Personal Exposure to Static and Time-Varying Magnetic Fields During MRI System Test Procedures

Transcription

1 JOURNAL OF MAGNETIC RESONANCE IMAGING 30: (2009) Technical Note Personal Exposure to Static and Time-Varying Magnetic Fields During MRI System Test Procedures Frank de Vocht, PhD, 1 * Floris Muller, MS, 2 Hans Engels, PhD, 3 and Hans Kromhout, PhD 2 Purpose: To assess personal time-weighted average (TWA) static magnetic field and time-varying magnetic field exposure for several system testing tasks routinely conducted by engineers at an MRI manufacturing plant. Materials and Methods: Personal exposure was measured using a personal dosimeter that measured B and db/dt in all three orthogonal directions with a 1 Hz frequency. TWA exposure was calculated and random effects and linear mixed effects models were used to assess exposure levels and variability. Results: Whereas full-body peak (2T) and TWA (200 mt/ 8 h) exposure limits for the static magnetic field originally proposed by ICNIRP were not exceeded, time-varying magnetic field peaks (db/dt) did exceed the proposed threshold value (767.9 mt/s) suggested by IEEE on several occasions at the 3.0 Tesla (T) whole-body cylindrical system. Conclusion: TWA exposure levels are well below the proposed limit values, but peak exposure limits are exceeded during certain procedures performed on 1.0T to 3.0T systems. Key Words: exposure assessment; system tests; exposure limits; static fields; time-varying fields J. Magn. Reson. Imaging 2009;30: VC 2009 Wiley-Liss, Inc. IN 2004, THE EUROPEAN UNION (EU) drafted a directive for workers exposure to electromagnetic fields (1), which were aimed at guaranteeing the health and safety of people who regularly get exposed to these fields at work. This directive was meant to be implemented in EU national regulations before April 30, 2008, but after criticism from international 1 Occupational and Environmental Health Research Group, School of Translational Medicine, Faculty of Medical and Human Sciences, University of Manchester, Manchester, United Kingdom. 2 Environmental Epidemiology Division, Institute for Risk Assessment Sciences, Utrecht University, Utrecht, The Netherlands. 3 Philips Medical Systems, Best, The Netherlands. *Address reprint requests to: F.d.V., Occupational and Environmental Health Research Group, School of Translational Medicine, Faculty of Medical and Human Sciences, The University of Manchester, Ellen Wilkinson Building, Oxford Road, Manchester M13 9PL, UK. frank.devocht@manchester.ac.uk Received May 15, 2009; Accepted August 13, DOI /jmri Published online in Wiley InterScience ( experts (2 4), the decision was made to postpone the implementation of the directive to 2012 (see EU Directive 2008/46/EC) so that the proposed exposure limit values could be re-evaluated taking into account results from expected new studies and evolving knowledge. The proposed guidelines for exposure to static magnetic fields translate into a 200 millitesla (mt) over 8-h time-weighted average (TWA) and a 2 Tesla (T) peak exposure threshold values for torso exposure, and a 5T peak exposure threshold value for limbs (5). Furthermore, for the bandwidth of 0 7 Hertz (Hz) a peak allowable limit of mt per second was suggested for exposure to time-varying magnetic fields (6,7). The 2T exposure threshold value for torso is similar to the exposure limit value for static magnetic field exposure more recently proposed by the International Commission on Non-Ionizing Radiation Protection (ICNIRP) (8). However, the TWA value of 200 mt over 8 h is no longer maintained in the ICNIRP 2009 Guidelines (8) compared with the 1994 Guidelines (5) because mechanistic considerations indicate that any effects are likely to be acute, and ICNIRP guidelines justify a much higher peak exposure limit of 8T for specific work applications if the environment is controlled and appropriate work practices are implemented to control movement induced effects. Whether or not these new limit values will be introduced in the postponed directive is not known at present. Regardless, valid measurements of personal exposure are required to (a) assess compliance with whatever threshold values will be introduced and (b) provide estimates of exposure level and variability for use in epidemiological studies to assess potential health and neuropsychological effects from magnetic field exposure. Until recently, personal exposure measurements were not possible and exposure was estimated from movement patterns of workers through the magnetic stray field (9) or based on simulations (10 14), because available measurement devices to measure the static magnetic field were too large and heavy to be worn during work without interfering with normal protocol. Relatively small and lightweight personal dosimeters have now been developed that can be worn for a working day without interfering with work VC 2009 Wiley-Liss, Inc. 1223

2 1224 de Vocht et al. practice (6;13 17), and thus enable assessment of personal exposure levels, and sources of variability that cause individual workers to be differently exposed even when performing tasks at similar systems. The first few exploratory quantitative exposure measurements during actual work procedures have been collected in the health care industry (6), but outside MRI rooms in hospitals, no personal exposure data are currently available. As such, there is a need for exposure data, and this study aimed at providing the first exploratory quantitative estimates of levels and variability of magnetic field exposure outside the health care industry. with the work shift of a subject and lasted until a specific task was completed (sampling time varied from 69 to 462 min) with exception of the lunch break because engineers would leave the building and exposure would not be relevant to the specific tasks anymore, in which case the dosimeter was switched off. The 20 Hz measurements were aggregated to 1 s averages and were used to calculate time weighted average (TWA) and peak static and time-varying magnetic field exposure for each measurement session. Details of the measurements are described in Table 1. Data were analyzed using linear mixed effects models to account for repeated measurements on systems, tasks, and workers, and were conducted using R statistical software (version 2.8.1). MATERIALS AND METHODS Personal exposure measurements were collected using a personal dosimeter (Mr.Dose TM (Wave Instruments Pty Ltd)). This 200-gram device is worn around the waist at a subject s left side and three Hall meters measure the magnetic field strength (B) as well as the change of the magnetic field over time (db/dt) in all three orthogonal dimensions (x,y,z) with a frequency of 20 Hz. As such, the total exposure at time t can be calculated for static field exposure as: B t ¼ p ðb 2 xt þ B2 yt þ B2 zt Þ (1) And for time-varying magnetic field exposure as: db=dt ¼ p ððdb x =dtþ 2 þðdb y =dtþ 2 þðdb z =dtþ 2 Þ (2) One dosimeter was available and was calibrated by the manufacturer. Calibration involved low magnetic field strength calibration for the linear region of the Hall sensor (by means of a 0.5T permanent magnet and portable Gauss meter using a zero flux chamber) and high magnetic field strength calibration to correct errors between the rate of change integration and the Hall sensor and to ensure linearity at higher field strength after calibration (6). Personal magnetic field exposure was measured during those tasks that are routinely conducted by MRI system engineers to test and optimize new MRI systems in an MRI manufacturing plant and were expected to yield above-average levels of exposure because the system engineers are required to enter the magnetic stray field and MRI magnet regularly. These tasks were monitored at 1.5T and a 3.0T cylindrical whole body MRI scanners and a 1.0T whole body open system and included shimming of the magnet homogeneity, turning the magnet field off, System Verification Tests (SVT), tag removal, tests of the Quadrature Body Coil (QBC), and tests of the Data Acquisition Cabinet (DAC). After explaining the procedures and obtaining informed consent, a total of 22 measurements were collected among 10 randomly selected engineers, although selection of the workers at the MRI manufacturing plant was depending on whether any of the above tasks would be conducted during a particular day (1 to 4 measurements per subject). Measurements started simultaneously RESULTS The pattern of exposure is shown graphically for static field exposure (Fig. 1) and time-varying magnetic field exposure (Fig. 2). As shown, depending on the task, the exposure pattern for static as well as time-varying magnetic field exposure can be characterized by recurrent peaks followed by periods of very low exposure. Summary statistics for each measurement are shown in Table 1. TWA exposure for different tasks at the 3.0T system ranged from 1.23 mt to mt, at the 1.5T system from 1.48 mt to mt, and at the open 1T system from 1.61 mt to mt. Similarly, peak exposure ranged from 187 to 3,975 mt/s, 277 to 944 mt/s, and 54 to 504 mt/s on the 3T closed, 1.5T closed, and 1T open systems, respectively. Whereas static magnetic field exposure peaks were lowest when taking the 1.0T system offline (54 mt), SMF peaks over 1.0 T were found during QBC tests and SVT at the 3.0T system. The mt/s guideline for time-varying magnetic field exposure (db/dt) was exceeded on routine basis in the tasks monitored during this study. Exceedance frequency ranged from never, when taking the system off-field, to almost three times per minute during 3.0T System Verification Tests. Although very variable, the tasks with the highest peak frequencies were shimming (range up to 1.45 peaks per minute) and System Verification Tests (range, 0 to 2.60 peaks/min). The peak frequency depended not only on the task, but also increased with the strength of the magnet. Furthermore, the magnitude of these peaks during a task was extremely variable and ranged from 105 mt/s, when taking the 1T system off field, to almost 4 T/s during shimming procedures at a 1.5T system. Table 2 shows the results of the linear mixed effects models. By including system and task as fixed factors in linear mixed effects models, 47% of between-worker and 33% of within-worker variability of individual mean TWA static magnetic field exposure could be explained. However, by adding system and tasks fixed effects to the TWA db/dt model, almost all of the between-worker variance was explained, but only 25% of the temporal variability in individual mean TWA db/dt exposure.

3 Static MF Exposure During System Tests 1225 Table 1 Summary Statistics of Measurements Task a System strength (T) TWA b B (mt) Peak c B (mt) TWA db/dt (mt/s) Peak b db/dt (mt/s) Duration (min) No. of peaks above 767 mt/s per minute Subject DAC DAC DAC Off-field QBC QBC QBC Shimming Shimming Shimming Shimming Shimming SVT SVT SVT SVT SVT SVT SVT SVT SVT Tags a Tasks: tests of the Data Acquisition Cabinet (DAC), turning the magnet field off (off-field), tests of the Quadrature Body Coil (QBC), shimming of the magnet homogeneity (shimming, System Verification Tests (SVT), and tag removal (tags). Units: Tesla(T), millitesla(mt), mill- Tesla per second (mt/s), minutes (min), static magnetic field (B), and time-varying static magnetic field (db/dt). b Time weighted average (TWA) based on arithmetic mean of 1 second averages. c At 20 Hz resolution. Average TWA static field exposure was as expected significantly higher during procedures at the 3 T system than at the 1.0 T system. Furthermore, statistically significant (P < 0.05) higher TWA exposure was found during shimming procedures (17 mt at the 1T open system, 25 mt at the 1.5T system, and 86 mt at the 3T system, respectively), but higher levels were also found (P < 0.10) for SVT (range, 4 mt to 21 mt) and QBC tests (range, 3 mt 17 mt), compared with DAC tests, tag removal, and taking the system off field. TWA db/dt exposure increased significantly when the system where the tasks was performed had a higher field strength. Similar to average SMF exposure levels, highest average exposure to time-varying magnetic fields was found for shimming (11 mt/s, 16 mt/s, and 22 mt/ s, at 1T, 1.5T, and the 3T system, respectively). Higher db/dt exposure was also found during SVT at all systems, with the TWA exposure levels ranging from 7 mt/s at the 1T system to 14 mt/s at the 3T system. Low average exposure was found during DAC tests; ranging from 4 mt/s to 7 mt/s. DISCUSSION These data provide estimates of personal exposure to magnetic fields in the MRI manufacturing industry where high exposure to magnetic fields are to be expected. Although these analyses were based on relatively few independent measurements of exposure to static and time-varying magnetic fields, they do suggest that, whereas the limits for static magnetic field exposure, both full-body peaks (2T) and TWA (200 mt/ 8hr) (5), were not exceeded, time-varying magnetic field peaks (db/dt) did exceed the proposed threshold value of mt/s (6,7) regularly (note that no limit values have been proposed for TWA db/dt); most notably during shimming and during SVT and QBC test at the 3.0T system. As such, if these proposed guidelines were to be used as exposure limit values the regular exceedance of the time-varying magnetic field peak exposure limits would suggest that with the current task-design most of these system testing procedures at 1.5T and 3.0T systems would not be allowed anymore, with the probable exception of DAC tests (although even here the proposed threshold value of mt/s was exceeded twice). However, additional measurement should be conducted to better estimate these exposure levels and peak frequencies during different tasks. Nonetheless, the measured TWA exposure levels for static field exposure at the 1.0T and 1.5T systems were comparable to those estimated from worker movement in the stray fields in systems with similar field strength in a previous study among system engineers from the same manufacturing plant (9). Conclusions on exceedance of proposed limit values during tasks in the vicinity of 1T to 3T whole body systems

4 1226 de Vocht et al. Figure 1. One-second geometric mean personal static magnetic field exposure (B 0 ) in millitesla for different tasks (DAC test, QBC test, shimming, SVT, tagging, off field, and no specific task) on different MRI systems (3.0T and 1.5T cylindrical and 1.0T open systems). therefore these testing procedures as they are conducted at present would only be permitted if the MR system can be considered to be a controlled environment (8) and these limits would be incorporated in new regulations. These data do suggest that to assess workers exposure levels in relation to TWA exposure limits, it will be important to take additional personal exposure measurements, instead of relying on mathematical calculations. Only 47% of the variability of individual mean TWA static field exposure between different workers could be explained by the system field strength and task descriptions alone. For TWA db/dt exposure however, system and task descriptions did explain almost all of the variability of individual mean exposure between workers. The estimate and contribution of the between-worker variance in this study should be interpreted with care and most likely underestimated the true contribution because only one engineer worker at two different systems and only two engineers conducted the same task at the same system. As such, additional measurement data are needed to properly estimate the components of exposure variability. The exposure patterns described here suggest that time-weighted averages are most likely not the most appropriate exposure metrics to be used for health- were similar to those described for standardized movements in which positioning of a test phantom inside the scanners was simulated (17). These exposure levels among the system testers were associated with increased reporting of vertigo, metallic taste, and concentration problems when compared with nonexposed colleagues (9). It should be noted that, because the dosimeter was worn at the waist by the engineers, the measurements described in these analyses most likely underestimated true exposure levels at the head, the most likely target organ for reported neurobehavioral effects (18 20). This likely underestimation of exposure can be ascribed to the fact that by placing the dosimeter at the waist, frequent additional movements of the torso and head, such as leaning into the bore of a magnet head first with the waist not reaching into the magnet bore, is not accounted for. These movements were conducted in areas with highest static magnetic field exposure in the magnet bore and highest spatial gradient at the edge of the bore. That these exposures were not registered is nicely illustrated by the fact that the dosimeter never measured the actual intensity of the homogeneous magnetic field inside the magnetic bore (1.0T, 1.5T, or 3.0T) in any of the tasks, even though engineers have entered the magnet bores at several occasions. Note that, for the 3.0T system, this would in practice suggest that the proposed exposure limit of 2T would be exceeded and that Figure 2. One-second geometric mean personal time-varying magnetic field exposure (db/dt) in millitesla/second for different tasks (DAC test, QBC test, shimming, SVT, tagging, off field, and no specific task) on different MRI systems (3.0T and 1.5T cylindrical and 1.0T open systems).

5 Static MF Exposure During System Tests 1227 Table 2 Time-Weighted Average (TWA) Personal Static Field (B 0 ) and Time-Varying Magnetic Field (db/dt) Exposure and Within and Between Worker Variance Components Estimated From Linear Mixed Effects Model TWA a B 0 TWA a db/dt System Task b mt SEE c System mt/s SEE c 3.0 Tesla* DAC T** QBC* Shimming** SVT** Tag removal Off field Tesla DAC T* QBC* Shimming** SVT** Tag removal Off field Tesla open DAC T open QBC* Shimming** SVT** Tag removal Off field Variance components Full model Random effects % explained Full model Random effects % explained only only Between worker % % Within worker % % a Time weighted average (TWA) in millitesla (mt) or millitesla per second (mt/s) based on arithmetic mean of 1 second averages. b Tasks: tests of the Data Acquisition Cabinet (DAC), turning the magnet field off (off-field), tests of the Quadrature Body Coil (QBC), shimming of the magnet homogeneity (shimming, System Verification Tests (SVT), and tag removal (tags). c Standard error of estimate (SEE). *Statistically significant different (P < 0.10); Statistically significant different (P < 0.05). Italic ¼ model estimates not covered by actual measurements. based exposure limits. Although no data are currently present on long-term health effects of static magnetic field exposure, data do exist on acute effects, which suggest these are associated to db/dt peaks in combination with the strength of the static magnetic field (18 20). In conclusion, although based on relatively few independent TWA measurements, these data provide the first estimates for time-weighted average and peak static magnetic field and time-varying magnetic field exposure during system testing procedures, and suggest that whereas TWA exposure levels are well below the limit values proposed in ICNIRP guidelines, the EU proposed limits values for peak exposures are being exceeded during certain procedures performed at 1.0T to 3.0T systems. However, to improve these estimates we propose to use the estimates from the mixed effects models presented here as prior information for updating using new measurements in a Bayesian analysis framework (19). ACKNOWLEDGMENTS The authors thank Professor Crozier and Dr. Fuentes for providing the dosimeters and for assistance during data management and analysis. Furthermore, we also thank all volunteers who participated in this study as well as the plant s management. REFERENCES 1. European Union. Directive 2004/40/EC of the European Parliament and of the Council of 29 April 2004 on the minimum health and safety requirements regarding the exposure of workers to the risks arising from physical agents (electromagnetic fields) (18th individual Directive within the meaning of Article 16(1) of Directive 89/391/EEC) Hill DL, McLeish K, Keevil SF. Impact of electromagnetic field exposure limits in Europe: is the future of interventional MRI safe? Acad Radiol 2005;12: COCIR (European Coordination Committee of the Radiological, Electromedical and Healthcare IT Industry). EMF Position Paper. ifa.pdf Keevil SF, Gedroyc W, Gowland P, et al. Electromagnetic field exposure limitation and the future of MRI. Br J Radiol 2005;78: ICNIRP, International Commission on Non-Ionizing Radiation Protection.Guidelines on limits of exposure to static magnetic fields. Health Phys 1994;66: Fuentes MA, Trakic A, Wilson SJ, Crozier S. Analysis and measurements of magnetic field exposures for healthcare workers in selected MR environments. IEEE Trans Biomed Eng 2008;55: IEEE (Institute of Electrical and Electronics Engineers). C95.6: standard for safety levels with respect to human exposure to electromagnetic fields (0 3 khz). New York: IEEE; ICNIRP.International Commission on Non-Ionizing Radiation Protection Guidelines on limits of exposure to static magnetic fields. Health Phys 2009;96: de Vocht F, van Drooge H, Engels H, Kromhout H. Exposure, health complaints and cognitive performance among employees of an MRI scanners manufacturing department. J Magn Reson Imaging 2006;23:

6 1228 de Vocht et al. 10. Bradley JK, Nyekiova M, Price DL, Lopez LD, Crawley T. Occupational exposure to static and time-varying gradient magnetic fields in MR units. J Magn Reson Imaging 2007;26: Crozier S, Liu F. Numerical evaluation of the fields induced by body motion in or near high-field MRI scanners. Prog Biophys Mol Biol 2005;87: Crozier S, Wang H, Trakic A, Liu F. Exposure of workers to pulsed gradients in MRI. J Magn Reson Imaging 2007;26: Riches SF, Collins DJ, Charles-Edwards GD, et al. Measurements of occupational exposure to switched gradient and spatially-varying magnetic fields in areas adjacent to 1.5 T clinical MRI systems. J Magn Reson Imaging 2007;26: Riches SF, Collins DJ, Scuffham JW, Leach MO. EU Directive 2004/40: field measurements of a 1.5 T clinical MR scanner. Br J Radiol 2007;80: Cavagnetto F, Prati P, Ariola V, et al. A personal dosimeter prototype for static magnetic fields. Health Phys 1993;65: Cavin I, Gowland P, Glover P, et al. Static B0 field monitoring at 3 Tesla and 7 Tesla using an MRI portable Dosemeter. In: Proceedings of ISMRM Workshop on MRI Safety: update, practical information and future implications. McLean, Virginia Kannala S, Toivo T, Alanko T, Jokela K. Occupational exposure measurements of static and pulsed gradient magnetic fields in the vicinity of MRI scanners. Phys Med Biol 2009;54: de Vocht F, Liket L, De Vocht A, et al. Exposure to alternating electromagnetic fields and effects on the visual and visuomotor systems. Br J Radiol 2007;80: de Vocht F, Glover P, Engels H, Kromhout H. Pooled analyses of effects on visual and visuomotor performance from exposure to magnetic stray fields from MRI scanners: application of the Bayesian framework. J Magn Reson Imaging 2007;26: Glover PM, Cavin I, Qian W, Bowtell R, Gowland PA. Magneticfield-induced vertigo: a theoretical and experimental investigation. Bioelectromagnetics 2007;28: