The evaluation of an on-site monitoring program for activity meter quality assurance with exemption-level sources

According to the International Commission on Radialogical Protection report No. 103, medical radiation should be considered in the personal radiation dose limit (International Commission on Radiological Protection 2007). Nuclear medicine applications include diagnostic and therapeutic aspects. The radiation dose of diagnostic nuclear medicine can use the concept of diagnostic reference levels (DRLs) (Vassileva et al 2015, International Commission on Radiological Protection 2017). The DRL can assist in setting up reference radionuclide activity for nuclear medicine diagnosis, ensuring image quality and radiation exposure of patients (Alessio et al 2015). For therapeutic nuclear medicine, drug delivery should consider the principle of personal radiation dose limit (International Atomic Energy Agency 2014) to avoid damage to normal tissue. It is essential to optimize treatment effect and radiation dose (Cicone et al 2021, Lassmann et al 2021, Mohan et al 2021). In nuclear medicine, when radioactive drugs are administered to patients, the biodistribution of the radioactive drugs follows the principle of physiological metabolism mechanism. The radiation dose assessment employs a medical internal radiation dose model encompassing radiopharmaceutical activity, radionuclide type, average energy, deposition time and target organ volume (Ellett et al 1964, Howell et al 1999, International Commission on Radiological Protection 2015). The absorbed dose in the body is calculated using the parameters above.

Since patient dosages can affect the determination of radiation risk (Huda 1986), it is necessary to standardize radionuclide activity measurements. In nuclear medicine institutions, radionuclide activity is measured using a radionuclide activity meter. The activity meter can display an accurate activity value of each radionuclide by adjusting the variable resistance inside the circuit (National Physical Laboratory 2006). In addition, the operational conditions of the activity meter (temperature, pressure, and shielding scatter), drug container type, liquid volume and position could affect the measurement value of radionuclide activity (Carey et al 2012, Santos et al 2009, Živanović at al 2021). Therefore, the activity meter needs quality assurance to ensure that the measured radionuclide activity is correct (Zanzonico 2008).

The related protocols for the quality assurance of activity meters have been proposed by the National Physical Laboratory, American National Standards Institute, American Association of Medical Physics and International Atomic Energy Agency (American National Standards Institute 1986, International Atomic Energy Agency 2006, National Physical Laboratory 2006, Carey et al 2012). Based on the available guidance, a quality assurance program should include radionuclide activity measurement accuracy, reproducibility, linearity and geometry efficiency. The radionuclide calibration factor of the activity meter can be ensured by conducting routine quality assurance tests (De Bessa et al 2008, Matsutomo et al 2016, Bailey et al 2018, Sanderson et al 2019). For standard sources, traceability has the importance of ensuring that all sites are measured with the same conditions. In Taiwan, the calibration factor of the activity meter was checked by the Institute of Nuclear Energy Research (INER) annual test with standard sources (Co-57, Cs-137 and Ba-133). However, the quality assurance status of activity meters was still self-monitored by each clinical institute. Therefore, INER has the potential to build up a program for monitoring the quality assurance status of activity meters.

In this study, the quality assurance of activity meters was evaluated through a questionnaire and on-site visits. First, we sent the questionnaire to nuclear medicine institutions including information on the activity meter (manufacturer, model and years of use) and the status of the quality assurance. The on-site visit was conducted by the authors of this study to evaluate the quality assurance of the activity meter. This includes radionuclide measurement accuracy, reproducibility and measurement efficiency inside the well. Considering flexibility and radiation safety, standard sources with exemption levels were used for the quality assurance tests. This paper discusses the implementation status of activity meter quality assurance in Taiwan. The measurement influence due to the use of exemption-level sources will be assessed based on the results from the on-site visit. This study intends to evaluate the feasible development of the monitoring program for the quality assurance of activity meters.

2.1. Questionnaire investigation

2.2. On-site visit

In this study, on-site visits to nuclear medicine departments were conducted, and standard sources were measured on the activity meters to assess accuracy, reproducibility and measurement efficiency when changing source position. In addition, the activity meters were visually inspected.

2.2.1. Standard sources

Three radionuclides were used in this study: Co-57, Ba-133 and Cs-137 (Eurostandard CZ, the Czech Republic). The activity of sources referred to the exemption-level criteria in Taiwan (International Atomic Energy Agency 1996). There were two types of source geometries. Co-57, Ba-133 and Cs-137 were sources for gamma spectrometry (EFS), and they were point sources with approximately 4π isotropic emission. The other source was a Ba-133 standard source for nuclear medicine (ENM), and it was the colloidal volume source in a penicillin 6 vial. When the standard sources were purchased, the radioactive sources were measured on the activity meter at INER (Capintec, Florham Park, NJ, USA). The reference activity, average energy, and measurement value from the INER of each source are shown in table 1. The uncertainties of the reference activities took into account the components other than the calculated mean values from the information of calibration reports. The uncertainties of INER were obtained from the standard deviation of measurement.

Table 1. Information from standard sources used in this study. The coverage factor was set to 1 (k = 1) for reference activity and INER measurement.

Standard source	Half-life	Average Energy a (keV)	Geometry type	Reference activity b (kBq)	Activity from INER measurement (kBq)
Co-57	271.8 d	121.6	EFS	34.0 ± 7.0	27.9 ± 1.4
Cs-137	30.1 y	661.7	EFS	8.7 ± 10.5	6.1 ± 1.9
Ba-133	10.5 y	362.6	EFS	216.3 ± 2.4	263.7 ± 2.4
			ENM	959.8 ± 6.8	990.2 ± 3.1

^a This is based on the average energy in gamma emission obtained from the Decay Data Evaluation Project (DDEP 2017). ^b The value was calibrated to the date of INER measurement.

2.2.2. Physical inspection of the activity meters

A physical inspection of the activity meter was carried out to check the regular operation of the activity meter system, including the power connection, digital display, switch function and system indicator. A background measurement check without radiation sources was performed to check for radionuclide contamination inside the instrument. The background reading was performed on the dial settings of commonly used radionuclides (Tc-99m, Ga-67, I-131, Tl-201). The background criteria should not fall outside three standard deviations from the mean background value without radiation sources determined at acceptance (International Atomic Energy Agency 2006).

2.2.3. Accuracy and reproducibility

To evaluate the measurement accuracy of the activity meter, the standard sources were measured ten times, and the average value was calculated ($\bar A$). Then, the value was compared to the decay-corrected reference activity (A_C), and the accuracy value was calculated using equation (1):

$$\begin{equation}{\text{Accuracy}} = \frac{{\bar A - {A_{\text{C}}}}}{{{A_{\text{C}}}}} \times 100\% .\end{equation} \tag{ 1 }$$

Referring to ANSI/HPS N42.13, the passing criterion of accuracy is ±10%.

To evaluate the measurement reproducibility of the activity meter, the source was removed and again placed in the activity meter for each of the ten readings (A_i ) and the average value ($\bar A$) was calculated. The measurements were decay corrected to a common reference time. The reproducibility was defined by equation (2):

$$\begin{equation}{\text{Reproducibility}} = \frac{{{A_{i}}-\bar A}}{{\bar A}} \times 100\% .\end{equation} \tag{ 2 }$$

Referring to ANSI/HPS N42.13, the passing criterion of reproducibility is ±5%.

The results from the on-site visit were later corrected with the calibration factor obtained from the annual INER test. The accuracy and reproducibility of the activity meters from the on-site visit in this study will be presented by the instrument model.

2.2.4. Measurement efficiency of the inner well position

The on-site visit conducted measurement efficiency of the inner well position to check the measurement sensitivity distribution inside the activity meter. A homemade poly-methyl methacrylate holder was used. The sources were placed on the holder's bottom disk, and the holder's top was placed at the wellhead of the activity meter. The holder's height was adjusted by additional support measuring 1 and 5 cm, respectively. The measurement efficiency of the inner well position used the Ba-133 ENM source for vertical direction (Z-axis) measurement. The source was placed at the well bottom (initial point) (figure 1(A)), then changed the vertical height of the additional support and recorded the measurement values moving from the initial point to the wellhead. Next, the holder's height was fixed at the position of the maximum measurement value in the vertical direction for measurement in the horizontal direction (X-axis) (figure 1(B)). The source was placed at the center of the disk (initial point), then the source was moved along both sides of the X-axis and the measurement values were recorded. The measurement efficiency of the inner well position will be presented in the instrument models from the on-site visit. One-way ANOVA was used to analyze whether there was a significant difference in the variance of measurement efficiency within each model. In this study, the one-way ANOVA was conducted by IBM SPSS Statistics 24 (IBM, USA). The statistical significance was hypothesized to be 0.05.

Zoom In Zoom Out Reset image size

3.1. Questionnaire investigation

3.2. On-site visit

3.2.1. Physical inspection of the activity meter

This study visited 20 nuclear medicine departments, and there were 21 activity meters comprising six models in total. The models are shown in table 2. For physical inspection, all activity meters in the visited departments functioned normally. For background checks without radiation sources, the standard deviations within all visited departments were from 0.1%–2.3%, which were within the three standard deviations of the criteria.

3.2.2. Accuracy and reproducibility

Figure 2 shows the measurement accuracy results corresponding to the four standard sources measured by the activity meters. Results are presented for six activity meter models (A)–(F). For Co-57 and Cs-137 of EFS-type sources, the accuracy of all activity meter models exceeded the 10% criteria proposed by ANSI/HPS N42.13. For the EFS-type Ba-133 source, models (B) and (D) met the 10% criteria. For the ENM-type Ba-133 source, all models met the 10% criteria.

Zoom In Zoom Out Reset image size

Figure 3 shows the results of measurement reproducibility corresponding to four standard sources within each activity meter model. For the Co-57 EFS source, the reproducibility of three models (C)–(E) met the 5% criteria proposed by ANSI/HPS N42.13. The reproducibility of the measurements of the EFS-type Ba-133 and ENM-type Ba-133 sources met the 5% criteria for all activity meter models with an average value of 0.4% and 0.2%, respectively.

Zoom In Zoom Out Reset image size

3.2.3. Measurement efficiency of the inner well position

Figure 4 shows the results of measurement efficiency of the inner well position corresponding to the Z-axis direction within each activity meter model, and they were all normalized to the position of the maximum measured value. The uncertainty bar was obtained from the mean value of five measurements at each point. The maximum measurement efficiency position in the Z-axis direction was between 6 and 7 cm of additional holder height. The results of the one-way ANOVA test showed that the variance of measurement efficiency had a significant difference (p < 0.001) between the initial point (0 cm) and wellhead (24 cm) in the single-activity meter model. However, there was no significant difference (p = 0.982) in Z-axis measurement efficiency among all models.

Zoom In Zoom Out Reset image size

Figure 5 shows the results of measurement efficiency of the inner well position corresponding to the X-axis direction within each activity meter model, and they were all normalized to the value of the center position. When the source position was moved from the center to both sides of the inner wall, the measurement deviation increased within each activity meter model, and the deviation range of all models was −0.1% to 1.2%. The results of the one-way ANOVA test showed that the variance of measurement efficiency had a significant difference (p < 0.001) between the center point (0 cm) and both sides of the inner wall (±2 cm) in the single-activity meter model. Nevertheless, there was no significant difference (p = 0.144) in X-axis measurement efficiency among all models.

Zoom In Zoom Out Reset image size

The survey results showed the implementation status of activity meter quality assurance in nuclear medicine departments. In this study, 95% of nuclear medicine departments participated in the annual calibration by INER. One activity meter did not carry out the annual calibration due to new acceptance. The statistical proportion of quality assurance, daily checks (physical inspection, background without radiation sources, and reproducibility) had at least 77% implementation. Since the results of the daily checks have a direct impact on the measurement function of the activity meter, it had the highest implementation rate. However, annual checks (accuracy, precision and linearity) were reduced to 50% and upon acceptance/after a repair check (geometry) reduced to 44%. The lack of comprehension and experience of these checks in clinical sites could be a possible reason for the results. In addition, some nuclear medicine institutions regard the INER check as the only annual quality assurance. As a result, the implementation of annual checks at the clinical end is reduced.

The measurement accuracy results of the activity meters showed that all models exceeded the ±10% ANSI/HPS specification criteria for the Co-57 and Cs-137 EFS-type sources. According to ANSI/HPS N42.13, the accuracy may fall short of the ±10% criteria if the activity in the sources is lower than 100 μCi (3.7 × 10⁶ Bq) (American National Standards Institute 1986). The initial activity of these standard sources was below the 1 × 10⁶ Bq referred to as the exemption-level limit (International Atomic Energy Agency 1996). Similarly, the measurement reproducibility results showed that some models exceeded the ±5% criteria of ANSI/HPS specification with Co-57 and Cs-137 of EFS-type sources. The utilization of exemption-level sources should consider some uncertainty problems. First, these kinds of sources may have large uncertainties in reference activity (Peralta 2004). This phenomenon also happened in the activity of INER reference (31% for Cs-137 EFS type), which made a primary contribution to the measurement uncertainty. The range of activities to be measured with an accuracy of ±5% or better will typically be between 1 MBq and 10 GBq (International Atomic Energy Agency 2014). Second, is the half-life period for decay calibration. Taking Co-57 for example, the low activity with shorter half-life (271.8 d) may cause less consistency for the long-term use of measurements. Third, the geometric choice of the standard source. The ideal source geometry should be the same as that of the source being assayed (Carey et al 2012). The ENM type (penicillin vial) was a source geometry for radiochemistry (Juget et al 2022). However, the EFS (point source) was basically used for spectrometry. From the perspective of source uniformity, the EFS type has larger decay contribution in the perpendicular direction, which leads to an angular dependence (Peralta 2004). The additional uncertainty budget of measurement results depended on the variations in background radiation and in the field instrument (Carey et al 2012). In addition, the years of use varied in the visited instruments (1–26 years), but there was no indication that the results deteriorated over the years. Since the activity meters performed annual calibration by INER, the years of use would have less affect on the measurement accuracy. Finally, the choice of exemption-level standard sources should be based on minimizing the uncertainty problem, which could lead to success of the accuracy and reproducibility tests. Although the source activities of Co-57 and Cs-137 (EFS type) were of limited use for this kind of assessment, they might offer a situation for measuring the ability of low-activity cases (e.g. pollution in the working environment). This could be seen as low-level radiation measurement and suggests the minimum detectable activity method for uncertainty estimation (Lee et al 2016). The sample geometry and the measurement time would also be considered as factors that influence the accuracy of limited activity source measurements (Done et al 2016).

From the results of measurement efficiency of the inner well position, it could be observed that there was a non-uniform sensitivity distribution in the Z-axis direction within each activity meter model. The activity meter was designed to simulate a 4π spherical geometry (Candelaria et al 2010). The electric field inside the instrument was uniform at the spherical geometry center. Therefore, this point has the highest measurement efficiency of the activity. As the source position moved towards the wellhead, the measurement efficiency gradually decreased, and the deviation increased. The guideline for the optimum region could take reference on the two regions: one was for the deviation from the maximum efficiency, which is less than 5% and another was for where it is less than 10% (Santos et al 2009). For the X-axis direction, the measured value was affected by the polarization scatter signal of the inner wall, which led to a small measurement variance (±2%) at the wall-side relative to the center position (National Physical Laboratory 2006). The position effect in the X-axis direction was insignificant compared to the Z-axis direction. However, deviation could still influence the patient radiation dose under the condition of therapeutic nuclear medicine using high-activity drugs (e.g. I-131 with 3700 MBq) (Valachis et al 2013). The efficiency test in this study could offer a detailed measurement of detection efficiency in order to check the efficiency variation so that we would be reminded of the importance of using the specialized holder provided by the manufacturer for reducing the uncertainty of radiopharmaceutical activity measurement.

The original INER annual test could only provide calibration for the standard source without including details regardingthe quality assurance status. Compared to the annual INER test, this study combined a questionnaire for quality assurance information with a simple performance test, which included accuracy, reproducibility and position efficiency. Standard sources with exemption levels were used for the on-site test, which could enhance flexibility and radiation safety for the on-site visit process. In this study, we recommended the Ba-133 ENM type with 1 MBq as the standard source. However, other standard sources used in accuracy and reproducibility tests should consider both the uncertainty of source activity and the dose rate safety according to the relative guidelines. Further development of this program can be a random on-site visit to investigate the clinical nuclear medicine department, which assists the enhancement of comprehensive monitoring of activity meter quality assurance.

This study evaluates the feasibility of developing a monitoring program for the quality assurance of radionuclide activity meters. The survey investigated the implantation status of activity meter quality assurance and helped us to understand the training insufficiency in clinical nuclear medicine departments. The application of exemption-level standard sources would have the benefit of radiation safety, but appropriate sources should consider the uncertainty including reference activity, half-life and geometry. The holder designed in this study could offer a quick check on the detection efficiency of the space dimension inside the activity meters. The establishment of this monitoring program would assist the current management on the quality assurance of activity meters in order to ensure the stability of radiopharmaceutical administration in nuclear medicine.

We thank the nuclear medicine institutions that participated in the on-site visit in this study. This study was supported by the Institute of Nuclear Energy Research [NL1070683] and the Atomic Energy Commission [AEC11012048L].

The data cannot be made publicly available upon publication because they contain commercially sensitive information. The data that support the findings of this study are available upon reasonable request from the authors.

Not applicable.