Real-Time Computed Tomography Volume Visualization with Ambient Occlusion of Hand-Drawn Transfer Function Using Local...

Area of Science:

• Medical Imaging
• Computer Graphics
• Scientific Visualization

Background:

• Computed tomography (CT) datasets require effective visualization techniques for accurate interpretation.
• Global illumination, specifically ambient occlusion, can enhance depth cues in rendered images.
• Real-time rendering of complex datasets like CT scans remains a computational challenge.

Purpose of the Study:

• To develop an efficient method for real-time visualization of computed tomography (CT) datasets using ambient occlusion.
• To improve depth perception in CT visualizations through global illumination techniques.
• To enable dynamic manipulation of transfer functions (TFs) during real-time rendering.

Main Methods:

• Preprocessing involves calculating local mean and standard deviation for CT data.
• Ambient light intensity is computed during rendering, accelerated by assuming normal distribution of CT values in local vicinities.
• Complex transfer functions (TFs) are approximated with line segments for further acceleration.
• Ambient occlusion is integrated with local illumination for real-time depth-enhanced image generation.

Main Results:

• The proposed method achieved real-time rendering speeds approximately 40 times faster than previous methods.
• Image quality was preserved, with an average signal-to-noise ratio of approximately 40 dB compared to conventional methods.
• The technique was successfully tested on various CT datasets with hand-drawn TFs.

Conclusions:

• The developed method enables real-time generation of depth-sensing images for CT data.
• This approach offers a significant speedup in CT visualization without compromising image fidelity.
• The technique facilitates interactive exploration of CT datasets through real-time rendering and TF adjustments.