Could someone explain the concept of detectors in CT scans?

Question

What happens if a detector is more sensitive?

Answer 1

Oh mercy, is this EVER my question...

There have been several different detector systems used successfully on CT systems, 3 that I'm at least somewhat familiar with are the xenon gas tube, photomultiplier tubes, and the solid state detector.

Tho I don't believe they're used anymore, the xenon gas tube was used by at least a couple of major manufacturers, including GE in their 8800 and 9800 series machines (c. 1980-1990). The tube was a large metal assembly with an aperture for each detector cell. I don't recall precisely how it worked, but I believe that the tube was highly pressurized with xenon, a relatively dense gas that was made even denser by pressure. This density improved the chance of an x-ray photon striking a gas molecule and thus causing a flourescence. I couldn't tell you the exact method of data amplification and read, it was not a machine I was very familiar with.

Photomultiplier tubes with an x-ray fluorescent crystal ("BGO" crystal comes to mind, but I can't recall exactly what that stood for) were used by the Picker in the 300/600SD series c.1978-1982. The SD600 had 600 such stationary photomultiplier tubes. Each output tube fed a current-to-frequency transducer, and 150 of these I/F were multiplexed into counters which output the counts for a variety of integration periods in the low to mid milliseconds range.

In the mid 1980's Picker went to solid state detectors. They were a slab of silicon about 1 by 4 cm, coated with, if I remember correctly, cadmium tungstate, an x-ray flourescent compound. This diode output a current in the range of a few picoamps to a few microamps, which, like the SD, was fed to an I/F converter. Also like the SD, these frequency outputs were fed to counters.

I believe that most current machines now use solid state detectors such as just mentioned.

Re detector response, there is a very impressive set of corrections done to detector data before it even gets close to the image processing hardware. All the detectors basically have different gain and offset levels, values for which are computed during a reference "air scan" prior to patient entry, as well as modification values to each applied during the patient scan by looking at the "wings" of the scan data which should have nothing but air in their path even tho a patient is on the table.. There is also background subtraction (which is not the same as offset). Slight variations in the kV and mA of the x-ray source during the scan also cause faulty detector response, and also must be corrected for. There are centering constants applied, for both detector position relative to frame and relative to x-ray tube since it is nearly impossible (or at least, excessively expensive) to have mechanically perfect geometry in such a large machine.

And if a detector just refuses to give a correctable output no matter how many corrections are applied to it's data, an operator or service tech will put it on "the bad list", Software will then estimate values for it based on adjacent detector data before reconstructing the image.

Answer 2

A CT scanner is a fancy X-ray machine, typically with a single X-ray source and a number of detectors (such as photomultiplier tubes) that are sensitive to X-rays. The attached computer captures the output of the PM tubes, and does some fancy mathematics (it's called a Fourier transform) to generate the resulting picture. The math tends to swamp out variations in the response of individual PM tubes, and a separate computation could be done if necessary to clean things up further.