DOSGEL 2006 Optical CT review talk

1
The history and principles of optical CT for
scanning 3-D radiation dosimeters 2008 update
Simon J Doran
2
Collaborators

• John Adamovics Heuris Pharma, Rider Univ. NJ
• Andy NisbetSteph Gerard-MartinAbdul
  Ismail Royal Surrey County HospitalChris Bunton
• John Brunt Clatterbridge Centre for
  OncologyAndrzej Kacperek
• David Bonnett Maidstone Hospital

3
Students

• Koen Kleinkoerkamp University of Twente, (NL,
  Socrates)
• Nik Krstajic University of Surrey (EPSRC)
• Philip Wai University of Surrey (EPSRC)
• Shamsa Al-Nowais University of Surrey (UAE)
• Mamdouh Bero University of Surrey (Sudan)
• Donna Talbot Maidstone Hospital

4
Summary of talk

• Contrast mechanisms in optical CT
• A brief history of optical CT
• ... or When the conditions are right, lots of
  people have the same idea!
• Scanned laser vs. pixelated detector systems
• Different optical scanner geometries
• Applications

5
Why do optical CT?

• Two orders of magnitude cheaper than MRI
• Every medical physics department could have a
  scanner
• You can build your own laser scanner for 10,000.
• Two orders of magnitude faster than MRI
• 32-echo MSE sequence, TR5 s 21 mins per 2562
  slice... Yes, Yves, you can interleave them, but
  still ...
• CCD or fast laser 20-30 mins for 2563 volume
• Kevins coffee cup test

6
Contrast in optical CT

• Source of contrast change in optical absorption
  or scattering

7
Contrast in optical CT

• Source of contrast change in optical absorption
  or scattering
• Fricke gels
• Mix with xylenol orange.
• Absorption change with Fe3

8
Contrast in optical CT

• Source of contrast change in optical absorption
  or scattering
• Fricke gels
• Mix with xylenol orange.
• Absorption change with Fe3
• Polymer gels
• Polymer particles scatter light.

www.iasa.gr/DosLab/
9
Contrast in optical CT

• Source of contrast change in optical absorption
  or scattering
• Fricke gels
• Mix with xylenol orange.
• Absorption change with Fe3
• Polymer gels
• Polymer particles scatter light.
• PRESAGETM
• Absorption change of dyeat HeNe wavelength 633 nm

10
History (1) Contrast mechanisms
OPTICAL MRI 1927 Fricke
and Morse (UV) 1950 Day and Stein 1970s,
1980s Gupta (xylenol orange) 1984 Gore (Fricke
solutions) 1984 Olsson Appleby Audet Late
80s Schreiner Jordan Kron 1990s and
others (Fricke gels) Maryanski (polymer
gels) 1993 1996 Gore (OT scanner) Late
1990s Jordan Bero (FXG) Many authors (polymer
gels) 1993- Fong (MAGIC gels)
2001 2003 Adamovics (PRESAGETM)
11
History (2) Invention of optical CT

• A reminder NMR was discovered simultaneously
  with two different flavours in 1945.
• In a similar vein, ideas of optical CT were
  developing simultaneously in several locations
  ...
• Gore et al. PMB 41, 2695 (1996) original
  submission Sept. 1995

12
History (2) Invention of optical CT

• A reminder NMR was discovered simultaneously
  with two different flavours in 1945.
• In a similar vein, ideas of optical CT were
  developing simultaneously in several locations
  ...
• Gore et al. PMB 41, 2695 (1996) original
  submission Sept. 1995
• Winfree et al. Chaos 6(4), 617 (1996)
  original submission Oct. 1996

13
History (2) Invention of optical CT

• A reminder NMR was discovered simultaneously
  with two different flavours in 1945.
• In a similar vein, ideas of optical CT were
  developing simultaneously in several locations
  ...
• Gore et al. PMB 41, 2695 (1996) original
  submission Sept. 1995
• Winfree et al. Chaos 6(4), 617 (1996)
  original submission Oct. 1996
• Gilboy (University of Surrey) independently
  sketches closely similar design to Winfrees a
  few weeks before the publication of either
  of these two articles. Bero et al. (1999)

14
History (2) Invention of optical CT

• A reminder NMR was discovered simultaneously
  with two different flavours in 1945.
• In a similar vein, ideas of optical CT were
  developing simultaneously in several locations
  ...
• Gore et al. PMB 41, 2695 (1996) original
  submission Sept. 1995
• Winfree et al. Chaos 6(4), 617 (1996)
  original submission Oct. 1996
• Gilboy (University of Surrey) independently
  sketches closely similar design to Winfrees a
  few weeks before the publication of either
  of these two articles. Bero et al. (1999)
• Wolodzko et al. Med. Phys. 26(11), 2508 (1999)
• Tarte (PhD thesis, 1995) Does anyone have a copy?!

15
History (3) Optical Projection Tomography (OPT)

• James Sharpe from the MRC in Edinburgh introduced
  OPT as a new imaging method in 2002, has
  patents and a company called Bioptonic.

Data source Sharpe, J. Anat. 202, 175 (2004)
16
History (3) Optical Projection Tomography (OPT)

• James Sharpe from the MRC in Edinburgh introduced
  OPT as a new imaging method in 2002, has
  patents and a company called Bioptonic.
• Essentially this is just optical CT microscopy.
• Two other groups currently performing this
• M Oldham, Duke

• S Doran, University of Surrey

17
History (4) Evolution
1999
2007
2001
18
History (4) Camera evolution
19
History (4) Light source evolution
20
What difference does it all make?

• Optical CT at the University of Surrey is a tale
  of slow but steady improvement between 1999 and
  2008.

21
Further improvements using laser scanner

• First results from the laser scanner are
  extremely encouraging.

2007 Laser scan
2006 CCD scan
Results for a slice containing significant
refractive index inhomogeneities (schliere)
22
Scanned laser vs. pixelated detector (1)

• Single-beam tomography(Gore, Maryanski et al.,
  1996)

• Pixelated detector (e.g., CCD) (Krstajic and
  Doran 2007)

Other workers in the field Oldham et al. Jordan
et al. Ravindran et al. Commercial vendor MGS
Other workers in the field Wolodzko, Appleby et
al. Jordan et al. Oldham et al. Commercial
vendor Modus
23
Scanned laser vs. pixelated detector (1)

• Single-beam tomography(Gore, Maryanski et al.,
  1996)

• Pixelated detector (e.g., CCD) (Bero, Doran et
  al., 1999)

Data source Mark Oldham
24
Scanned laser vs. pixelated detector (2)

• LASER
• Advantages
• Only one beam to worry about
• High bit depth from cheap detectors
• Disadvantages
• Slow
• True 3-D with thin slices is impractical

• PIXELATED DETECTOR
• Advantages
• Very fast! (e.g., CMOS)
• High spatial resolution and large FOV possible
  simultaneously
• True 3-D
• Disadvantages
• Good detector is expensive
• Schliere are a problem
• More prone to artefacts

25
Scanned laser vs. pixelated detector (3)

• Scanners based on pixelated area detectors are
  undoubtedly more sensitive to the following
  problems than laser systems
• Refractive indexinhomogeneities
• Sample non-uniformity
• Poor mixing in tank
• Convection currents
• Also problematic
• Bubbles
• Floaters
• Scratches
• All optical CT scanners tend to exhibit ring
  artefacts.

26
Scanned laser vs. pixelated detector (4)

• Does the laser system have to be slow?

27
Scanned laser vs. pixelated detector (4)

• Does the laser system have to be slow?
• No! Just stop scanning with a stepper motor and
  do it the way they do in bar-code readers!
• Wuu et al. Med. Phys. 30(2), 132 (2003)

28
Scanned laser vs. pixelated detector (3)

• Does the laser system have to be slow?
• No! Just stop scanning with a stepper motor and
  do it the way they do in bar-code readers!
• Wuu et al. Med. Phys. 30(2), 132 (2003)
• Van Doorn et al. Austral. Phys. Eng. Sci. Med.,
  28(2), 76 (2005)

29
Scanned laser vs. pixelated detector (3)

• Does the laser system have to be slow?
• No! Just stop scanning with a stepper motor and
  do it the way they do in bar-code readers!
• Wuu et al. Med. Phys. 30(2), 132 (2003)
• Van Doorn et al. Austral. Phys. Eng. Sci. Med.,
  28(2), 76 (2005)
• Conklin et al. DOSGEL 2006

30
Scanned laser vs. pixelated detector (3)

• Does the laser system have to be slow?
• No! Just stop scanning with a stepper motor and
  do it the way they do in bar-code readers!
• Wuu et al. Med. Phys. 30(2), 132 (2003)
• Van Doorn et al. Austral. Phys. Eng. Sci. Med.,
  28(2), 76 (2005)
• Conklin et al. DOSGEL 2006
• Krstajic and Doran Phys. Med. Biol. 52, N257
  (2007)

31
Scanned laser vs. pixelated detector (3)

• Does the laser system have to be slow?
• No! Just stop scanning with a stepper motor and
  do it the way they do in bar-code readers!
• Wuu et al. Med. Phys. 30(2), 132 (2003)
• Van Doorn et al. Austral. Phys. Eng. Sci. Med.,
  28(2), 76 (2005)
• Conklin et al. DOSGEL 2006
• Krstajic and Doran Phys. Med. Biol. 52, N257
  (2007)
• MGS (Maryanski) DOSGEL 2008

32
Two geometries of CCD scanners

• Cone-beam
• Simple and inexpensive
• Easily scalable
• Like x-ray CT in reverse
• Correction for refraction effects at wall
  required
• More sensitive to scatter

• Parallel beam
• Light path through sample is simple
• Scalable, but at a cost
• Focussing optics very flexible see Krstajic and
  Doran PMB 51, 2055 (2006)
• We expect it to be less sensitive to scatter.

Image source Kevin Jordan, DOSGEL 2006)
33
Applications (1) External beam RT

• Collaboration with David Bonnett, Maidstone
  Oncology Centre
• 2 x 2 cm2 square field, 10 MeV photons
• Depth dose measured to within ?2 (best
  measurement) and ?6 (worst measurement)

34
Applications (2) Brachytherapy

• Collaboration with Abdul Ismail, Royal Surrey
  County Hospital
• Amersham Model 6711 OncoSeed, 125I, LDR

35
Applications (3) Proton therapy

• Collaboration with John Brunt and Andrzej
  Kacperek, Clatterbridge
• 62 MeV proton beam for ocular therapy

36
Applications (4) IMRT verification

• Kevin Jordans group, in collaboration with RPC
  (Houston)

37
Applications (5) Respiratory-gated RT

• Samuel Brady and Mark Oldham, Duke University

Axial plane
Coronal plane
Eclipse
EBT
PresageTM
38
Applications (6) Commissioning IMAT

• Collaboration with Philip Wai and Jim Warrington,
  Royal Marsden

39
The competition

• The alternatives to gel dosimetry are getting
  more sophisticated!

• Delta4 phantom from ScandiDos
• Two diode arrays arranged in cross formation
• Real-time calibration of IMAT and IMRT,
  beam-by-beam

40
Conclusion

• Lots of people independently had the idea to do
  optical CT when the technology to do it with
  became available.
• Three main chemical systems Fricke-xylenol,
  polymer, PRESAGE
• Two broad classes of hardware laser and CCDEach
  has a number of possible geometries.
• Good results can be obtained with all of these
  combinations.
• Optical CT is suitable has been tested for many
  applications.
• Still artifacts we are some way behind MRI in
  scanner development. Lots more work needed! QA!!

41
And finally ...

• Between 2006 and 2008, 78 full papers on gel
  dosimetry were published (Web of Knowledge).
• 34 used MRI
• 34 used optical techniques
• 10 used other readout techniques.