Quality Control and Quality Assurance of linear accelerators

Autor: Dr. István Polgár

 
The linear accelerators have gone through a notable development since their appearance in the 50's. The conversion from X-ray mode (Bremsstrahlung) and the electron mode is completely automatic which decreases significantly the risk of the treatment with inappropriate modality. The technical advantage like the comprehensive application of latches ensures that the modern linear accelerators are safer than the older ones. In spite of this the equipment failures can cause inaccurate treatments.
Many of the tests presented here are written in the report on Installation and Quality Assurance of Linear Accelerators: IPSM Report 54 (1988) on the Commissioning and Quality Assurance of Linear Accelerators. The emphasis is on the continuous quality control procedures.
Since checks must be carried out with different frequencies, detailed annual quality control program should be created not to miss the less frequent checks in proper time. In some cases it is preferable to carry out checks of similar properties in a rotation system. Thus, the common parts of the system are checked more often than if all the tests are carried out together.

 

Responsibility

 
The quality control is the legal responsibility of the medical physicists who must synchronize the quality control checks performed by different groups. The physicist needs to interpret the discrepancies in terms of what impact it has for the treatment and should give the appropriate advice to the physician. Often there is a contradiction whether a patient should be treated or the equipment failure should be repaired. There is increasing evidence that the omission of treatment - especially fast-growing tumours - has a negative effect on patient outcome (Hendry et al 1996, RCR 1996), and it is better to allow the treatment with a beam different from the optimal than interrupt the patient's treatment. In such cases the radiotherapist must be fully involved in the decision.
The therapy assistants (radiographers) - as the regular users of the equipment - always need to be aware of the device performance and any problems with the equipment. Therefore, they should be involved in the QC program that they can trust to the machines used by them day to day. Often, therapy assistants carry out the daily constancy checks together with the functional safety checks. Other tests are carried out by the medical physicists or by - if any- technicians as required in the protocol. The medical physicist has to coordinate the various groups involved in QC program. If this is not possible, the physical quality control program should be established with the assumption that no other checks are carried out. It is important that when some work is in progress on the linear accelerator a suitably qualified physicist has to make a decision on the availability of this equipment for clinical use. This rule can be alleviated if the work is purely of mechanical-nature, but if there is any doubt it is necessary to consult with an experienced physicist. Because of the diverse nature of the linear accelerators perhaps the application of this rule is more important here than for other equipment. A useful discussion of the issue of maintaining is given by Colligan and Mills (1997).
It should be noted that in our Hungary - unfortunately - in spite of the effective regulation of the minimum requirements the staffing in most Medical Physics Departments is inappropriate. Thus, the regular completion of the QA / QC checks adopted in international protocols is not possible.

 

Tolerances

 
The results of the alignment and dosimetry checks have to be compared with certain criteria in order to decide what action- if it is necessary - need to be taken. These requirements must be clearly stated in departmental protocols.
Routine quality control checks can be done if it is not suitable to take the device out of service to make adjustments. Therefore, an appropriate 'action level' must be defined at which the device should be taken out of service. This can be the same as the manufacturer's specified tolerance or in some cases greater latitude may be allowed. The 'action level' must be locally determined according to the types of treatments being executed. Another tolerance could be established in what the error is and what is acceptable on completion of maintenance. However, the set-up of an individual parameter should be better than the tolerance limit.
The determination of tolerance must be kept in mind that the specification provided by the manufacturer indicates the design tolerance of the equipment.
In spite of this tighter tolerances can be prescribed which probably means more frequent preventive maintenance than it is economically justified. The recommended tolerances are given in this document. These are based on documents such as the IEC Publication 977 (1989) and WHO in 1988, although it must be recognized that the IEC specification is a published standard and the information contained in it are 'guidance to the expected values.' The European Community has defined standards which should be met by all the treatment machines (European Commission 1997). These data can be found in Table 1. (The meaning of data concerning calibration and constancy is not the same as that elsewhere in the book. The calibration relative to the difference between measurement and absolute dose and constancy refer to daily variation rather than to the measurement obtained with a constancy meter as in Section 12.3). It is more important to monitor trends in performance rather than just to state that a given parameter is within specification. Recording the measured value provides better evidence than a tick in a box that the check has been conscientiously performed. On the basis on this it may be predicted the future problems and we can take action in a planned way before the action level is reached.

The level of checks

 
Many functions of an accelerator can be checked using a fast routine procedure, the more strict checks being reserved for occasional checks or when a problem is identified. The quick checks must be reliable to avoid unplanned
shutdowns resulting from wrong measurement equipment and devices intended to quick check must be included in the quality control programme.
For concerning many parameters it will be described in detail which claims quick checks and more strict checks.

 
1. Table, Criteria for acceptibility of linear Accelerators (European Commission 1997)

Test type	Remedial Action level
Gantry and collimator rotation indication	±1 º
Yoke rotation	±0,2 º
Diameter of the Isocentre	±2 mm
Source distance and beam axis indicators	±2 mm
Numerical field size indicators	±2 mm
Light field compared to radiation field	±2 mm
Treatment couch scale	2 mm
Couch deflection under load	5 mm
Immobilisation devices (e.g., casting, etc.)	±2 mm
Patient alignment devices	±2 mm
Light field indication – field size	±2 mm
Light field indication (density measurements)	±1 mm per edge
Dose calibration at reference point	±3% for photons, ±4% for electrons
Output constancy (including accelerators, cobalt and orthovoltage)	±2%
Timer of cobalt unit	±0.01
Electron/photon beam type	Correct type
Beam flatness and symmetry	±3% (photons, electrons and cobalt)
Orthovoltage beam symmetry	±6%
Transmission factor of wedge or compensator	±2%
Dose monitoring system
Precision	±0.5%
Linearity	±1%
Dose rate effect	±2%
Stability	±2%
Variation with gantry angle	±3%

 

References

 
1. IPEM 81 Physics Aspects of Quality Control in Radiotherapy (Edts.: W.P.M. Mayles, R. Lake, A. McKenzie, E.M. Macaulay, H.M. Morgan, T.J. Jordan and S.K. Powley) The Institute of Physics and Engineering in Medicine. York, 1999 ISBN 0 904181 91 X

 

---