6. SCINTILLATION CAMERAS
6.3.3. Test of intrinsic flood-field uniformity
To test the intrinsic response of a scintillation camera to a spatially uniform flux of incident gamma radiation over the field-of-view.
Materials
Point source (see p. 147) consisting of 10-20 MBq (0.3-0.5 mCi) 99-j-cm or 113 jnm ^n solution in suitable container, giving a count rate not greater than 30 000 c/s with a 20% PHA window.
Source mounting for point source (see p. 147) Lead mask (see p. 147)
Procedure
1. Remove the collimator from the detector head. Align the head and the source mounting.
2. Position the lead mask centrally on the crystal housing.
3. Mount the source in the source mounting.
4. Centre a 20% PHA window on the photopeak (see test 6.4.2:
Check of Energy Calibration of PHA).
5. Acquire an analogue image on the display device with hard copy, at a preset count of 1.5x10'. If a digital image processor is available, also acquire a digital image. For the latter, use a 64x64 matrix with the diameter of the flood-field image fitted to 60 pixels.
The above preset count will result in a count of about 4 000 in the centre pixel.
6. Remove the source and lead mask. Replace the collimator.
Data analysis
METHOD 1: ANALOGUE IMAGE METHOD
Visually inspect the image for variations in brightness or density.
METHOD 2: DIGITAL IMAGE METHOD
1. Smooth the image data in the image processor once using a nine-point smoothing function having the following pattern of weightings :
1 2 1 2 4 2 1 2 1
2. Delineate the half-height circumference of the image by locating the pixels around the edge having a count one half of that in the centre pixel. This may require interpolation between pixels adjacent to the half-height position. Then define the useful field-of-view, UFOV, on the digital image as that within the circle with a radius which is 95% of the mean half-height radius. Similarly define the central field-of-view, CFOV, as that within the circle with a radius which is 75%
of the mean half-height radius (Fig. 6-14).
3. Determine the maximum (Max) and minimum (Min) counts in the pixels lying within the UFOV and the CFOV. The integral uniformity, IU, is then given by;
= ioo/Max~Min
Max + Min
4. Determine for each row or column of pixels in the X and Y directions within the UFOV and the CFOV, the maximum count difference in any 6 contiguous pixels. Determine the highest value of this maximum count difference in the sets of rows and columns. The differential uniformity, DU, is then given by:
DU=100/Hi~Low
Hi + Low
where Hi and Low are the pixel counts giving the highest value of the maximum count difference.
Observations
This test is intended to be performed as an acceptance and reference test, and at quarterly intervals.
In both the analogue image and the digital image method, if the scintillation camera is fitted with a uniformity correction circuit, the test should, if possible, be performed with and without the circuit enabled.
The digital image method can be followed with appropriate software to perform the analysis automatically or, more laboriously, with a print-out of the count in each pixel of the 64x64 matrix.
4000
2000
UfOV
______CFOV_______
75% OF HALF-HEIGHT RADIUS
HALF-HEIGHT RADIUS
DISTANCE FROM CENTRE (PIXELS) 30
Fig. 6-14. Test 6.3.3: Test of Intrinsic Flood-field Uniformity.
Definitions of useful field-of-view (UFOV) and central field-of-view (CFOV) from digital count profile of intrinsic flood-field image.
Interpretation of results
METHOD 1: ANALOGUE IMAGE METHOD
At acceptance testing, the images should be compared with those acquired by the manufacturer at the factory or by his representative at installation.
At routine testing, the images should be compared with the reference images.
METHOD 2: DIGITAL IMAGE METHOD
At acceptance testing, the values of integral and differential uniformity for the useful and central fields-of-view should be compared with the manufacturer's worst-case values.
At routine testing, the values should be compared with the reference values.
The uniformity of most scintillation cameras with a uniformity correction circuit but with the circuit disabled will be poorer than with the circuit enabled. This does not represent a malfunction. Uncorrected reference images should be obtained at the time of acceptance or after major repair and further images obtained weekly thereafter to monitor for defects which may be hidden in the corrected images (see Observations, test 6.4.2: Check of Flood-field Uniformity). If the defects appear to worsen, maintenance should be scheduled, as eventually the correction device will be unable to produce a uniform response. Further, if count addition or subtraction processes are employed in the uniformity circuit, the number of added or subtracted counts will become an increasing significant fraction of the total as the uncorrected uniformity worsens, and any quantifications based on the corrected images, such as ejection fraction calculations, will contain increasingly significant errors.
Limits of acceptability
METHOD 1: ANALOGUE IMAGE METHOD
There are no absolute limits of acceptability for this method. At acceptance testing, if the image obtained from the display device appears to differ from that obtained at the factory or at installation, corrective action should be initiated through the manufacturer's representative.
At routine testing, the image should be comparable to the reference image. Its shape should be round or hexagonal, as appropriate, and its uniformity adequate for clinical imaging. Evident non-uniformities would call for follow-up action.
METHOD 2: DIGITAL IMAGE METHOD
At acceptance testing, a value of integral or differential uniformity that is 10% or more above the manufacturer's worst-case value would call for corrective action initiated through the manufacturer's representative.
At routine testing, a value 20% or more above the reference value would call for follow-up action.
Conclusion
Record whether or not the results confirm acceptable performance.
If not, indicate follow-up action taken.
6.3.4: TEST OF INTRINSIC FLOOD-FIELD UNIFORMITY OVER