Session Item

Measurements were conducted on two LINACs: a TB STX, equipped with PerfectPitch Couch with six degrees of freedom (DoF), and a TB with a four-DoF couch. A Gafchromic EBT3 film was located on the couch at SSD 100 cm and irradiated using an asymmetrical field collimated by jaws with X1=-0.5, X=0.5, Y1=-6, Y2=7 cm, with couch rotation of 270°, 315°, 0°, 45° and 90° around the Y-axis (IEC 61217) and collimator angles of 240°, 0° and 120°. After exposures, the film showed fifteen dark not-overlapping regions of about 1x1 cm², visually arranged in a circle with radius of 6.5 cm (fig. 1). An Epson 10000XL scanner was used to obtain a digitized image of the film with a resolution of 150DPI, corresponding to a pixel spacing of 0.17 mm. A homemade software (Matlab, The MathWorks Inc.) was used to apply a threshold to the film image and identify fifteen regions of interest (ROI); the beam central axis position at film plane was automatically calculated as the barycenter of the ROIs irradiated with the same couch rotation. The mean distance between beam central axis positions and couch rotation axis, determined through two complementary couch angles (90° and 270°), was assumed as metric to quantify the amount of the induced shift (fig .2). Results were compared to MPC data, performed with the Enhanced Couch option activated. In this configuration, MPC determines the center of rotation through a motion on all available couch rotational axes; the rotation-induced couch shift measured with MPC is the offset of the center of rotation and the radiation isocenter. The sensibility of the film-based method was determined by processing a film having the same ROI pattern, which was obtained irradiating the film with fixed couch angle. The repeatability was evaluated processing three consecutive scans of the same film.

A sensibility of 0.02 mm was achieved by measuring the rotation-induced couch shift for the film exposed with fixed couch angle. The repeatability resulted within 0.01 mm. The table reports the offsets measured with film and MPC, performed in the same day, for three different days of measurement. Results showed a good agreement between MPC and films. The mean difference between the two outcomes was -0.01±0.03 mm and -0.02±0.05 mm for TB STX and TB, respectively.

We demonstrated the validity of MPC tool for evaluating couch rotation-induced shift using an independent method based on films. Since the overall MPC procedure is less cumbersome than film handling, it represents an ideal tool for daily QA, especially when high accuracy in the LINAC geometric setup are mandatory.