[A quality control test of the image obtained with electronic portal imaging devices]

A quality control of the digital image obtained from two electronic portal imaging devices is discussed. The devices are used to verify the radiotherapic treatment setup by comparing online images of the irradiated volume with those of the simulation devices.

Both iView and Target View devices, respectively installed on a dual energy SLi Precise Elekta and Saturne 42 Ge linear accelerators, consist of highly efficient phosphor screen and high quality videocamera, controlled by a workstation, able to generate digital portal images from few cGy doses. A phantom and software package are used to assess the spatial resolution and signal to noise ratio, and to compare tha data obtained. Spatial resolution and signal to noise ratio of both systems were studied as a function of energy, gantry angle and image acquisition parameters.

The mean spatial resolutions obtained from the first 30 measurements were 0,265+/-0,012 and 0,220+/-0,010 lp/mm respectively for 6 and 18 MV of Saturne 42 (Target View) and 0,241+/-0,006 and 0,239+/-0,005 lp/mm for 4 and 6 MV of SLi (iView). Spatial resolution decreases as a function of energy, meanwhile there are no significant statistical differences as regards of the acquisition parameters; signal to noise ratio, instead, increases with integration time. Different values of the spatial resolution as a function of gantry angle are due to changes in the screen-camera distance and flexing of the detector housing. The quality control test is performed every 15 days by the technicians of our Radiotherapy Department. We set the reject level of spatial resolution and signal to noise ratio to be three standard deviations below the mean value obtained during the initial EPID characterization: if the measures fall below these values preventative maintenance is scheduled.

The efficacy of the use electronic portal imaging devices for visualizing and quantifying the relative positions of anatomical structures within the radiation field depends on the image quality. It is therefore essential to devise quality control tests for the devices themselves, to guarantee an optimal level of system performance in a fast and efficient manner.