Introduction: One of the most used imaging systems in nuclear medicine is gamma camera. It is used to detect single gamma photons emitted from radioactive substance. The main component parts of gamma cameras are collimator, crystal (NaI), photomultiplier tubes and processing electronics. If we want gamma camera to work correctly, we have to regularly perform quality control tests. One of the most important tests, that we have to perform daily before we start clinical work, is called uniformity. The uniformity can be measured intrinsically (without collimator and with technetium source) or extrinsically (with collimator and cobalt flood source). In both cases we measure it central, global, integral and differential. Purpose: We wanted to determine how uniformity varies with extrinsic test if we change specific parameters (number of counts and the size of matrix). We also wanted to determine if there is any difference in uniformity between automatic and manual test. Methods: The measurements were carried out on SPECT/CT (Siemens Symbia series T2) in Clinic of nuclear medicine, where we placed cobalt flood source directly on detector 2 and then moved the detectors closely together. Then we used the protocol for manual uniformity where we changed the number of counts (3 millions, 4.5 millions and 10 millions) and the size of matrix (1024 x 1024, 512 x 512, 256 x 256 and 128 x 128). The time duration of the tests was between 10 and 40 minutes. We have collected 10 measurements for each variable separately. Uniformity was calculated with the IAEA application in the program of ImageJ. Data that we collected with manual test where we used 10 million counts and the matrix size 1024 x 1024 was also used for comparison with automatic test, where it was used linear source (Gd-153) and the tomographic way of imaging. Results: The measurements were statistically tested with SPSS. One of the tests that we used was ANOVA, where we decided for the statistical significance to be at 0,01. ANOVA showed us that uniformity values are lower, if we use bigger number of counts (p < 0,001). In most cases was determined that the size of matrix does not affect uniformity (p > 0,01). Exception was shown at integral uniformity with useful field of view on detector 1 (p = 0,001) and detector 2 (p < 0,001). When we compared automatic test with manual we used Mann-Whitney test with statistical significance 0,05. We determined that significant difference does exist between the two tests at integral uniformity. Discussion and conclusion: With our research we have came to conclusion that both the differential and integral uniformities for useful and central field of view are improving as the number of acquired counts increase, because there are fewer statistical fluctuations. The size of the matrix does not affect uniformity. However there have been some deviations in relation to coincidentally events which happen in nuclear medicine, because we deal with radioactive sources. This sources are constantly decaying so we never get the same results. That is why our measurements have to be inside the written tolerance.
|
