HUP 5/6 Projection Setup for fMRI Experiments
   

Introduction

This page describes the projection system installed in the HUP 5/6 magnets in the basements of Founders and Rhoads respectively. The system is evolving, and we will try to keep this page up to date with description of its current state.

Equipment
Power Up Sequence
Power Down Sequence
Video Input
Screen Placement
Spatial Dimensions
Spectral Calibration
Replacing the Bulb
Known Problems/Issues

Equipment

Both HUP 5/6 have projection systems installed. The basic system for each magnet is the same.

  • Sanyo SXGA 4200 lumens projectors with Sanyo Long Throw zoom lens. FiberLynx optical-fiber VGA connection into magnet room. Custom RF shield box. Mylar rear projection screen at back of bore. Image projects to screen off of 45 degree mirror on back wall of magnet room. Image is viewed through mirror mounted on head coil. Before aligning the projectors, we reset all projector parameters to factory default. We then put them into rear projection mode.
  • There is an energy saver feature on the projectors that darkens the screen if the video input doesn't change for a while. This is currently turned on, but it is possible we want to turn it off.

Power Up Sequence

  • Turn on master power switch for system. This is located on the vertically shielded box near the floor. Slide open door on top of projector box. If you lean over and look back into the hole, you will see that the projector light is yellow and not flashing. Push the power button on the projector. The projector light should now flash green.
  • When the light goes to steady green, the system is ready for action.

Power Down Sequence

  • Slide open door on top of projector box. If you lean over and look back into the hole, you will see that the projector light is green and not flashing. Push the power button on the projector and hold it for a few seconds. The projector light should now flash yellow. When the light goes to steady yellow, the system is ready to power down. This takes about a minute.
  • Turn off master power switch for system.

Video Input

  • Your computer hooks to the control room end of the FiberLync fiber box. It takes VGA input. Ideally, the projector should be driven at its native spatial and temporal resolution: 1024 by 768 pixels, 60 Hz refresh. Driving with these parameters is likely to minimize the image processing applied by the projector electronics to your stimulus.
  • I have had trouble driving it at 60 Hz -- the image comes out shifted and incomplete. Therefore I have been driving at 75 Hz refresh. Visually this looks fine.

Screen Placement

  • The screens are mounted on stands. At the back of each bore, blue tape on the floor indicates the position of the stand for which we focused the projectors. The stand should be oriented so that the screen is on the bore side of the stand.
  • The projectors were aligned and focussed on the screens, you shouldn't need to adjust the focus or alignment.

Spatial Dimensions

HUP 5
  • The image is considerably smaller than the bore because of where we have to put the head coil mirror to see it. The image measures 32.9 cm (w) by 29.6 cm (h). With the current mirror (mounted above head coil), the distance from eye to screen is about 101.5 cm.
  • At this distance the image subtends a visual angle of 21.85 (w) by 16.6 (h).
HUP 6
  • The image is viewed from a mirror inside the head coil, which seems better than the HUP 5 setup. The image measures 50.4 cm (w) by 38 cm (h). With the current mirror (mounted inside the head coil), the distance from eye to screen is about 124.25 cm.
  • At this distance the image subtends a visual angle of 22.9 (w) by 17.4 (h).

Spectral Calibration

HUP 6
  • We've made spectral measurements of the red, green, and blue LCD channels of the projector, as well as of the projector's "gamma" function. These are tabulated in an Excel spreadsheet, HUP6Cal.xls. Spectral calibration of display devices is discussed in a recent chapter by Brainard, Pelli, and Robson (PDF). This chapter may be of interest for those who want to use the data in the spreadsheet.
  • We have not yet made a detailed study of the appropriate model for characterizing the projectors. More to follow.

Replacing the Bulb

  • Sooner or later the bulb is going to give out in one or the other of the projectors. Probably at the most inopportune possible moment. We have a spare bulb, but replacing it involves removing the projector from the magnet room, something that needs to be done very carefully to avoid having it sucked into the bore. Don't try this yourself!
  • Ask one of the MRI techs to contact Mark Elliott, PhD at MMRRCC or John Pluta at the CfN. Both have access to the replacement bulb as well as the expertise to replace it. Call Margaret Ryan at 215-614-0630 if you need immediate assistance.

Known Problems/Issues

There are a number of problems and issues with the projections systems. These are to be expected with any new system, and as a group we need to work through them towards better solutions. Ideas and volunteers welcome. The following is a list of currently known issues and some thoughts about them.

  • HUP 5, head coil mirror vibrates during scans.
    • It is possible this can be fixed by making a more solid link between the head coil mount and the mirror.
    • Another possibility is to try to construct a mirror inside the coil, like the one on HUP 6. It is not clear whether the HUP 5 coil is large enough to accommodate this.
    HUP 5, it's hard to see full image.
    • If you get the mirror in just the right place, you can see the full image. It is a bit tricky to wiggle it around to where you want it. Getting the mirror inside the head coil may help, see above
    • If anyone can look into what solutions were found at other Siemens sites (MGH?, Princeton?) that might also give us some ideas.
    HUP 5/6, image ghosting
    • The head coil mirrors are not front-surface. We should try to get front surface mirrors to reduce ghosting.
    HUP 5/6, spots visible in image
    • With certain backgrounds, ugly spots are visible across the image. The origin of these spots is not yet clear. On April 18, we tried lining the RF pipe with black velveteen but this had little effect. The spots are visible when you project directly rather than across the mirror on the back wall, so it's not the mirror. These spots may result from light scattering around inside the RF box and out to the screen, they may result from something funky about the Buhl lens, they may result from back reflections off the UV filter, or they may be intrinsic to the projector. We will continue to try to track this down.
    HUP 5/6, rf noise?
    • So far we've heard no reports of RF noise since we put the covers back on the projectors, but keep an eye out for it. We should be able to run some good tests with a phantom.
    HUP 5/6, possible to get larger FOV?
    • It would be nice (for me at least) to have a larger angular field of view (FOV). At some point we can explore mounting the screen in the bore, closer to the subject. If the head coil mirror can be arranged to let us see the whole thing in this case, it should increase the FOV.
  • HUP 5/6, projector imaging properties?
    • LCD projectors are an evolving technology and each projector has some idiosyncratic properties. It is likely, for example, that the image is delayed a frame or two (~16=32 msecs) from when the video was sent out from the computer. It would be good to characterize these aspects of our projectors carefully.
   

Author: David Brainard, brainard@psych.upenn.edu
Last modified: 19 September, 2013