Computed Tomography 1

Computed tomography, or CT for short, acquires images from delivering x-rays from multiple directions around the patient and detecting the intensity of the x-rays once transmitted through the patient. A cross-sectional layer of the patient is divided into small volume elements termed voxels. The linear attenuation coefficient of each element is calculated and expressed as a grey scale.

The Hounsefield Unit or CT number takes the linear attenuation coefficient of a voxel and normalises it to that of water. Only after the CT number is calculated is a grey scale value assigned. CT numbers for specific tissues are exampled below:

• Air: -1000
• Water: 0
• Bone: 1000
• Lung: -300
• Fat: -90
• White matter: 30
• Grey matter: 40
• Muscle: 50
• Trabecular bone: 300-400

Generations of CT

     1.   First generation: translate - rotate

In first generation CT, a pencil beam is used with a single x-ray tube and single detector. The beam travels perpendicular to the detector through the patient. The x-ray then moves to the side and delivers another exposure, continuing the 'sweep' until ~160 exposures have been made. The x-ray tube and detector then rotate 1 degree and perform another 'sweep'. This continues until 360 degrees of information has been acquired. First generation CT has a duration of approximately 5 minutes and does not cope well with motion.

     2.   Second generation: translate - rotate

In second generation CT, the same principle is applied but for a narrow fan beam and a small array of detectors (~30). The sweeping motion is applied again, however fewer individual exposures are made because of the fan geometry. The duration of this scan is 15 seconds.

     3.   Third generation: rotate - rotate

In third generation CT, the most commonly used method, a larger fan beam is used with a larger array of detectors. The sweeping motion is lost because the fan beam now covers the area of the patient and only rotational movement of the x-ray tube and detectors is required. There is one scan taken per rotation, with a total duration adding up to 1 second. Third generation CT produces less artefact due to a shorter scan duration.

     4.   Fourth generation: rotate - fixed

In fourth generation CT, there are 360 degrees of fixed detectors, requiring only rotational movement of the x-ray tube. Again a fan beam is used, but the detectors are costly. Thus, fourth generation CT is uncommon.

     5.   Fifth generation: electron beam CT

There is no mechanical motion in fifth generation CT. Instead there is a curved tungsten target. Magnetic fields steer electrons towards the tungsten target. Multiple, closely spaced targets can be employed to get separate slices in quick succession. This is particularly good for imaging the beating heart, as it has a quick duration of 50ms. Ultrafast CT is also common. However, it is again relatively expensive and third generation CT is preferred. 

Slip Ring Technology

Previously electrical cables used to be responsible for gantry rotation, however they did not allow free, continuous 360 degree rotation. Slip ring technology allows for continuous rotation, with transmission of both power and data from stationary mount to the rotating gantry. Thus, without slip ring technology, helical scanning would not be possible.

The CT X-Ray Tube

The x-ray tube used in CT is often up to 10 times more powerful than that of general radiography. The tube is only for a longer duration of time in helical CT and thus more heat is produced. The tube must have efficient heating and cooling capacities, and so not only does it have a rotating anode but the body itself rotates.

The anode to cathode axis, or heel effect, runs parallel to the z axis of the scanner, or the long axis of the patient. However, as only a slice taken in the centre of the beam is of importance, the heel effect and any induced spectral changes along the beam are eliminated.

X-ray scanners in CT use 80-140keV, depending on the vendor. The x-ray spectra in are hardened in CT using significant thickness of aluminium, somewhere between 5-10mm.

Bow Tie Filter

Most areas scanned by CT scanners are approximately circular in size. This means that the photon fluence at the periphery is higher than at the centre. Thus, a beam shaping filter is used to reduce the dose delivered to the periphery with no effect on image dose. Based on the geometry of the filter, the beam increasingly attenuates x-rays towards the periphery.

Detectors used in CT

The detectors used in CT are scintillating solid state detectors, mostly made of gadolinium oxysulphide. The crystal is sintered to increase its density, producing greater light output from lower signals to help reduce patient dose. The light produced by the crystal is converted to charge by a photodiode, which is converted into a digital number by analogue-digital converters.

An opaque filter is placed between the detector elements to prevent optic cross talk, and all detectors are mounted on an aluminium frame.

Single Detector Computed Tomography

Single detector CT consists of a single detector, typically 10-15mm wide. Slice thickness in SDCT is determined by collimator width, which varies between 1-10mm. 1mm is thin and 10mm is thick. Thin slice acquisition provides high spatial resolution but scan duration is longer.

Multi Detector Computed Tomography

Multi detector CT consists of an array of detectors, allowing acquisition of multiple slice thicknesses in one scan. For example, a 0.5cm detector can be binned into a 2cm detector by combining with 3 other 0.5cm detectors. Image acquisition is faster in MDCT and thinner slice thicknesses are possible. Slice thickness in MDCT is determined by detector configuration and x-ray beam width is determined by collimator width. 

Scanned Projection Radiograph

A scanned projection radiography is a preliminary scan taken of the patient with both the x-ray tube and the detectors stationary, but the patient moving through the gantry. It is also called a tomogram, localiser, scanogram, scout view or CT radiograph. The scanned projection radiograph can be used to measure patient parameters which help determine kVp, mAs, mA modulation, pitch etc. 

Axial Scan

An axial scan is done by turning the x-ray generator on and off, and moving the table when the x-ray tube is deactivated. It is a basic step and shoot mode, where there is a rotational speed of approximately 0.5 seconds. The x-ray tube is activated and 360 degrees of data is acquired. The tube is then deactivated and the table moves into the next position. The tube is then re-activated and another 360 degrees of data is acquired.

Helical Scan

In a helical scan, the x-ray tube is always activated and the table moves continuously through the gantry, forming a helix of radiation around the patient. The helical scan is much faster, but because the x-ray tube is always on, generates a lot more heat and requires greater heat dissipation. A large volume of anatomy can be scanned with little effect of patient motion on image quality. However, because of the helical geometry of the beam around the patient, some data is lost in between and interpolation is required. Duration of the scan is decreased from 1.5 to 0.3 seconds.

Pitch

Pitch is defined as the table travelled per rotation divided by beam width. It should be between 0.75 to 1.5 for CT scanning.

• A pitch <1 means overscanning the patient because the scan itself is low and dose is increased.
• A pitch >1 means under scanning the patient because the scan itself is fast and dose is decreased.

Faster gantry rotation results in greater dose deposition because pitch is decreased.

Pitch is inversely proportional to dose. This means a higher pitch is preferred for decreasing patient dose. However, lower pitch results in higher spatial resolution because high pitches require interpolation (can result in artefacts). Thus, choosing pitch is a trade off between patient dose and image quality.