Abstract
Background
Quality assurance (QA) of computed tomography (CT) systems is one of the routine tasks for medical physicists in the Nordic countries. However, standardized QA protocols do not yet exist and the QA methods, as well as the applied tolerance levels, vary in scope and extent at different hospitals.
Purpose
To propose a standardized protocol for acceptance and constancy testing of CT scanners in the Nordic Region.
Material and Methods
Following a Nordic Association for Clinical Physics (NACP) initiative, a group of medical physicists, with representatives from four Nordic countries, was formed. Based on international literature and practical experience within the group, a comprehensive standardized test protocol was developed.
Results
The proposed protocol includes tests related to the mechanical functionality, X-ray tube, detector, and image quality for CT scanners. For each test, recommendations regarding the purpose, equipment needed, an outline of the test method, the measured parameter, tolerance levels, and the testing frequency are stated. In addition, a number of optional tests are briefly discussed that may provide further information about the CT system.
Conclusion
Based on international references and medical physicists’ practical experiences, a comprehensive QA protocol for CT systems is proposed, including both acceptance and constancy tests. The protocol may serve as a reference for medical physicists in the Nordic countries.
As described in the first policy statement of the International Organization for Medical Physics (IOMP), the function and responsibilities of medical physicists are oriented towards clinical service (including technical and radiation safety), management, education, research, and development. The role of a medical physicist may vary depending on a country's socioeconomic background, academic community, education, healthcare and regulatory policies, and multi-professional practices in healthcare. One of the principal responsibilities and functions of a medical physicist is the establishment, implementation, and supervision of quality assurance (QA) programs. This includes the acceptance and commissioning of new equipment and the technical supervision of maintenance. Combined with subsequent annual constancy controls, such testing is a valuable tool for monitoring the condition of a radiological imaging system and to detect subtle changes in performance to ensure that equipment always meets regulatory and vendor specifications (1). The significance of such tests has been recognized by the international community, such as The International Electrotechnical Commission (IEC), The American Association of Physicists in Medicine (AAPM), and Institute of Physics and Engineering in Medicine (IPEM). Publications by such organizations form a basis for the regular monitoring of computed tomography (CT) systems (2–6). Further, optimization of the whole diagnostic process, including the delivery of new clinical applications and procedures are dependent on these tests. Acceptance and constancy testing are therefore an essential part of QA, required in international standards and legislation in the Nordic countries for parameters influencing image quality and radiation dose (7–11).
Finnish QA recommendations for digital radiology modalities were published by the Radiation and Nuclear Safety Authority of Finland (STUK) in 2008 and cover all main X-ray modalities, including CT. The recommendations supplement the acceptability requirements for X-ray equipment used in healthcare (radiography, fluoroscopy equipment, and CT) which were given by the Finnish authority in 2006 (8, 12). Swedish minimum requirements regarding QA for digital radiology modalities were revised by the Swedish Radiation Safety Authority (SSM) in 2009 and cover all main X-ray modalities, including CT. These regulations do not specify what methods should be used, or which acceptance levels that should be applied. The general advice revised by SSM in 2009 regarding performance specifications when purchasing equipment for X-ray diagnostics, states recommendations regarding what acceptance levels to use for X-ray equipment used in healthcare (9, 13). Danish minimum requirements regarding QA of digital radiology modalities were published in 1998 by the National Institute of Radiation Protection (SIS). The regulation does not include the specified methods to be used, nor which acceptance levels to comply (10). The Norwegian legislation states that all X-ray equipment used for medical imaging should be included in a QA programme. Acceptance testing should be performed on all new modalities and a periodical quality control regime should be established. Parameters influencing image quality and radiation dose should be tested. No recommendations are given regarding modality specific tests, methods, or tolerance levels (11).
Despite the similar technical and safety cultures in the Nordic countries, there has been a tradition to develop local routines for acceptance and constancy testing in each healthcare region, based on national recommendations. Previous attempts have been made to propose a standardized QA program for radiology equipment within the Nordic region (14). However, there is still no inter-Nordic recommendation for CT stating which tests should be performed and which tolerance levels should be used. A standardization of the Nordic QA practices is warranted in order to facilitate the acquisition of comparative QA data among the neighboring countries, and to enable more data coverage for further optimization of the imaging process.
The purpose of this paper is to provide medical physicists in the Nordic countries with a comprehensive QA protocol for acceptance and constancy testing of clinical CT systems. The recommendations are based on international standards, national publications, accumulated data, and practical experiences. The most essential QA tests for diagnostic CT scanners are presented, while tests related specifically to radiotherapy CT simulators have been excluded.
Material and Methods
With the aim to develop a recommendation for a standardized QA regime for CT, a work group within the Nordic Association for Clinical Physics (NACP) was formed. The group consisted of medical physicists represented from all the Nordic countries (except Iceland). Regular meetings were arranged both by telephone and in person where these issues were discussed. The recommendations describe essential CT tests, how often they should be performed, and what tolerance levels should be applied.
The terminology regarding tolerance levels has been adopted from IPEM Report 91 (6). A remedial and a suspension level are given for each test, where they are applicable. If a test exceeds a remedial tolerance level, corrective actions should be made, based on the severity of the fault, at latest during the next regular service. Exceeding the remedial level may warrant restrictions on the use of a scanner, depending on the test. Also, the follow-up of such parameters and an initial survey of the cause of the fault are typically advisable if the remedial actions are scheduled later. If a test exceeds a suspension level, the corrective actions to restore the performance of the equipment should start immediately. For tests where a suspension level has not been specified, failure to meet the remedial criteria may never result in a critical risk for suspending the system. Actions should therefore be identical to those taken when a test exceeds the remedial tolerance level, described above. Guidance on actions if a test exceeds a given tolerance level has been described in reference (3). It should be the responsibility of the person performing the tests (e.g. the medical physicist) to report to the owner of the equipment or to the service department of the hospital, if any test exceeds a tolerance level. Details on reporting may be found in reference (4).
The frequency at which each test is suggested to be performed is classified by either acceptance (A), to be performed only once when a CT scanner is installed, or constancy (C), to be performed annually. Tests that should be performed at both acceptance and constancy are marked A, C. For some of the tests, relative tolerance levels at constancy are given from a baseline value, which is determined during acceptance testing. Additionally, a subset of the described tests may also be performed after service, or whenever vital parts of the CT scanner have been repaired, replaced, or after major software updates (corresponding to a new acceptance test for the applicable parameters). Many of the recommended tests have been described in the literature (2–6, 15, 16). For others, the methods and tolerance levels are based on practical experience and discussions within the work group. These have been referenced as NACP CT group. If the test methods, frequencies or tolerance levels given by the vendor are stricter than stated here, vendor instructions should be applied to ensure conformity with the technical specifications.
In each radiological organization, the extent of testing should be defined in the QA program and coordinated by a qualified medical physicist. Tests can be performed by a medical physicist, service engineer or, in part, by a responsible technologist with appropriate training. Measuring equipment should have a valid calibration traceable to primary standards. Unless otherwise stated, all scans should be performed in axial scan mode, i.e. data acquisition should be performed while the patient table remains stationary. Prior to the acceptance testing, for each new scanner, initial scan settings for a standard head and a standard body protocol should be identified. Such settings are commonly stated by the vendor in the scanner documentation. Unless otherwise stated, all scans should be performed with the phantom, test object, or the active volume of the detector positioned in the isocenter of the scanner, as defined by the positioning lasers. The utilized measurement device, phantom, scan parameters, and geometry settings for all tests should be documented for repro-ducibility at later constancy tests. Image analysis may be performed either manually or with dedicated software. Commercial or software packages developed in-house may both be appropriate for this purpose. The definition of the geometry of a CT scanner, which will be used in this paper, is presented in Fig. 1.
The geometry and axes of a CT scanner used in the detailed test descriptions
The pixel values in a CT image represent a mapping of the linear attenuation coefficients onto a linear scale of numbers, called Hounsfield units (HU). The HU for a given material, with linear attenuation coefficient, m, has been defined in reference (2, 3). The full width at half maximum (FWHM) of a line profile peak is defined as the interval parallel to the abscissa between the points on the curve at one-half of the peak value corrected for background (3). The volume CT dose index (CTDIvol) is a standardized measure of the radiation output of a CT system, measured in a cylindrical acrylic phantom that enables users to gauge the amount of emitted radiation and compare the radiation output between different scan protocols or scanners. Complex calculations are required to approximate patient dose from CTDI metrics. Several publications have suggested new dosimetry methods which focus on a more robust description of the radiation output from modern CT scanners. These methods can account for, among other things, wide beam collimations and dynamic pitch scanning, which are not readily accounted for using the current standard formalism (17, 18). For QA purposes, however, it is sufficient to utilize the well described and standardized CTDI formalism (15).
To date, there exists no standardized terminology used by the available CT vendors on the market for scan acquisition and reconstruction parameters. The terminology used in this paper may be biased towards a few specific vendors. It should however be possible to translate a scan setting to the corresponding one from another vendor from the context of the test description. AAPM has started working on translating the terminology between vendors to make an intermanufacturer standard (19).
Results
The recommended QA tests are presented in Appendix 1 as a numbered list of detailed descriptions. Each description provides the purpose for a test as well as a suggested method for performing the test, including equipment used, evaluation of the test, the measured parameter, tolerance levels (remedial and suspension), and testing frequency. The detailed test descriptions are summarized in Table 1 in the Appendix 2. In addition to the tests presented in these recommendations, optional tests exist that provide a more comprehensive overview of the CT system. However, due to the practical scope of this paper, and to the limited number of available references, the optional tests are mentioned only briefly in Appendix 3.
Discussion
The purpose of this study was to propose a standardized protocol for acceptance and constancy testing of diagnostic CT systems based on international literature and practical experience within the work group. In this paper we recommend the most essential tests needed for performing QA, together with basic operational parameters. In order to make the testing procedures more accessible to the reader, phantom and dosimeter models have not been specified in detail. Dosimeters calibrated for diagnostic CT beam qualities, are available from a number of vendors. Many commercial image quality phantoms with a wide variability of features exist. One example of an image quality phantom, used by the majority of the NACP CT group, is the Catphan 500 or 600 phantom (The Phantom Laboratory, Salem, NY, USA).
Future work
It is important that the recommendations provided in this paper are revised and updated regularly. The frequency of such an update should not necessarily be defined by time but should rather be triggered by changes in CT technology, and by publication of new and updated literature and standards. Also, the authors of this report hope to implement the proposed test regime on diagnostic CT imaging systems at their respective hospitals. A database could then be created to store the results from acceptance and constancy testing from all of the participating hospitals. By analyzing the collected data, tolerance levels and test frequencies could be optimized and above all, it could be evaluated which tests are necessary and which should be considered as optional.
Most of the tests outlined in this report are also applicable on radiotherapy dedicated CT imaging systems. However, CT simulators have higher demands on mechanical functionality and system calibration (considering treatment planning). Test procedures that should be considered should include additional table and laser positioning controls and electron density calibrations. Further, modern CT scanners include advanced image reconstruction and dose reduction tools that have not been discussed in this paper. Specifically, technology such as dual energy, dual tube-detector configurations, partial rotation radiation techniques, and iterative reconstructions should be evaluated by separate tests. Finally, future dosimetry standards should be implemented, taking into account the ever increasing beam width of modern CT scanners.
In conclusion, this report has proposed a comprehensive test regime for acceptance and constancy tests for diagnostic CT scanners. Test procedures and tolerance levels are based on international guidelines and the authors’ practical experiences in testing CT imaging systems. The work group hopes that the recommendations given in this report could serve as a reference for medical physicists in the Nordic countries.
Footnotes
Acknowledgements
Although not officially approved by the NACP, the authors would like to thank the NACP for encouraging this work and for supporting us with practical matters during our meetings. In addition to the authors, the following physicists contributed to this work: Toni Ihalainen (Finland), Ingerid Skjei Knudsen (Norway), Inger Mathilde Larsen Kofoed (Denmark), and Hans Bomholt Rasmussen (Denmark).
Appendix 1
Detailed test descriptions