In helical CT, overranging is the difference between total exposed scan length and total imaged scan length (cf.
). To assess overranging, several approaches exist; some rely on displayed data at CT console [1,2], while others measure the dose profile using computed radiology (CR) imaging plates or Gafchromic films [2,6,7,8], or even through in-air ionization chamber measurements [2,5,9,10,11,12] or finally considering more elaborated measurements suitable for modern tomographs equipped with dynamic z-collimation [13].
In the present work, overranging length (LOR) was defined and computed as the difference between total exposed scan length (Lexp = DLP / CTDIvol) and total imaged scan length (Lim = [( |first slice position – last slice position| ) + section width]) following Schilham et al [1].
C2i santé database
The C2i santé database was built under Microsoft Excel® (16.0.14026.20202). All data was manually recorded by seven medical physicists between January 2017 and December 2021, containing information on the hospital name, anatomical site of CT examination, patient demographics (age, gender), CT equipment (model, commissioning year, slice number), acquisition parameters (kV, mA, rotation time, pitch, beam collimation, section width, scanning length, iterative reconstruction level) and exposure data (CTDIvol, DLP).
From the C2i santé database, the three most frequent CT scans currently used in France and considered for this study are shown in the
.
Overranging values for head and chest procedures were computed from the C2i santé database based on clinical acquisition parameters and subsequently compared to measured and computed values from the benchmarking exercise.
Overranging measurement using Gafchromic films
For these experiments, overranging was tabulated considering different pitch, collimation and rotation time values (cf.
). Certain of these parameters do not correspond to a clinical use, nevertheless this allowed us to study their influence. CTDI
vol and DLP values displayed on the screen were recorded together with the planned start and end positions of the CT table. All others acquisition parameters (kV, mAs, rotation time, section width, beam collimation) were collected from the main console (DICOM tags). Scans were performed using lengths around 8 cm for every acquisition.
XR-SP2 Gafchromic films (Ashland Specialty Ingredients G.P., NJ, USA) were used to measure overranging values. Experimental setup involved the use of an in-house wood (extender) and foam (film) support which helped placing the Gafchromic film at the CT isocenter while reducing backscattered radiation (cf.
). The film was cut into 25.4 cm-long and 1.5 cm-wide strips.
Gafchromic films were read using a multifunctional scanner (Xerox, AltaLink C8030, 2017) following the method introduced by Farah et al. [14] to reduce scanner-related uncertainties. Scanned films were analyzed using ImageJ software (Wayne Rasband, National Institute of Health, US, version 1.53e) [15] where the dose profile was first plotted. Next, an in-house Python (version 3.9.6) script was used to compute overranging as the difference between the FWHM (Full Width at Half Maximum) and the imaged length of the dose profile following Trevisan et al. 2014 [2] (cf.
right). For this goal, a Savitzky–Golay filter [16,17] was first applied to the dose profile plotted with the purpose of smoothing the data. Subsequently, the imaged length was determined by the script as the inflection point of both sides of the dose profile plateau. Each film was scanned 20 times and the subsequent ImageJ and Python analysis was repeated to ensure consistency. Median overranging value obtained over each of the 20 readings was compared to the value computed following Schilham et al 2010 [1].