantly better than 6D mouse, walkaround, walk/rotate, and immersion tasks.