Visualizing 3D objects from 2D cross sectional images displayed in-situ versus ex-situ

Area of Science:

• Cognitive Psychology
• Medical Imaging
• Human-Computer Interaction

Background:

• Medical imaging relies on reconstructing 3D objects from 2D cross-sectional slices.
• Current medical imaging often displaces these slices to a remote screen (ex-situ).
• This displacement may hinder accurate 3D mental visualization.

Purpose of the Study:

• To investigate the impact of image display location (in-situ vs. ex-situ) on 3D object visualization from cross-sectional images.
• To assess how spatial information integration is affected by image displacement.
• To determine the influence on judging object orientation (pitch and yaw).

Main Methods:

• Three experiments were conducted using a virtual rod reconstructed from cross-sectional images.
• Participants visualized the rod using either in-situ (directly in space) or ex-situ (remote screen) displays.
• Participants adjusted a stylus to match the virtual rod's pitch and yaw, with performance analyzed via error and regression.

Main Results:

• Simultaneous judgment of pitch and yaw revealed systematic errors in ex-situ displays compared to in-situ.
• Ex-situ imaging showed reversed directional errors, though magnitude was often correct.
• In-situ judgments relied on spatiotemporal visualization; ex-situ judgments used ad hoc, screen-based heuristics.

Conclusions:

• In-situ displays facilitate more accurate 3D mental visualization from cross-sectional images.
• Ex-situ displays introduce systematic errors, particularly when judging multiple dimensions of spatial orientation.
• In-situ displays show potential for improving radiologists' 3D anatomical visualization in clinical practice.