Week 2 summaries

By on January 16th, 2013

Merging Virtual Objects with the Real World: Seeing Ultrasound Imagery within the Patient

The paper describes how a small video camera mounted on a HMD can be useful for seeing ultrasound imagery within the patient. This approach basically merges virtual images with the real world view and can be extended for use in image guided surgical procedures, seeing through smoke in burning buildings, on site architecture preview and so on. However ultrasound imagery has several disadvantages as well such as low SNR, poor spatial resolution, speckle, increase in attenuation with frequency, phase aberration and reflection and refraction artifacts. It is yet apt for fetal examination, cardiac study etc because ultrasound imagery is safe and generates images in real time.

The system deals with 3 major areas such as:

  1. Real time ultrasound data acquisitions and rendering – The system incrementally visualizes a 3d volume dataset from multiple 2d data slices because when the system was constructed only 2d ultrasound scanners were available for real time volume data acquisition.  They sample the target function at irregular points and then reconstruct a 3d volume from this time series of spatially irregular sample points. This is useful as it limits the storage and computation requirements. The reconstruction is a 4d convolution process i.e a 3d Gaussian kernel is used for spatial reconstruction followed by a temporal reconstruction based on ARMA filtering. The reconstructed volume is then rendered with an incremental volume rendering technique. A hierarchical ray cache that stores ray samples is used for fast rendering of polygons.
  2. Creating a virtual environment that displays this ultrasound data in real time –  As each echography image is acquired by the ultrasound scanner, the scanner’s and HMD’s position and orientation in 3d world space are tracked with 6 DOF. Using this geometry, 3d renderings of the 2d ultrasound images is constructed by an image generation system. A small TV camera mounted on the HMD captures real world images which are video mixed with the 3d ultrasound image renderings.  The resulting composite image in turn shows the 2d ultrasound data registered in its true 3d location.
  3. Recovering structural information for volume rendering – This is concerned with surface definition and not covered in this paper.

The systems hardware is composed of 3 major components:

  1. Image acquisition and tracking system- It includes the ultrasound scanner and a Polheums tracking system which has 2 receivers, one tracks the HMD and other tracks the ultrasound transducer.
  2. Image generation system – It includes Pixel Planes 5 graphics multicomputer
  3. Video-see through HMD – It includes a portable TV camera, a video mixer and VPL EyePhone.

After testing it in the lab, the first live experiment was on a 38 week pregnant volunteer. The results were quite impressive though there were technical problems their system couldn’t resolve such as conflicting visual cues, system lag, tracking system range and stability etc. Most of these problems were due to the technology constraints and hence their approach seems to be great contribution to Computer graphics and visualization.

Question: Considering the number of applications of this technique the cost of HMD’s should reduce due to mass production. But even in 2013 good quality HMD’s are still too expensive for public use. Why? Why are most of the technological constraints faced by Bajura still existing?

Virtual Environments for Treating the Fear of Heights

The paper is quite interesting as it deals with how virtual environments can be useful for treating the fear of heights.  Acrophobia is treated by exposing the subject to anxiety producing stimuli. Studies show that in vivo(real) graded exposure is more effective than systematic desensitization(imaginary). What distinguishes VE ‘s from Computer Graphics and multimedia is the sense of presence and immersion. The paper validates that this sense of presence can be positively used to produce the same results as in vivo graded exposure. While VE’s had been applied to psychology prior to this paper, no study documented clinical data or analytical results. The authors of this paper so focus on proving how VE’s are effective in treating acrophobia rather than the technical aspects of developing such a Virtual Environment.  For the therapy, three virtual environments were created: an elevator, a series of balconies and a series of bridges of which most subjects founds bridges as least scary. Tactile feedback provided in each of these environments increased the subject’s sense of presence and immersion.  Questionnaires were used to find suitable subjects and record the study results and subjects feelings. The good feature about the experiment I feel was that the system was flexible as it allowed the subjects to spend as much time as they need in each situation to overcome their fears.  In general, subject’s anxiety increases as they experience more threatening situations and decreases as they spend more time in the same situation. The study shows that subjects experienced similar change in anxiety even in the Virtual Environment situations. Thus, using VE for psychological treatment is useful as it is less expensive in terms of both time and money as compared to in-vivo graded exposures. Also VR exposures can be made extreme that can be life threatening in real world.  The only two problems with VR exposure is nausea and being ineffective for people with poor eyesight. They treated nausea by limiting sessions to 30-40 minutes which seems as a reasonable solution. VR is ineffective for people with poor eyesight as they lose the sense of presence and immersion and this problem can be tackled by using specifically programmed images for them.  In totality, I feel Virtual Environments are apt for treating different forms of fears and for various psychological treatments.

Question: The results show that it reduces a small group of peoples’ anxiety but has/can it be scientifically proven as a therapy?

Designing Interactive Theme Park Rides

This paper is all about the design considerations Disney made to make their Theme Park ride fun. Disney realized that people will pay for the ride only if it is entertaining enough and it involves social interaction.  To ensure that the ride is fun they needed to prevent people from steering to dull places. By having sneak attack ships suddenly appear, ensuring at least one of the islands is in view, using water force to redirect the ship Disney ensured that people sailed to interesting spots. Since 5 minutes is too less to guarantee that guests would experience an exciting climax, they scripted the game such that the guests have to battle with Jolly Roger in the last half minute to win/lose the game.  Since 5 minutes are too less for guests to understand complex user interfaces they used steering wheel and cannons which every player can intuitively use. 3d surround sound, tactile speakers, strobe lights, 3d stereo glasses made the virtual world experience more real i.e increased guests sense of presence and immersion.  Since people visiting theme parks are always in group, this ride involving 4 players helps them experience a shared social interaction. Guest testing assured that the game is balanced properly and allowed to deal with unpredicted social behaviors.  All these features make the ride different than a video-game or an experience that can be replicated in the living room and ensure that the players enjoy the ride.

Question: Considering the technological innovations, why can’t such an experience be re-created in one’s living room even if on a small scale?

Categorized under: week2.
Tagged with: .

Comments are closed.