ABSTRACT
The CT scanner developed by EMI is often called the firstgeneration scanner in which a single detector cell is used to collect the projection signals. After the detector and the x-ray source travel synchronously along a straight line to collect a set of parallel projection samples, the entire apparatus rotates 1° and the above-mentioned process is repeated to collect parallel projection samples along a slightly different orientation. Such data acquisition takes a substantial amount of time, so to speed up the data acquisition process, a second-generation scanner was built in which multiple detector cells were used to collect projection samples of multiple projection angles simultaneously. If six detector cells are used, for example, the angular increment for successive acquisition can increase from every 1° to every 6°, a factor-of-six increase in speed. Despite the improvement, the data acquisition speed was still fundamentally limited by the translation-rotation motion of the CT system, because the cross section of the entire object could not be covered by the small detector size. This limitation led to the development of a third-generation scanner in which a large number of detector cells were used to create a field of view large enough to cover the entire object of interest within the imaging plane. With this configuration, the linear translational motion of the x-ray source and the detector was no longer needed, and the entire data acquisition could be completed with the rotational motion of the gantry. The data acquisition speed was dramatically increased with this type of scanner. The third-generation design, however, came with a series of engineering issues. For example, because the x-ray focal spot and the detector cells were stationary with respect to each other, the projection samples violated the Nyquist sampling criteria that required the collection of two independent projection samples per detector cell. The fixed geometric relationship between the samples collected by each detector cell relative to the isocenter of the CT system also placed a stringent demand on the detector performance in terms of the fidelity of the measurement. Although these design issues have been resolved in later years of CT development, they paved the way to the development of the fourth-generation CT scanner in which the x-ray source and detector are no longer stationary to each other. In the fourth-generation scanner design, a ring of stationary detectors completely surrounds the patient while the x-ray tube rotates about the patient. A single projection view is formed with the readings of a single detector cell collected over time with the x-ray tube at different locations. Although this design overcomes some of the limitations of the third-generation scanners, it has issues of its own, such as the lack of ability to reject scattered radiation and a large increase in the number of detector cells. Figure 1.5 presents schematic diagrams of the four generations of scanners. Interestingly, the state-of-the-art CT scanners in the market today are all third-generation scanners.
