Scanning Protocol Objectives
The two main objectives when choosing scan parameters for CT QA measurements are:
- To operate the scanner as described by the manufacturer and/or a regulatory body in order to measure a given variable.
- To generate image data that matches the requirements of the Phantom and QA processing software and/or visual assessment of a given variable, in order to establish constancy.
Correctly chosen scan parameters are needed to secure consistent and accurate test results in a vendor neutral phantom. It is highly recommended to store the test settings to be used for status and constancy testing as scan protocols with descriptive names. The selection of scan parameters depends on the type of variable being measured. Most manufacturers will list recommended values for the acceptance testing.
Slice Thickness (Modules: CTP591, CTP404, CTP401, CTP682)
| Parameter | Value | Notes |
|---|---|---|
| Scan Protocol | Regulatory or vendor recommendation may be available. | |
| Scan FOV | 300>FOV>220mm | The targets of interest must be found in the image, thus the FOV must contain the target of interest, this presumes reasonable phantom centering. |
| Reconstructed FOV | 130>FOV>90mm | See comments above. For high resolution, the pixel size should be 0.25 mm or less. |
| mA | 200 mA at 120 kVp |
For measurements of the thinnest slices (1mm or less), and/or high resolution, it may be important to select the small focal spot. This often requires a mA lower than 200 mA (check the tube specs) to ensure selection of the small focal spot. |
| kVp | 120 kVp | Any calibrated kVp can be used. |
| Scanning mode | Sequential | Often the small focal spot is automatically selected for the thinnest slices. Spiral may be used but the result may depend on the pitch of the scanner and be different from sequential value. |
| Kernel | High resolution |
Other kernels may be used, but the decreased resolution may somewhat increase slice thickness. |
Sensitometry (Modules: CTP401, CTP404, CTP682)
| Parameter | Value | Notes |
|---|---|---|
| Scan Protocol | Regulatory or vendor recommendation may be available. | |
| Phantom diameter | 20cm head / 30cm or more for body | The CT number can vary considerably depending on phantom size. |
| Scan FOV | 300>FOV>200mm | Head vs. Body protocols can yield different results depending on beam hardening, scatter etc., as well as vendor approach to calibration. If annuli are added, the FOV will need to expand to ensure the complete phantom is in the scanned FOV. |
| Reconstructed FOV | 250>FOV>160mm | See notes above. |
| mAs | >200 mAs | The noise value should be kept relatively low, although the CT values are probably reasonably well determined even for lower dose scans. |
| kVp | Depends on the kVp station to be evaluated | The CT numbers do depend on kVp. See notes above. |
| Reconstruction kernel | Standard | High resolution or low contrast kernels may slightly affect CT number. |
| Scanning mode | Sequential | Presuming this is the mode of interest. Spiral mode can be used, but ensuring the spiral does not go out of the sensitometry targets. |
| Slice thickness | 1 to 10 mm | CT number can vary depending on slice thickness and beam width/scatter effect. |
| Beam width/ number of slices | 1-512 (typically 16-64) | This will depend on the vendor choices and the scan protocols of interest.
Values may vary as slice of interest moves out from central slice. |
Spatial Resolution - MTF (Modules: CTP528, CTP591, CTP682)
| Parameter | Value | Notes |
|---|---|---|
| Scan Protocol | Regulatory or vendor recommendation may be available. | |
| Scan FOV | 300>FOV>200mm | The pixel size may affect the final resolution. The resolution module needs to be within the scanned FOV and for higher (targeted) resolution, the reconstructed FOV may need to be reduced. See comments below. |
| Reconstructed FOV for standard res | 250>FOV>210mm |
The targeted reconstruction is used to avoid the image matrix Nyquist frequency limitations. The pixel size needs to be ½ or smaller dimension of the FWHM of the PSF(LSF). |
| mAs | >200 mAs | The noise should be kept reasonably low so that the pixel values are significantly higher than the noise. |
| Focal spot | Small or large focal spot | Focal spot selection often depends on the selected slice thickness, mA value, and scan protocol. The thickest available slice and the highest mA allowed for the small focal spot will often give the best results. |
| kVp | 120 kVp | Any calibrated kVp can be used. |
| Reconstruction kernel | Select the kernel to be measured | The resolution will depend on the kernel. |
| Slice thickness | Select the slice thickness to be measured | The slice thickness should be chosen so that the SNR of pixel values is sufficient. Thus, thick slices can produce reduced contrast in the bead or wire. Thin slices increase contrast for beads in particular, being careful not to saturate (over range) the CT number. This is not as important for wires. Of interest, the in-plane resolution itself is not dependent on the slice thickness. |
| Scanning mode | Sequential or spiral |
There can be slight differences between the two modes. |
Low Contrast Sensitivity (Modules: CTP515)
| Parameter | Value | Notes |
|---|---|---|
| Scan Protocol | Regulatory or vendor recommendation may be available. | |
| Phantom size/shape | Select the one of interest (eg, head vs. body phantom) | The larger the phantom, the higher the attenuation and increased noise, which will lower the LC performance. Oval shape annuli may be subject to different level calibration and uniformity results and give different LC performance. In particular, any dose modulation approaches need to be carefully monitored, particularly for oval annuli, whereas circular phantoms are not so subject to differences in dose modulation. |
| Scan FOV | 300>FOV>200mm | See previous notes on the module being contained within the FOV. |
| Reconstructed FOV | 250>FOV>150mm | See previous notes. |
| mAs | Typical 300 mAs or about 35mG /3.5Rad) | The low contrast performance will depend on the noise from the dose level to be evaluated and the CT scanner performance. Dose reducing algorithms should be approached with caution in terms of interpretation and extension of the methods used here. We cannot currently report reliable results on low contrast below 50 mAs. |
| mA | Select the mA value to be measured | The mA and scan time achieve the mAs. |
| kVp | 120kVp | The composition of the low contrast targets has been chosen to have minimal dependence on kVp from 100 to 140 kVp. However, kVp does effect penetration and noise, and therefore low contrast performance can depend on kVp when all other scan parameters are constant. |
| Reconstruction kernel | Select the kernel to be measured | The low contrast performance depends on the kernel. |
| Scanning mode | Sequential or Spiral | Although spiral may be used, it is important to ensure that the spiral range does not exceed the low contrast target dimensions. In terms of constancy, a centered sequential slice may be advised. |
| Slice thickness | Select the slice thickness of interest | The slice thickness influences the noise level. All other factors being constant, thicker slices have lower noise and thus better low contrast performance for supra-slice target (see Catphan® manual), but may diminish the sub-slice targets, where small spherical objects may be volume averaged and contrast reduced. |
| Number of repeated slices | From 1 to several slices | Because low contrast performance is statistical in nature, multiple repetitions will give an estimate of variation in statistical performance. |
Uniformity (Modules: CTP486, CTP712)
| Parameter | Value | Notes |
|---|---|---|
| Scan Protocol | Regulatory or vendor recommendation may be available. | |
| Phantom diameter and shape | 20cm head / 30cm or more for body | The CT number vary considerably depending on phantom size and shape (eg, round or oval). Various annuli are available. |
| Scanning FOV | 300>FOV>200mm | Note whether you intend to use a head or body protocol. The calibration (beam hardening, etc) may be different. The scan FOV must include the uniformity module. |
| Reconstructed FOV | 210>FOV>200mm | Should match the phantom diameter. If body annuli are used, the scanned FOV and reconstructed FOV must be large enough to include the uniformity module. |
| mAs | Chose the current and scan time of interest | Rings and bands and other nonuniformities may not be seen with high noise - low dose scans, but will possibly be seen with higher dose and lower noise. |
| kVp | Select the kVp of interest | The calibration of the system needs to be matched to the kVp used. |
| Reconstruction kernel | Select the kernel of interest | Noting if this is in head or body mode for certain scanners and reflect the scan protocol of interest. |
| Scanning mode | Sequential or spiral | Spiral may be used, but making sure there is enough water or uniform material so that the slice of interest in spiral scanning has the same material above and below the slice of interest. Sequential centered slices will minimize this risk. |
Comments
0 comments
Please sign in to leave a comment.