Can anyone explain why x-rays don’t look like an ordinary photograph?
X-rays aren’t fuzzy, you can and have to be able to make out very five detail. They look the way they do because the objects you’re typically shooting are half transparent (to X-rays).
Unlike regular photographs where you have an image focused by a lens, an x-ray is more like a shadow where there can be a significant distance between the object being imaged and the film/sensor. This means the rays will spread out somewhat rather than being focused to a point.
An X-ray image is formed on either film or a sensor by X-rays that travel directly from the anode of the X-ray tube source. X-rays cannot be usefully focussed in diagnostic machines, and so the smallest feature you can resolve on an image is limited by how small you can make the spot on the anode. (In many ways a simple X-ray image is formed much like a pinhole camera, except that the spot on the anode is the pin-hole, and the subject is on the same side as the image.)
The problem with getting the spot on the anode small is that the production of x-rays is pretty inefficient. A huge amount of energy is dissipated as heat on the anode. Since the anode is sitting is a hard vacuum the only ways of getting the heat out is thermal radiation or conduction though the metal. The geometry of the anode is also optimised so that the effective spot is as big as possible as seen by the source of electrons, by thinner when seen by the subject. The anode is typically a disk that is spun, so as to even out the heat load, and is made of high temperature materials, such as tungsten. Even so, the spot can only be made so small before the point where the x-rays come from melts.
Reconstructed x-ray images - such as from a CT scanner are further limited by the resolution of the detectors.
Finally an X-ray is a shadowgram. All the features in the 3D subject appear projected onto a 2D image. This will lead to indistinct outlines simply due to the geometry.
In reality X-ray images can be pretty sharp. You can see remarkably fine detail - like hairline fractures, or with contrast media (X-ray opaque fluid either drunk or injected to outline vessels or the gut) you can see quite small features. But most of a human is squidgy stuff that is not all that opaque to X-rays and has a 3D shape that is not going to translate to sharp outlines. This is probably the main issue when the images are perceived to be fuzzy. Interpreting X-rays includes understanding this. You are not looking at something that is equivalent to photograph. Think of it more like trying making a bowl of jelly, with lumps of jelly of a different colour added. You can hold the bowl up to a light (but you can’t move it around or rotate it, and you can only look with one eye.) You have to work out what shape the lumps of the different coloured jelly are.
I’ve wondered if it’s because X-ray photons are higher energy, so for the same power, there’s fewer of them than for visible light, so they’re more grainy? Could that be true? Also medical X-rays are probably operating at as low power as possible, to reduce risks.
I was wondering why X-ray analysis methods are so noisy and “counts” based, eg XRD, but say, UV-VIS isn’t.
The problem with X-ray photons is catching them at all. Film bases systems used a sheet (actually two - one each side - as X-ray film is doubled sided) of an X-ray luminescent material that converts the X-ray photons to visible light - which exposes the film. Without that, the capture of the photons is much less efficient. Real time imaging systems also use a screen of florescent material and then use light detectors to build the image. Actual efficiency is very good.
There are two parts to the X-ray dose. The energy of the photons and the number. The energy is tunable by changing the voltage on the anode of the tube. This is important as it changes the image contrast depending upon what it is you want to see, and which part of the body you are imaging. With lower energy photons you can differentiate some tissues more easily, but you want higher energy to outline broken bones (for instance). Then you vary the forward current in the tube to vary the actual flux of photons, and finally time the pulse. There is clearly a lot to be gained with better detection to lower the overall dose. CT scanners have made astounding strides here. The early ones were pretty bad, the latest ones very good indeed.
The so-called digital X-ray systems use a sheet of material that directly captures X-ray photons, and can be read out (it is very neat - the X-ray photon casues the material to move to a higher energy state, and then an laser of just the right energy causes the material to drop down to the base level again but emit a photon which can be detected. So it uses a machine that is a glorified flatbed scanner to read the X-Ray.)
Grain is a manifestation of signal to noise in a spatial world. It works with the same rules. So indeed, the grain is a function of dose. The dose will be limited to the point where an acceptable image is formed, and the effective grain is one of those determinants. Just fine enough grain and good enough contrast range, work backwards via the film/detector sensitivity and you have your dose. Because the conversion efficiency is very good, some of the grain/noise probably is a result of quantisation to the individual X-ray photons being detected.
Thank you very much team, very interesting and my question is answered.
A normal photo is composed of reflected light whereas an x-ray photo is of blocked light (high energy photons).
How bout them gamma rays thou? Just a tiny dose don’t want to burn any holes.