In a typical X-ray center, the end of the beam line is equipped with an X-ray horn in which the electron beam is scanned on an X-ray converter to produce the X-ray field. The scan horn may also include a pseudo-parallel magnet to adjust the divergence of the electron beam to offer a focusing or defocusing effect.

The calibration of those magnets is typically performed by measuring the X-ray profile along the scan direction. Methods are proposed in ISO/ASTM 51649 for electron beam width determination, but users have developed their own method for X ray scan width assessment. In these methods, no distinction is made between the beam width and scan width while those are two distinct quantities. While such approximation was acceptable until now, it should be revisited due to the context evolution. In particular, the emphasis put on Monte Carlo simulations, to predict dose in product and to ‘calibrate’ X-ray sites, forces manufacturers to calibrate the magnets to the physical scan width. Failure to do so may generate confusion and hurdles when comparing measurements with simulations. Also, the trend towards parametric release might benefit from a closer look at this variable which has a direct impact on the dose.

Measuring accurately the scan width in X-ray is more difficult than it may appear. Difficulties come from the non-accessibility to the electron beam, possible misalignment of the detector with the beam, contamination of the X-ray field by electrons, non-uniformity of the beam etc.

This paper will present a method for determination of the physical scan width, its calibration as well as the pseudo-parallel magnets setting to minimize the different sources of inaccuracy. Experimental work was performed on the 7 MV X ray beam line of the Feerix® facility of Aerial.