Circles are used to sample circular regions of the EPID as an measurement of dose. They are used by the center dose, wedge and symmetry/flatness/constancy tests.
The dose from a circle is calculated by calculating the average CU (calibrated units) of all pixels that lie within the circle as measured in the isoplane.
To determine the center of the circle in the isoplane, the expected center (determined by the test) is added to the collimator's center of rotation, as determined in the Collimator Centering test.
To determine the pixels that should be averaged, the positions of pixels in the image plane are projected back to the isoplane (LINK: using SID and SAD). All pixels whose centers fall within the radius of the circle are used.
Because a circle is being projected onto a rectangular pixel array, pixels on the edge of the circle will only have part of their area within the circle. If the center of the pixel is inside the circle or not, then the entire pixel is used or not; no attempt is made to partially weight the pixel's influence.
Also note that although the collimator's center of rotation is generally a small number, even smaller than the width of the pixels, it may cause some pixels to be include or not included that otherwise would be.
