By convention, all X-ray beams of energy greater than 1 MV (1 million electron volts) are classified as megavoltage. This range of beam energies used in clinical practice are all produced by the linear accelerator, or in the trade, the LINAC. The linac is a wonderful piece of machinery and owes its entire existence to the early quantum physicists of the early 20th century, and to lots of engineers who took these theoretical findings and produced things such as nuclear reactors, atomic bombs and …. the LINAC!
The linac uses high-frequency electromagnetic waves (microwaves actually, just like those that heat your coffee each morning, but more tuned to doing the job required!) to accelerate electrons to high energies through a linear tube containing a vacuum (there is that vacuum again!). This highly accelerated and fast (0.99c) electron beam can be used as a stream of electrons, or the stream of e- can strike a tungsten target to produce x-rays (almost exclusively bremsstrahlung x-rays).
The components of a linac consist of a power supply that provides DC power to modulator (I feel like I'm singing "Them Bones"! If you are interested in the words & music!. The Modulator has a pulse forming network which stores electrical energy to provide flat topped DC pulses to the thyratron. The thyratron uses these pulses as a high-tech switch to delivers the pulses to the electron gun and simultaneously to the magnetron/klystron (the linac will have either the magnetron or klystron, not both). The electron gun (you have one of these in the back of the old CRT television - the one's you watched in your childhood before plasmas and LCDs became a moderate excess and necessary component of modern living) produces a stream of electrons that enter the proximal part of the wave guide with an energy of about 50keV. The magnetron/klystron produces pulsed microwaves which are introduced into the wave guide by a hollow rectangular copper piping system which happens to be filled with $SF_{6}$. The wave guide is a copper tube with the interior divided by copper discs or diaphragms, and is evacuated to a very very high vacuum by an ion pump. The ejected electrons interact with the tuned microwave produced by the magnetron/klystron, absorb energy (like the water in your coffee does each morning) and so are accelerated. The wave guide has a fancy arrangements of side pockets which have the effect of continuously accelerating the electron down the entire 1-1.5m length of the wave guide. The high velocity, and therefore, energy electrons then exit the thin ceramic window at the end of the wave guide (you have to maintain the vacuum!) in the form of a 3 mm 'pencil beam'. The beam then travels straight.
What happens then depends on the construction of the linac. Linacs that produce 4MV or less only need a short wave guide which can sit in line with the target and eventual photon beam direction (remember you can change the direction of an electron - that's how you get TV! - but you can't change the direction of a photon unless you have the mass of a Sun or a black hole. For linacs that are producing beams of 6MV or higher, the wave guide is too long and so it becomes more helpful to bend the electron pencil beam. Most linacs have a device called a 'bending magnet' to change the electron beam's direction. It does this by applying a magnetic field (for those a bit light on high school science, moving electrons generate a magnetic field, and so if they enter a magnetic field they will spiral according to the right hand rule, … or is it the left hand rule? Look it up!
The bending can be either through 90 or 270o. The problem of the 90o bend is that the pencil beam will smear and become 'chromatic' - let me explain! All the electrons will not have identical speeds when they exit the wave guide. Some will be faster and some slower. On entering a magnetic field attempting to change the direction of flight by 90o, the faster electrons will bend less than the slower electrons and so the 'pencil' will now look decidedly oblong with all the slow electrons acute to the 90o angle and all the faster electrons obtuse to the 90o angle. This produces a spectrum of electron, hence the use of the term 'chromatic' which derives from the splitting of light into different frequencies by a prism.
There is a very tricky way of minimising this effect (damned clever physicists!!), and this is by turning the beam through 270o (turning by 180o or 360o would seem useless!). When turning the beam through 270o, the additional thing that happens is that the slower electrons take a tighter circle and the faster electrons take a wider circle (as described before). But here is the really helpful bit, the combination of less speed and less distance and faster speed and more distance counter-balances, so that after a 270o turn, the faster and slower electrons arrive at almost the same time. Which is analogous to me getting ready for a ball and travelling via the next capital city to the ballroom and reaching there the same time as my wife who travelled directly from home to the ball! So now the 'chromatic' effect has been removed, and the beam is described as 'achromatic'.
After exiting the bending magnet 'race track' within the treatment head, the electrons continue in a straight line. The first item that the electron beam meets depends on the selection of beam required. If electrons are required, the electrons will meet a scattering foil. This is designed to spread the electron pencil beam (like a laser) into an electron beam (like a torch light). If photons (the insider's name for the X-ray!) are required, the electron beam will meet a target made of tungsten. because of the velocity of the electrons at this point, the bremsstrahlung production will be anisotropic and largely forward projected. It is necessary to shield the treatment head with lead because of the photons projected and scattered laterally. Following the target/foils, there are two ion chambers which monitor the beam's dose rate, integrated dose and field symmetry. This is to prevent the use of too much radiation and radiation beams that do not meet expected profiles. (Remember there are people under these beams so we don't want to fry them!) Below these chambers are the primary, secondary and sometimes tertiary collimators. The primary and secondary collimators are large lead blocks about 15cm thick that move on tracks to open an rectangular aperture for the radiation to escape. The actual construction of these collimators varies greatly between manufactures (as do some of the components and arrangements already described).
| Manufacturer (A..Z) | Primary Collimator | Secondary Collimator | Tertiary Collimator |
|---|---|---|---|
| Elekta | focused MLC | focused block | none |
| Siemens | focused block | focused MLC | none |
| Varian | focused block | focused block | unfocused MLC |
In addition to these beam defining devices, the treatment head also has a field light system which duplicates the radiation beam size so as to make set up easier and more accurate. There is also an ODI or optical distance indicator which is used to define skin source distance.
All of this is mounted on a gantry which can rotate around the patient. More than this there are at least three permitted movements of the linac - rotation of the gantry like a windmill, collimation in the head like an electric mixer, and movement of the couch like a gate. The engineers have deliberately designed these three movements to occur around axes that all intersect at a common point which is situated 100cm from the source (the focal point of the radiation beam). This point is called the isocentre and is very important in modern radiotherapy and therefore it is very important that a regular quality assurance program document frequently that this point is where everyone thinks it is and that the linac is not wearing and making the point change with movement.