THE USE OF DESKTOP SCANNERS IN DOSIMETRY: MEASUREMENT PROTOCOLS
W.Zealey1
1 Centre for Medical Radiation Physics, University of Wollongong
INTRODUCTION: Radiographic and radiochromic film are being increasingly used for quality assurance and radiation dosimetry in radiotherapy. Desktop scanners provide a low cost option for scanning such films. Suchowerska, Davison, Drew and Metcalfe1 (1997) evaluated radiographic film for use in dosimetry and identified sources of inaccuracy. Most recently Yang2,3 (2005) has developed quality assurance protocols for using radiographic film which yielded accuracies of 2-4%. In parallel with work on radiographic film International Specialty Products, the vendors of GAFCHROMIC film, have published measurement protocols for their transparent and reflective media. In this paper we evaluate the use of scanners in reflection and transmission mode for digitising film and establish procedures to compensate for desktop scanner’s nonuniform illumination and variations in response over time.
METHODS: A Microtek Scanmaker i700 desktop scanner was used to scan in both reflection ( GSVtransmission) and transmission scanning (GSVreflection) to produce 16 bit greyscale images. Flat field scans were made to assess the spatial uniformity of the scanner. Transmission and reflection step wedges manufactured by Stuffer Industries were digitised using both transmission and reflection modes. All images were analysed using ImageJ to provide mean Grey Scale Values [GSV] and standard deviations for each step. Plots of GSV against Optical Density were compared for reflection and transmission measurements. Each step wedge was digitised at the top, centre and bottom of the scanner’s field of illumination and the variations in GSV values for each step evaluated
RESULTS:.Variations in the flat field for both the reflection mode and transparency mode of measurement were less than 4%. Variations are seen to occur at the edge of the scannable area when the scanner is used in reflection mode, possibly due to reflections within the scanner.
The GSV-Optical Density plots for the calibrated film wedge show that the GSVtransmission is prportional to the film’s transmission, T, and the GSVreflection is proportional to T2. The use of a desktop scanner to measure film in reflection mode therefore reduces the dynamic range but increases the contrast of the measurements. The position of the film within the scan area has a minor effect on the GSV for a given OD.
DISCUSSION & CONCLUSIONS: The uniformity of the flat fields clearly depend on the reflective backing to the scanner lid or the cleanliness of the transparency illumination panel
When using scanners in reflection mode the use of a photoquality backing sheet is recommended. The backing sheet should be printed with a horizontal greyscale calibration across the top and bottom and 1cm uniform vertical strips at the extreme left and right edges. GSV measurements should be normalised to the extreme values of the greyscale calibration.
Variations in illumination and detector response across the scanning bar are corrected for in the scanner self calibration process. Changes in background illumination due to scattering within the scanner can be corrected ofr by obtaining flatfield scans prior to digitising film and reflective media. These flatfields should be subtracted from subsequent digitised images.
REFERENCES:
1 N. Suchowerska, A. Davison, J. Drew, P. Metcalfe, (1997) Australasian Physical and Engineering Sciences in Medicine, Vol.20 No1: 20-27
2Yang Wang, (2005) PhD Thesis University of Wollongong “Digital Film Dosimetry In Radiotherapy “
3Y. Wang and P. Cross, (1999) Australasian Physical and Engineering Sciences in Medicine,, Vol 22 No. 1