The linear electron accelerator is a so-called multiple accelerator, in which the electrons obtain a high energy by a multiple repetition of low energy accelerations. The heart of the linear accelerator is a tube that is formed by a sequence of dozens of acceleration cavities. At the input cavity an electron gun fires electrons into the acceleration tube in short pulses with an initial energy of about 50 keV. In travelling down the tube, the electrons gain an amount of energy in each of the cavities until the final energy is reached. The acceleration cavities are resonance cavities in which a strong axial electric acceleration field is built up by short pulses of electro-magnetic energy, at a frequency of about 3000 MHz. The pulse length is about 2 microseconds and the peak energy in the pulse is about 2 megaWatt. Typically the accelerator is operated at 100 to 500 pulses per second. The acceleration tube can vary in length from 20 cm for a small 4 MeV accelerator to more than 2 meters for an energy of 20 MeV and higher.
The illustration shows the interior of a Siemens accelerator in which the vital parts are indicated. A magnetron or klystron electron tube (not shown here) generates the necessary electro-magnetic waves that build up the resonance energy. Hollow metal waveguides are used to feed the electro-magnetic wave to the linear acceleration tube. A vacuum pump maintains a very low pressure in the acceleration tube to prevent the accelerated electrons from colliding with gas atoms, by which they would be scattered out of the beam and into the walls of the tube. From the electron gun the beam travels in a straight line through the resonance cavities and is then bent by the magnet in the direction of the patient and focused onto the target plate. The electrons are absorbed in the target plate and in that process the absorbed energy is converted into an intense megavolt photon beam. In some types of accelerators an electron window can be moved into place instead of the target plate so that the patient can be treated with fast electrons. With a beam filter (equalising filter) the dose distribution within the available field size is made homogeneous. To assist in the aiming of the photon beam at the patient a beam of light can be projected with the use of a mirror, to visualize the path of the exiting radiation beam.
Most linear accelerators for medical use are constructed as rotating structures with the acceleration tube in a horizontal arm (gantry) in which at the end, the electrons are bent towards the patient and hit the target plate in the beam collimator system. The figure shows a linear accelerator where arrows indicate the axes of the rotating parts. The gantry rotates around a horizontal axis. Also the diaphragm mechanism of the collimator ("radiation head") and the treatment table can be rotated. The intersection of these axes is called the isocentre. The positioning of the patient in relation to the isocentre is made easier by laser line projection from the walls and ceiling. The lasers are trained at the isocentre and they project lines that form crosses on the skin of the patient. During the preparation for the treatment, marks are made on the skin where these crosses should be positioned in order to place the tumour exactly in the isocentre.
Linear accelerator with isocentric rotation axes