Optimising the number of thermoluminescent dosimeters required for the measurement of effective dose for computed tomography attenuation correction data in SPECT/CT myocardial perfusion imaging

Background: The use of thermoluminescent dosimeters (TLDs) is regarded as time consuming and laborious. As part of our dosimetry research it was necessary to optimise the use of our resources, both physical and time. Experimental work was carried out to develop a method that allowed for a reduction in the number of TLDs needed for accurate effective dose measurement. For this work specific reference to computed tomography attenuation correction (CTAC) for myocardial perfusion imaging (MPI) acquisitions is made although it is proposed that the developed method could be applied to dose assessments using TLDs. Research to measure and compare the effective dose from CTAC for MPI was to be carried out using an ATOM 701 dosimetry phantom, Harshaw 3500 manual TLD reader and TLD-100s. Method: To establish the areas of the phantom where dose measurements should be carried out, a batch calibrated TLD-100 dosimeters were placed along the centre of the phantom. A simulated CTAC for MPI was performed. After reading the distribution of the dose was recorded and areas where dose levels were below the sensitivity threshold dose of 50 μGy were noted. To test the effect of excluding dose measurement for some areas on the final calculation of effective dose and the time taken to acquire the data a repeat acquisition was performed with the full complement of TLDs placed in the phantom in organ locations recommended by the manufacturer. The time taken for loading, unloading and reading was recorded. Effective dose and organ doses were calculated. The calculation was repeated with TLDs outside the established range excluded and the potential time saved calculated. Results: Excluding TLDs from areas where doses were below the 50 μGy threshold resulted in 82 fewer TLDs being used (268–186) leading to a time saving of around 2 h per batch. The results of the experiment showed that effective dose measurements were 1.75% lower with the reduced chipset and organ dose measurements were not significantly different (p > 0.10). Conclusion: It is proposed that this methodology could be applied to TLD dosimetry work to establish the areas that should be included in the measurements. In some cases significant savings in time could be made.