P09052: Molecular Imaging System Upgrade
/public/

Test Product Concept(s)

Table of Contents

Optics Proof of Concept

Lenses

Experiment 1 - Keystoning

public/OpticsExp.jpeg

We found that in order to get an image the size of a hand in the space we have we probably need a compound lens.
Experiment 2 - OptiGrid Diffraction

public/LaserDiffraction.jpeg

We shined a laser through the OptiGrid to look at the diffraction. We get similar results to those using a thin slit diffraction.

Experiment 3 - Projecting Grid

public/projectgridset.jpeg public/gridprojection.jpeg public/gridhanproject.jpeg

We were able to project and focus the square containing the grid several feet away and make it big enough to fit a hand in. We could not see any lines

Experiment 3 - Rear Projection Mode

Operating the system in rear projection mode consists of turning the structured illumination optics upside down and using the Image Station camera to test concepts.

Rear Projection Mode Experiment Diagram

public/Rear Projection Mode.jpeg

Furthermore, the Image Station was turned on its side to allow for easy experimentation on the optics bench. The system was tested using straight down 90 degrees projection as well as at 45 degrees using a Scheimpflug lens system.

public/Rear Projection Setup Actual.jpeg

public/Measuring Angles.jpeg

Results

The grid was successfully projected using the Computar C mount lens at approximately 75x magnification. Images were captured for both 45 and 90 degree configurations.

90 degrees

public/90.jpeg

45 degrees

public/45.jpeg

Florescence

public/phand.jpeg public/phandimaged.jpeg public/phandimagedneg.jpeg

The first image shows what was being imaged, the second the positive image and the third the negative image.

The image of the hand is the natural florescence of the palm of the hand. The bright strip is a piece of highly florescent plastic that was wrapped around the top of the finger, away from the imaging surface. Theoretically if we can filter out the natural florescence we can see the florescence of the plastic behind the finger.

Depth Resolution

We are creating a Phantom to focus the OptiGrid on.
public/phantomblock.jpeg
Phantom Block PDF

Penetration of Tissue

Mechanical Proof of Concept

Angle of Grid

public/GridDiagram.jpeg
public/UsableGridGraph.jpeg

% Usable Gird Spreadsheet

public/UsableGridDrawing.jpeg

Coarse Calibration of Optigrid

Images were taken in rear projection mode in the range of positions of the Optigrid. The grid voltage was varied from 0 to 130 V in 5 V increments. The phase of the grid was calculated by taking the FFT of a line profile of the image and taking the inverse tangent of the ratio of the real and imaginary parts of value of the FFT at the first harmonic. A phase versus voltage curve was generated using this data.

public/PhasePlot.bmp

Voltage positions were chosen that were approximately 120 degrees apart. The voltages selected were 20 V, 45 V, and 70 V. The .tif files can be downloaded below.

20 V Image

45 V Image

70 V Image

As a proof of concept, these three images were combined, and the Optigrid Algorithm was applied. The resulting images was free of most of residual lines.

public/OG_Coarse.bmp

Fine Calibration of Optigrid

Several attempts were made to finely calibrate the Optigrid with limited success. The results had very high merit scores, and in general resulted in poor image quality.

Several ideas for improving the experimental setup were suggested. First, using reflected light instead of transmitted light will provide a clearer image. Second, an improvement in the motor control GUI was made, and it is anticipated that this will provide more accurate and repeatable phase shifts from the Optigrid. Third, padding the FFT in the calibration MATLAB scripts will potentially increase the resolution of the processed frequency information.

In order to support the use of reflected light, a separate camera was used for image capture. The optics rail being used in the final design was also incorporated into the experiment for improved flexibility and precision. The experiment is documenting in the link below.

Calibration Results - April 10th, 2009

Simulated Images

In order to test the Merit Map script, three simulated images were generated and processed. The images, the processed images, and the spectrum graph can be seen below.

public/0 Degrees.jpg

0 Degrees TIF

public/120 Degrees.jpg

120 Degrees TIF

public/240 Degrees.jpg

240 Degrees TIF

Combined Image

public/combined image.bmp

Combined Image TIF

Spectrum

public/spectrum.jpg

Phantom Captures

Pictures were taken in rear projection mode using a phantom test object. The phantom object was constructed from sheets of transparent plastics combined with sheets of fluorescent plastic (#96, Lime Green)placed in between them. Three images were taken at voltages approximating 0,120, and 240 degrees of phase shift. Downloadable images (.tif and JPEG) can be found below. Other details of the experiment include:

- Pictures were taken at an angle of approximately 49 degrees. This helped reduce glare while also simulating the finished system

- A blue excitation filter was used, and a green emission filter was used.

Phase 0: 20 V

public/V1_20volts.jpg

Phase 0: 20 V TIF

Phase 120: 48 V

public/V2_48volts.jpg

Phase 120: 48 V TIF

Phase 240: 72 V

public/V3_72volts.jpg

Phase 240: 72 V TIF

Calculated OG Image

public/OG_1.jpg

Calculated OG Image TIF

Repeatability

Accuracy

Program Proof of Concept

Accessible/Readable by Carestream Software

Home | Planning | Concept Development | System Level Design | Detail Design | Testing and Refinement