Compared with conventional X-rays, the method of computed tomography (also: computed tomography; CT) is comparatively young, but it is hard to imagine clinical routine without it. Its versatility and rapid technical developments make it indispensable for a wide variety of problems in almost all regions of the body. Can X-ray measurements taken from different projection directions be combined in such a way that they provide a complete, superposition-free image of a body layer – similar to a jigsaw puzzle?
X-rays: Image of the interior
In conventional X-rays, the rays are sent through the body and – depending on how much they are transmitted by different tissues – reach the other side. There, they are recorded by a kind of photographic plate. A two-dimensional image is obtained, similar to a silhouette on the wall, in which the various structures are superimposed.
What is lost is the information at what depth they are located. This is a crux that can be partly solved by taking images in different projection planes – for example, from front to back and from left to right. Computed tomography also uses X-rays, but solves this problem in a different way.
How does computed tomography (CT) work?
The difference between CT and traditional imaging is that CT images the body in thin slices. Each of these slices, which are only a few millimeters thick, can be assigned to exactly one location in the body – as if it had been cut crosswise a thousand times with a sharp knife.
But the device can do even more: the images can be post-processed, enlarged, measured, stored and viewed from different angles. And – particularly helpful – a spatial image can be assembled from the sectional images if required, which can be viewed from all sides and allows doctors to precisely assign and extend structures and their surroundings, for example, in preparation for an operation. To obtain such thin slices, a fine beam of X-rays is sent through the body and collected by detectors on the other side.
Different types of CT
The trick is that the CT machine rotates once around the patient during the examination, taking a great many measurements. These are transmitted to the computer, which stitches them together – according to the differences between the intensity of the beams sent and the intensity of the beams received – to create a cross-sectional image with different shades of gray.
The device is then moved a small distance along the patient and the process is repeated layer by layer until the desired area is scanned. This conventional technique is also known as incremental CT. During the scans, the patient must lie still and adjust his or her breathing movements to the staff’s instructions so that the image is not blurred.
The newer machines work even more efficiently by having the tube move continuously in a spiral shape around the patient (spiral CT), often firing multiple units of X-ray beams picked up by multiple rows of detectors (multi-detector CT = multi-slice CT). This allows large sections of the body to be scanned very quickly and with high resolution, an advantage especially for moving structures such as the heart.
History of computed tomography
The mathematician Radon proposed a theory as early as 1917, and its converse enabled the physicist Cormack to find a computational solution to this problem in the early 1960s. The electrical engineer Hounsfield took advantage of this finding and developed a machine with which he scanned the brains of pigs and oxen starting in 1967. In 1972, the brain of a human being was examined for the first time, and the triumphal march of computer tomography began. Cormack and Hounsfield received the Nobel Prize in Medicine in 1979 for their pioneering work.
The first prototype of a computer tomograph still took nine days to acquire and two hours to calculate 28,000 measurements. Today’s devices manage to process hundreds of thousands of measurements in just a few seconds, and it takes between two and ten minutes to examine the head, for example.
