Proton Calorimetry/Meetings/2025/02/04

From PBTWiki
Jump to navigation Jump to search

Minutes for UCL Proton Beam Therapy Group Meetings, 4th February

Present

Joe Bateman, Sonia Escribano, Simon Jolly, Raffaella Radogna (via Teams), Kirsty Hall (from Hamamatsu)

  • Discussion with Kirsty Hall from Hamamatsu about photodiode arrays for scintillating fibre array to allow for better integration with QuARC.
  • Kirsty to investigate two possible routes for going forward:
    • Possibility of having on board DDC or way of connecting S13885-128 photodiode array to DDC board.
    • Possibility of having surface mounted photodiodes (such as S12859-021) shrunk to a smaller size so that pixel pitch is comparable with 0.5mm fibre diameter.
  • Also will look into questions Raffy has a about current experience with S13885-128 photodiode array, i.e.:
    • If observed response at saturation (e.g., decreasing signal rather than plateau or charge sharing among neighbours) is expected
    • How response of individual photodiodes within array vary and how to characterise this.

Raffaella Radogna

  • Discussed with Bari mech eng Producing second prototype with resin material harry suggested - use inventor files (should be able to open fusion) will share with them once harry has finished with the fusion file
  • Produce 2 new prototypes (frame + fibres + screws etc) and have one sent here
  • Beam test in Trento
    • ~5s per beam
    • 10 beam deliveries each for CONV and UHDR calibrations (but may only need 8)
    • Background measurement at FSR for 12.5pC acquisition and for for FLASH
    • Happy to cut out as many pristine Bragg peaks as necessary to keep within FLASH time limits at Trento
    • I.e. 1 minute = 12 Bragg peaks at ~5s per measurements
    • Or Bragg peaks at lower measurement time per FLASH Bragg peak (i.e. <5 s)
    • Spot scanning at CONV dose rates
    • Potentially WET measurement of fibres at next UCLH if using giraffe on new QuARC electronics


Sonia Escribano

  • Looking at new calibration method - now works with txt file that is generated with calibration values
  • Problem with calibration
    • Need calibRMS - need separate file with these values?
      • Important for the fit
      • Include rms beneath calibration coefficients
    • How do we save name of header file for calibration file
    • Include in header file which files were used to generate calibration coefficients
  • In new file structure maybe have new file directory for background data, i.e. calibration, background, date
    • For now do it manually
  • New calibration and fitting procedure
    • generate calibration txt files using shoot throughs
    • Apply calibration values generates calibrated txt with every single measurement
    • Fit data averaged at 25 Hz -read calibrated data and fit at desired frequency (default 25hz)
  • Things for new experiment
    • Raw data = 160*8 (hex length and 8 modules)
    • Calibration: 4 calibration_i files
    • Calibrate data
    • Initially split the raw data in two files for each Bragg peak
  • Messaged harry about scintillators waiting to hear from John and Wayne but still hasn't sent them scintillator sheets yet
  • Simon and Sonia to check final design on fusion
  • How do you ensure consistency between modules if there's difference in coupling etc
    • Take Manchester data and do shoot-through calibration in current way then apply it to one of the shoot-through calibration files and see how it looks for each module (i.e. treat each module independently)
    • Need to find a way to scale to dose in each of the modules


Joe Bateman

  • Re-did simulations in TOPAS (ongoing) and FRED with adjusted input beam size from DICOM for 3x3 and 5x5 cm box fields
    • Calculated new input beam size as: σ_P = (σ_iso × (SAD - d)) / SAD Where:
      • σ_P is spot size at position P
      • σ_iso is spot size from DICOM (at isocenter)
      • SAD is virtual source-axis distance
      • d is distance from isocenter to position P
    • Using d as 10 cm upstream from phantom surface
  • Resulting shape looks more characteristic of TPS distribution.