Test of tomographic resolution in air

4.3.4. Test of tomographic resolution in air

Purpose of test To measure the tomographic resolution of the system in air and to ensure

that the reconstruction process is not degraded by either the tomographic that the reconstruction process is not degraded by either the tomographic

Materials

A small point source of 99m Tc, as used in the test for centre of rotation and alignment (see Section 4.3.6).

Procedure (1) Place the point source in air within 1 cm of the centre of rotation, near the

centre of the field of view. (2) Set the radius of rotation to be approximately 15 cm, or if this cannot be

achieved, to be as small as possible. Use a circular orbit of rotation. (3) Perform a tomographic acquisition using the matrix size and number of angles used clinically, collecting about 10 000 counts per view. (4) Reconstruct the data with filtered backprojection, using either a ramp filter or the sharpest filter that the system will permit. (5) Perform a normal planar (static) acquisition at the home position, using the same acquisition matrix size, etc., as for the tomographic acquisition. (6) Repeat steps (1)–(5) with the point source placed about 8 cm off axis. (7) Repeat steps (1)–(5) with the point source placed on the axis of rotation,

but close to the edge of the field of view (close to +Y MAX and –Y MAX ), as indicated for the centre of rotation test (see Section 4.3.6, step (5)).

Data analysis (1) Draw a profile through the image of the point source in the reconstructed

image and calculate the FWHM in both the horizontal and the vertical directions, estimating the FWHM as described in Section 2.3.8.

(2) Measure the FWHM in the horizontal direction on the planar image acquired at the home position.

Observations This test is intended to be performed as an acceptance and reference test

and at half-yearly intervals. There should be no significant difference between the FWHM calculated from the horizontal and vertical profiles in the reconstruction, when a circular orbit is used, in air. There may be differences for non-circular orbits, in scatter, and when the source is offset from the axis of rotation.

The interpretation of the results does not change when a larger radius of rotation than recommended has been used, but the absolute values of the FWHM will increase for both tomographic and planar images.

Interpretation of results This is a useful system test to ensure that the centre of rotation of the

system has been accurately calibrated, that the acquisition and reconstruction software is functioning correctly and that adequate performance can be obtained. Any error in the centre of rotation, any errors due to vibration, etc., will result in a loss of tomographic resolution with respect to planar resolution. If a filter other than a ramp filter has been used in reconstruction, there will be some degradation of tomographic resolution by comparison to planar resolution.

Set the display so that there is a very low maximum cut-off level, for example, by observing the background. The reconstructed image of the point source should appear to be round and, in particular, not distorted into shapes such as a comma, ellipse or ring. There are likely to be streaks radiating symmetrically from the point source but there should not be so-called ‘preferred directions’ where the streaks appear to be considerably more significant than others. There may be an ‘edge’ surrounding the image but this should not be much more significant than the amplitude of the streaks.

There should be no difference (except for the field of view) between the results obtained, as described above, with the point source in the central slice and the results obtained with the point source close to the edge of the field of view.

Limits of acceptability Depending on the collimator, reconstruction filter and the radius of

rotation, various values for the FWHM may be obtained. If the FWHM measured on the planar view is 12 mm (a typical value for a radius of rotation of 15 cm and a high resolution parallel hole collimator), then the FWHM on the tomogram should not be worse than 13.2 mm (i.e. 12 mm +10%) when a ramp filter has been used. The difference between planar and tomographic resolution should be no more than 2 mm or 10% (of the planar resolution), whichever is less. If it is worse, then the system has probably not been well centred. If this is the case, a new centre of rotation calibration should be performed. If, after recalibration, the FWHM is still poor, then this may be the result of vibration, or other such error, and the system needs adjusting. If a rotation, various values for the FWHM may be obtained. If the FWHM measured on the planar view is 12 mm (a typical value for a radius of rotation of 15 cm and a high resolution parallel hole collimator), then the FWHM on the tomogram should not be worse than 13.2 mm (i.e. 12 mm +10%) when a ramp filter has been used. The difference between planar and tomographic resolution should be no more than 2 mm or 10% (of the planar resolution), whichever is less. If it is worse, then the system has probably not been well centred. If this is the case, a new centre of rotation calibration should be performed. If, after recalibration, the FWHM is still poor, then this may be the result of vibration, or other such error, and the system needs adjusting. If a

Conclusion Record whether or not the results confirm acceptable performance. If

not, indicate the follow-up action taken.