To achieve 4D CT imaging of an object in the presence of respiratory motion, one must acquire data at each location for the duration of at least one breath cycle plus the duration needed to acquire sufficient projections for one image reconstruction. This is the data sufficiency condition (DSC) for 4D CT imaging [16]. The time needed to acquire one image is equal to one gantry rotation cycle for full-scan reconstruction (FSR) or of a gantry rotation cycle plus fan-angle for half-scan reconstruction (HSR) [18], which is plus about , amounting to roughly gantry rotations. Additional data acquisition is needed because it takes at least of a gantry rotation cycle in CT for reconstruction of an image and to ensure that there are images at both ends of a complete breath cycle. This additional data acquisition is not necessary for projection x-ray imaging such as fluoroscopy in which each projection is an image by itself [13]. Two acquisition modes, helical and cine, can be used to realize the 4D CT imaging. A helical CT scan acquires data when the table translates at a constant speed programmed by a pitch factor , which is defined as the table travel per rotation divided by the width of the x-ray beam projected onto the rotation axis (Fig. 2.1).

To satisfy the DSC with FSR, the pitch values
and with HSR
where and are the durations of the gantry rotation cycle and the breath cycle, respectively. Let us examine the formula for FSR. If the breath cycle is the same as the , then after one (e.g.,  s), the detector moves in and out of the x-ray beam in exactly one . Taking into account of the extra acquisition of for one image reconstruction,  0.5. If the breath cycle is 4 times of (i.e.,  s,  s), then . Typically  s and  s and , almost the same gantry rotation cycle as the typical pitch factor of 0.1 or less in the helical 4D CT in Tables 2.1 and 2.2. Similar reasoning can be applied to Eq. (2.2) for HSR.

Although HSR is used in cardiac CT for better temporal resolution, FSR is typically used in 4D CT because FSR gives rise to fewer image artifacts between 25 and 50 cm reconstruction field of view (FOV) than HSR (Fig. 2.2). The FOV with HSR is typically less than 25 cm only to cover the heart. The FOV for radiation therapy simulation is at least 50 cm. Since FSR uses more data in the image reconstruction than HSR, its image will also be less noisy. One very important observation of the pitch value selection is that the longer the breath cycle or the shorter the gantry rotation cycle is, the smaller the pitch factor becomes. In diagnostic CT imaging with patient breath-hold and without gating, is about 1, which is 10 times faster than in a helical 4D CT scan with respiratory gating!

A MSCT can be characterized by the number of data channels that can be simultaneously read-out (SRO) in the patient table direction. This number of SRO is also frequently referred to as the number of slices in MSCT. A MSCT with a higher number of SRO typically means a newer MSCT that can scan the same coverage faster. To acquire the cine 4D CT data with GE scanners, the numbers of SRO data channels are 4, 8, 16 and 64 (Fig. 2.3a). To acquire the helical 4D CT data with Siemens or Philips scanners, the numbers of SRO data channels are 16, 20, 40, and 64 (Fig. 2.3c). The reason that helical 4D CT started at 16-slice was because Siemens/Philips only introduced the low pitch helical CT scan () in their 16-slice and up CT scanners. This practice is typical in commercial CT that a new feature is only introduced in newer CT scanners, which was 16-slice at the time.

The number of physical data channels on a MSCT is at least the number of data channels of SRO. Both GE’s 4 and 8 channels have 16 physical detectors. For GE’s 16 and 64 channels, there are 24 and 64, detectors, respectively. During data acquisition, the cine 4D CT utilizes the data acquisition modes of , , and  mm on GE’s 4, 8, 16 and 64-slice CTs, respectively (Fig. 2.3b) to produce the images of slice thickness 2.5 mm for treatment planning. Combination of smaller data channels to 2.5 mm slice is conducted in data acquisition. A 16-slice CT may not have a significant advantage over an 8-slice CT in terms of acquisition speed for the cine 4D CT because the detector size is at the maximum of 2 cm for both the 8- and 16-slice GE CT.

In helical 4D CT data acquisition, the raw data before helical data interpolation are smaller than 1.5 mm (Fig. 2.3d). For the 16-slice, the mode of data acquisition is  mm. For the Siemens 20/40-slice, it is  mm. For the Philips 40/64-slice, it is  mm. Siemens 16, 20 and 40-slice CTs have the same detector coverage of 2.4 cm. Therefore, the speed of helical 4D CT is the same for the Siemens 16, 20 and 40-slice CTs. The Siemens 40 and 64-slice scanners use the z flying-focal spot to fast-switch the focal spot between two positions to double-sample each angle of projection data to achieve the effect of 40 and 64-slice data even though their detector channels are 20 and 32, respectively [6]. The detector size of 16, 20 and 40-slice Siemens CTs is 2.4 cm, and it is 2.88 cm for the 64-slice Siemens CT. The Philips 40 and 64-slice CTs have 4 cm detector coverage, and their data acquisition is at  mm for the helical 4D CT, which will shorten the acquisition time for their large coverage. It is important to note that slice broadening of 180 % in the low-pitch helical 4D CT prevents the smallest data element in data acquisition to be greater than 1.5 mm.

Helical 4D CT scan data allow for CT image reconstruction at any location by permitting data interpolation between two neighboring detector elements (Fig. 2.4), whereas cine 4D CT scan data allow for reconstructions only at the scan position. When using 4D CT to scan patients with lung cancer, it is important to have a complete coverage of the lungs for tumor delineation and dose calculation. The flexibility of image reconstruction at any location is not important as long as the images of uniformly spaced slices can be reconstructed to create a 4D volume. Typical image slice thicknesses in radiation treatment planning are 2–3 mm. There is a penalty from data interpolation in helical 4D CT as it widens the slice sensitivity profile, a measure of the slice thickness of the CT image. It is important to use thin slice collimation of  mm in helical 4D CT on a 16-slice MSCT scanner because data interpolation will widen the slice thickness to almost 2.7 mm (that is, the 1.5 mm slice will be thickened to 1.8 times its original width). Similarly, this means that if the detector configuration is  mm, the slice thickness will become 4.5 mm in helical 4D CT. Same  mm collimation in cine 4D CT will generate the slices of 2.5 mm because no data interpolation is in the cine CT image reconstruction. In diagnostic imaging, data interpolation causes only about 20 % widening when the pitch is greater than 0.5; the amount of widening reaches 180 % when the pitch is less than 0.2 (Fig. 2.5). To better understand this, we can take two independent slices of 2.5 mm at two neighboring locations in cine 4D CT and interpolate for an image located equidistant to the two slices; in such a case, interpolation would assign 50 % weight to each image, and the composite image will become a 5.0 mm slice (resulting in 200 % broadening of the slice thickness). Image reconstruction at pitch values of less than 0.1 in helical 4D CT significantly thickens the CT images as shown in Fig. 2.4. Once the pitch factor becomes 0, helical CT interpolation is no longer needed and the helical 4D CT becomes cine 4D CT. That is the reason that the typical detector configuration in a 16-slice MSCT is  mm to keep the slice thickness  mm.

Stay updated, free articles. Join our Telegram channel

Join