1. S Optical CT scanning of PRESAGE TM polyurethane samples with a CCD-based readout system S J Doran 1*, N Krstajic 1, J Adamovics 2 and P M Jenneson 1 1 Department of Physics, University of Surrey 2 Heuris Pharma, Skillman, NJ

2. Structure of talk Basic design of optical CT scanner New design features of scanner (including works in progress)  Light source  Collimation arrangement  Scanning tank Irradiation and imaging of PRESAGE TM samples Characterisation of scanner and future prospects

3. Scanner schematic Hg lamp Cylindrical lens, pinhole and filter  pseudo point-source Lens  parallel beam Scanning tank with matching medium Exposed gel Unexposed gel Diffuser screen on which real shadow image forms CCD detector Standard 50mm camera lens PC with frame- grabber card Turntable controlled by acquisition computer via stepper motors

4. Disadvantages of existing design Mercury lamp is too weak  Light intensity is split between three colours.  Fricke gels are moderately absorbing even before irradiation.  Light is at wrong wavelength for scanning PRESAGE TM, which was optimised for a HeNe laser (632 nm). Diffuser screen wastes light and introduces “noise”  Introduction of a properly designed collimation system greatly improves performance.  “Noise” from diffuser screen was coherent between projections and gave rise to ring artifact. Perspex is not a suitable material for the end walls  New design will use higher quality optical glass.

5. Solutions (1): Light source Light source changed to ultra-bright LED  Considerable investigation of different options (expanded laser beam, laser + fibre optic, more powerful discharge lamp)  New LED products have appeared within the last year.  Range of wavelengths available: one can pick one to suit any of the current families of gel.

6. Solutions (2): Collimation New lens introduced at the exit of the tank  Light still passes through the sample as a parallel beam.  Second lens focuses light down to an appropriate acceptance angle for the camera lens.  Standard 50 mm camera lens focuses light onto CCD chip. * LED light source Scanning tank CCD chip

7. Sample irradiation PRESAGE TM samples supplied as cylinders of radius ~7 cm Two experiments undertaken  2.7 x 2.7 cm 2 square field  grid irradiation using purpose built lead collimator Top view Cross-section X-ray beam ~30 Gy at lead surface 2 mm diameter holes PRESAGE TM

8. Results on PRESAGE TM samples Results of this first test show early promise, but highlight a number of problems  resolution is excellent (> 7 pixels across 2 mm spot)  contrast / artifact ratio is poor  difficult to achieve good contrast in original projections for these very small features

9. Discussion (1): Artefacts Artefacts are currently the limiting factor in both resolution and dose sensitivity.  Slice thickness was very large (~2 cm) in images presented.  Ring artefact: These initial experiments were performed on the original system with diffuser screen. Solution: With the modified system, significantly better results will be obtained.  Refraction artefacts: These have two causes, poor mixing of the matching liquid and (possibly) internal inhomogeneities of the PRESAGE TM. CCD optical-CT appears more sensitive to these than laser optical-CT. Solution: Commercial matching liquids under investigation  Absolute attenuation coefficients: Current camera is not yet calibrated over full dynamic range.

10. Discussion (2): Sensitivity CCD tomography scanner originally designed for use with Fricke gels.  PRESAGE <  Fricke for a given dose. At present, light needs to pass through a relatively large path-length of gel to obtain a measureable effect Solutions (both can be attempted in parallel)  Modify PRESAGE TM characteristics  Increase digitisation depth of camera (10  16 bits) once the artifact problem has been solved.

11. Characterisation of scanner (1) In order to investigate the resolution and 3-D capabilities of the scanner, a phantom was made from wires (c.f. needle phantom of Oldham et al.) Raw projectionSinogramSingle slice3-D reconstruction

12. Characterisation of scanner (2) A series of experiments investigated the effect of the diffuser screen and the system resolution. Reconstructed slice from experiment with diffuser screen Corresponding surface plot

13. Characterisation of scanner (3) Replacing the diffuser screen dramatically reduced the ring artefact, as expected. Reconstructed slice from experiment without diffuser screen Corresponding surface plot Residual artifacts are due to the problem of an “infinite absorber”

14. Characterisation of scanner (4) A profile through the thinnest wire demonstrates the extent of the ring artifact problem. Original arrangement with diffuserNew arrangement with no diffuser

15. Characterisation of scanner (4) From a knowledge of the true profile, we can estimate the MTF of the system by constrained deconvolution.

16. Conclusions We have performed the first CCD-based optical-CT scans of the new PRESAGE TM dosimeter. We have identified a number of issues that need addressing before this method will yield high-quality results. However, these problems appear to be soluble using available technology. The excellent resolution and speed of the CCD optical-CT system should be compatible with the excellent dosimetric properties of PRESAG TM.