16 Quality Assurance in Radiation Oncology
James A. Purdy, Eric Klein, Srinivasan Vijayakumar, Carlos A. Perez, and Seymour H. Levitt
J.A. Purdy, PhD, S. Vijayakumar, MD
Department of Radiation Oncology, University of California Davis Medical Center, Sacramento, CA 95817, USA
E. Klein, MS, C.A. Perez, MD
Department of Radiation Oncology, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St.
Louis, Missouri 63110, USA S.H. Levitt, MD
Department of Therapeutic Radiation Oncology, University of Minnesota, Minneapolis, MN 55455, USA
16.1
Introduction
Dose precision in radiation therapy is expected to be on the order of ±5%, based on the fact that cer- tain tumors and normal tissues exhibit steep dose response curves (Herring and Compton 1971).
Delivery of radiation with this criterion places great demands on the entire process, although such a level is believed to be achievable (ICRU 1976).
Uncertainties in treatment are due to many fac- tors including: (a) dose calibration at a point in phantom; (b) patient-specific data used for treat- ment planning; (c) dose calculation in the patient;
(d) transfer of the treatment plan to the radiation therapy machine; and (e) day-to-day variations in patient positioning and internal motion of tumor volume and organs at risk. These uncertainties may be categorized as systematic and random. Random uncertainties vary in magnitude and sign and cannot be totally controlled (e.g., the position of the radiation field on the patient may vary from day to day by a few millimeters). Moreover, the degree with which treatments can be reproduced differs among clinical sites and between institutions. Systematic uncertainties maintain their magnitude and direc- tion over a period of time. For example, the use of an incorrect factor in the calibration of a treatment unit would have the same effect on the dose delivered to all patients. Systematic errors, in principle, should be controllable: for example, the degree of misregistra- tion of field defining apertures can be reduced with periodic review of beam localization films; however, many systematic errors remain, e.g., the approxima- tions used in dose calculation algorithms.
We first need to establish some definitions. Qual- ity assurance (QA) is defined as the set of policies and procedures instituted to ensure the proper and safe delivery of the prescription dose to the patient.
Quality control constitutes the actual tests taken to maintain and improve the quality of the treatment.
We must also understand that a QA program is an interdisciplinary effort involving radiation oncolo-
CONTENTS
16.1 Introduction 395
16.2 Goals and Structure of a QA Program 396 16.2.1 Physics Staffing 396
16.2.2 Training 397
16.2.3 Structure of a QA Program 398 16.3 Dosimetry Instrumentation QA 399 16.4 Medical Linear Accelerator QA 400 16.4.1 Linac Annual QA Tests 405 16.5 Linac Advanced-Technology QA 405 16.5.1 Linac Computer Control System 406
16.5.2 Asymmetric Jaws (Independent Collimation) 407 16.5.3 Dynamic Wedge 408
16.5.4 Multileaf Collimation 408
16.5.5 Online Electronic Portal Imaging 409 16.5.6 Intensity-Modulation Radiation Therapy 411 16.6 Quality Assurance of Cobalt Teletherapy Units 412 16.7 Treatment-Machine Maintenance 413
16.8 Treatment-Planning Computer System QA 414 16.9 Treatment-Planning QA 415
16.9.1 Patient Immobilization and Data Acquisition 415 16.9.2 Critical Structure, Tumor, Target Volume Delineation 416
16.9.3 Designing Beams and Field Shaping 416 16.9.4 Dose Calculation 416
16.9.5 Computation of Monitor Units (Time) 416 16.9.6 Plan Evaluation 417
16.9.7 Treatment-Plan Review 417 16.9.8 Beam Modifiers 417
16.9.9 Plan Implementation and Verification 417 16.10 Clinical Aspects of QA 418
16.10.1 Planning Conference 418 16.10.2 Chart Checking 418
16.10.3 Port Film/Image Verification Review 419 16.11 Conclusion 420
References 421
gists, radiation physicists, dosimetrists, and radia- tion therapists. Although medical physicists and radiation therapists are more involved in the tech- nical aspects of QA, and radiation oncologists in the medical aspects, the efforts of each group substan- tially overlap.
Material presented in this chapter relies heavily on recommendations given in various AAPM Task Groups, including TG 35 Report on Medical Acceler- ator Safety Considerations (Purdy et al. 1993), TG 40 Report on Comprehensive QA for Radiation Oncol- ogy (Kutcher et al. 1994), World Health Organiza- tion (WHO 1988), American College of Radiology (ACR 2004), American College of Medical Physics (ACMP) (9), government regulations (NRC 2003), and the authors’ previous QA reports (Parrino and Purdy 1983; Purdy 1983, 1991a,b; Purdy et al. 1986, 1995; van Dyk and Purdy 1999).
16.2
Goals and Structure of a QA Program
A series of publications known as the “Blue Book”
have provided a strong rationale for the development, purpose, and need for QA in radiation oncology. Five such reports were published over the period 1968 to 1991 including: A Prospect for Radiation Therapy in the United States (1968); A Proposal for Integrated Cancer Management in the United States: The Role of Radiation Oncology (1972); Criteria for Radiation Oncology in Multidisciplinary Cancer Management (1981); Radiation Oncology in Integrated Cancer Management (1986); and Radiation Oncology in Integrated Cancer Management (1991). The 1991 version published by The Inter-Society Council for Radiation Oncology (ISCRO) provides the following statement regarding the purpose of a QA program (ISCRO 1991):
“The purpose of a Quality Assurance Program is the objective, systematic monitoring of the quality and appropriateness of patient care. Such a program is essential for all activities in Radiation Oncology.
The Quality Assurance Program should be related to structure, process and outcome, all of which can be measured. Structure includes the staff, equipment and facility. Process covers the pre- and post-treat- ment evaluations and the actual treatment applica- tion. Outcome is documented by the frequency of accomplishing stated objectives, usually tumor con- trol, and by the frequency and seriousness of treat- ment-induced sequelae.”
The report emphasizes that the complexity of radiation therapy requires a teamwork approach among radiation oncologists, medical physicists, dosimetrists, nurses, and therapists as no one indi- vidual has all the skills necessary. This series of pub- lications has not been updated in over a decade, and now more than ever, there is a definite need to do so. Most important is for administrators to under- stand the need for a robust radiation oncology QA program, and to work with the radiation oncology team and ensure that adequate funding is available to support such a program. For a QA program to be effective, all of the faculty and staff involved with providing radiation therapy to patients must be committed to the QA program.
16.2.1
Physics Staffing
Appropriate physics staffing is an essential compo- nent of the radiation oncology QA program. In the past, staffing guidelines were promulgated via the Blue Book and were based on patient load and treat- ment equipment. The 1991 Blue Book recommended at least one clinical physicist per center for up to 400 patients treated annually (Table 16.1a; ISCRO 1991).
Additional clinical physicists are recommended in the ratio of one per 400 patients treated annually.
This report makes clear that these staffing levels are for clinical duties only, and additional full-time equivalent (FTEs) medical physicists will be required for translational research, teaching, and adminis- tration duties; however, the present physics staffing levels must also take into account the complexity of treatments being performed in the clinic such as IMRT, brachytherapy, and stereotactic radiosurgery, as such procedures are physics intensive.
The most detailed information currently avail-
able regarding medical physics work effort is in
the reports by Abt Associates (1995, 2003). These
reports were the result of the American College of
Medical Physics (ACMP) and the American Associa-
tion of Physicists in Medicine (AAPM) engaging Abt
Associates to conduct a study that measured what
was termed Qualified Medical Physicist (QMP) work
for medical physics services, and to develop a rela-
tive work value scale depicting the relative amount
of QMP work required for each medical physics
service. The results of that survey were published
in 1995 (Abt Associates 1995). This report was
updated in 2003 due to the recognition that the many
changes in medical physics practice and technology
that had occurred since the original report may have affected QMP work-related values (Abt Associates 2003). Herman et al. used the data from the first Abt report along with a manpower study conducted by the American College of Medical Physics (ACMP) and AAPM, accounting for IMRT and other special procedures, to show that current reimbursement models do not adequately support the needed phys- ics QA effort, particularly for those clinics involved with IMRT (Herman et al. 2003). Herman et al. at the 2005 AAPM Annual Meeting presented an algo- rithm making use of the Abt-II study and the pre- viously referenced work survey data to determine medical physics FTE recommendations depending on number of procedures and types of procedures.
An example of this type of FTE estimation is given in Table 16.1b (Herman et al. 2005).
Blue Book recommendation on dosimetrist staff- ing is 1 per 300 patients treated annually (ISCRO 1991). This number appears to still be valid; how- ever, it should be noted that in some institutions, dosimetrists not only provide treatment-planning services, but also participate in QA tests, and in some cases, perform the simulations. In those cases, staffing levels must be increased accordingly.
16.2.2 Training
The education and training of the radiation oncol- ogy team and their continuing education are of critical importance to a QA program. In the past, clinical physics training and dosimetrists’ training
Table 16.1a. Minimum personnel requirements for clinical radiation therapy. (From 1991 Blue Book: Radiation Oncology in Integrated Cancer Management: Report of the Inter-Society Council for Radiation Oncology)
Category Staffing
Radiation oncologist-in-chief One per program
Staff radiation oncologist One additional for each 200–250 patients treated annually. No more than 25–30 patients under treatment by a single physician
Radiation physicist One per center for up to 400 patients annually; additional in ratio of 1 per 400 patients treated annually
Treatment planning staff
Dosimetrist or physics assistant One per 300 patients treated annually Physics technologist (mold room) One per 600 patients treated annually Radiation therapy technologist supervisor One per center
Staff (treatment) Two per megavoltage unit up to 25 patients treated daily per unit, 4 per megavoltage unit up to 50 patients treated daily per unit Staff (simulation) Two for every 500 patients simulated annually
Staff (brachytherapy) As needed
Treatment aid As needed, usually one per 300-400 patients treated annually
Nursea One per center for up to 300 patients treated annually and an addi-
tional one per 300 patients treated annually
Social worker As needed to provide service
Dietitian As needed to provide service
Physical therapist As needed to provide service
Maintenance engineer/electronics technician One per 2-mV units or 1-mV unit and a simulator if equipment serviced “in house”
Additional personnel will be required for research, education, and administration. For example, if 800 patients are treated annually with three accelerators, one 60Co teletherapy unit, a superficial X-ray machine, one treatment-planning computer, the clinical allotment for physicists would be two to three. A training program with eight residents, two technology students, and a graduate student would require another 1−1.5 FTEs. Administration of this group would require 0.5 FTE. If the faculty had 20% time for research, a total of five to six physicists would be required
a For direct patient care. Other activities supported by LVNs and nurse aides
have been the weakest link, as the training programs lacked organized clinical training beyond individ- ual apprenticeships or self-training on the job. This was probably adequate in the early days of phys- ics involvement in radiation oncology; however, as radiation oncology has become increasingly more sophisticated and complex, this strategy is no longer acceptable. The practice of hiring inadequately trained medical physicists, who are allowed to per- form patient-related tasks, must be discontinued.
The lack of proper clinical training of medical physi- cists reached a serious level in the late 1980s. There was, and continues to be, an acute shortage of quali- fied clinical physicists, e.g., physicists with adequate clinical training and board certification. There was, and continues to be, a growing abundance of phys- ics graduates (including medical physics graduates) with inadequate clinical training applying for hos- pital positions. The AAPM recognized this prob- lem, and in 19881989, developed a comprehensive document entitled AAPM report no. 36, “Essentials and Guidelines for Hospital-Based Medical Physics Residency Training Programs,” which sets down the educational and administrative requirements for a hospital-based residency training program (AAPM 1990). The AAPM report recommends 2 years of clin- ical physics training beyond an M.S. or Ph.D. degree in physics or a closely related field. The organization of the recommended program was patterned after
physician residency programs. In the words of the committee that developed the recommendations,
“This document will hopefully encourage the devel- opment of a high-quality clinical medical physics instructional environment on a nationwide basis and make an important contribution to the protec- tion of the public health, safety, and welfare.”
In October 1992, the Barnes-Jewish Hos- pital/Washington University Radiation Oncology Center established the first Radiation Oncology Physics Residency Program accredited by the Com- mission on Accreditation of Medical Physics Educa- tion Program (CAMPEP) in the United States. More physics residency programs are now being instituted, although at a much slower pace than needed.
Certification boards for physicists exist, but the entry requirements for the examination still do not mandate residency-type clinical training as is required of the physicians. As a result, there is an unchecked influx of inadequately trained physicists into the field.
16.2.3
Structure of a QA Program
At the heart of any modern QA program is a QA/
continuing quality improvement (QA/CQI) commit- tee. The need for such a committee is contained in reports by several United States radiation oncology organizations including the ACR (ACR 2004) and the AAPM (Kutcher et al. 1994), and international organizations such as the World Health Organization (WHO 1988). To properly function, the QA/CQI com- mittee should be created by the radiation oncology departmental chairman and should report directly to the chairman and the hospital administration.
Its function is to design, implement, and maintain a multidisciplinary QA/CQI program whose goal is to improve the quality of patient care. The QA/CQI committee should meet regularly, preferably on a monthly (at least quarterly) basis to review the ongoing QA/CQI program, and these deliberations should be reported to the department chairman in writing. It is important that the committee have the full support of the chairman of radiation oncology and all of the faculty and staff; otherwise, maintain- ing quality and implementing improvements in care will prove extremely difficult.
The structure of a typical QA/CQI committee is shown schematically in Figure 16.1. The commit- tee’s work encompasses numerous areas, and needs the participation of many individuals, including the
Table 16.1b. Example of full-time equivalent (FTE) needs for a modern radiation therapy clinic having three treatment machines (including advanced technology procedures such as IMRT), and including physician residency training and one medical physics resident. A total of 9.4 FTE medical physicists were required. (From Herman et al. 2005)
Procedure Quantity
Total new patients 800
IMRT 250
HDR 50
Radiosurgery 60
Prostate implants 40
TBI 30
Task FTE
Patient procedures 6.67 Commissioning and QA 1.04
Education 0.11
Research 0.78
Administration 0.78
Total FTE 9.38
committee chairman (preferably a radiation oncol- ogist or medical physicist), radiation oncologists, radiation therapists, dosimetrists, nurses, and phys- icists involved in simulation and treatment machine quality assurance, treatment planning, brachy- therapy, nursing, and radiation safety. The number of committee members varies, and in smaller insti- tutions, the committee might be comprised of only three individuals: a radiation oncologist; a medical physicist; and a radiation therapist.
Depending on the institution, the responsibili- ties of the QA/CQI committee members will differ.
Moreover, there are many gray zones in responsibil- ities in the radiation oncology practice. In fact, one of the roles of the committee is to define the lines of demarcation between tasks so that some impor- tant functions are not overlooked. Typical QA/CQI functions and lines of demarcation are listed in Table 16.2.
It is also suggested by the ACR that there should be an annual review of outcome (ACR 2004). Such a review is highly complex, since it includes an under- standing and acceptance of standards of care in the sites under review.
16.3
Dosimetry Instrumentation QA
A list of the type of equipment typically considered most useful in a QA program for treatment machines is given in Table 16.3. Special QA devices for treat- ment machines are now commercially available for checking beam alignment, field symmetry, and the output of the machine. Generally, the radiation QA devices consist of an array of ionization cham- bers or diodes positioned in a plastic phantom (see
Fig. 16.2). Also, a plastic constancy phantom that can be attached quickly to the treatment machine, allow- ing measurements to be made at a fixed and repro- ducible distance, is a useful QA device as it allows measurements to be performed on a daily basis with a minimum of setup time. (see Fig. 16.3).
Accurate data acquisition with automated beam data scanning systems and scanning film densi- tometers requires that the systems be subjected to a systematic performance test prior to use and also undergo periodic QA tests thereafter. Details of acceptance testing and QA of such devices have been reported in the literature (Mellenberg et al.
1990).
Film is satisfactory for assessing beam symmetry and flatness, but one must be cautious if it is used to measure dose (Williamson et al. 1981). For exam-
Chairman
Radiation Oncology Administration
Clinical Engineering
Treatment Outcome
Brachytherapy
Treatment Delivery
External Beam Treatment Equipment
Treatment Planning
Nursing Personal Safety Quality Assurance
Continuing Quality Improvement
Committee
Patient Safety
Fig. 16.1 Structure of a quality assurance (QA) committee il- lustrating typical areas addressed
Table 16.2 Typical QA responsibilities and functions Review by a medical physicist of ongoing QA of external beam equipment including treatment units, calibration equipment, and imaging modalities (CT simulators, conven- tional simulators, port films or electronic portal imaging devices, cone-beam CT, megavoltage CT, ultrasound, etc.) Review by a medical physicist of ongoing QA in brachy- therapy, including instrumentation, handling of sources, treatment planning, and remote afterloaders/applicators operation
Review by a medical physicist and a medical dosimetrist of QA in treatment planning which includes the treatment planning system and peripherals, graphical planning, in vivo dosimetry, and plan review
Review by a radiation therapist of procedures for dose deliv- ery to the patient, which includes verification of patient treat- ment setup parameters, chart checking, portal films/images, and patient safety
Review by a physician of patient QA procedures which includes weekly chart review, port film/images review, and mortality and morbidity assessment (typically reviewed in a separate committee)
Review of any cases in which there have been deviations out- side the action levels set by the QA/CQI committee, depart- ment, hospital, and regulatory bodies
Regularly scheduled audits of charts, films, and QA proce- dures. The results should be presented to the QA/CQI com- mittee
Recommend actions to be taken as a result of the problems encountered in the reports and audits. Approve or suggest modifications to corrective actions
Assurance that the recommended actions have been taken Report the QA/CQI committee’s results, actions, and recom- mendations to the chairman of radiation oncology and to the hospital QA committee
Supervision of staff continuing education programs
ple, for some types of radiographic film the dose is not a linear function of optical density so that it is necessary to calibrate the film and then make corrections to the densitometer readings. It is also important to make sure that processing conditions are maintained so that the measurements are repro- ducible. Another problem with the film technique is that the results of measurements are not immedi- ately known, i.e., after exposing the film, it must be developed and then read on a densitometer.
Accurate data acquisition with automated beam data scanning systems and scanning film densitometers requires that the systems be sub- jected to a systematic performance test prior to use. Details of acceptance testing of such devices have been reported in the literature (Holmes and McCullough 1983; McCullough and Holmes 1985). These devices should also undergo periodic QA tests thereafter.
16.4
Medical Linear Accelerator QA
The decision to purchase and implement a megavolt- age radiation therapy linear accelerator for clinical use carries with it a commitment to provide adequate staff, test equipment and instrumentation, and allow the necessary machine time in order to assure that the unit is performing according to specifications (Nath et al. 1994).
An effective QA program for treatment machines establishes criteria for optimum machine perfor- mance, monitors adherence to established criteria, ensures the accuracy of the dose delivered, mini- mizes treatment machine downtime, and enhances communication between radiation oncologists, physicists, dosimetrists, therapists, and mainte- nance technicians.
The prescribed dose planned and the dose deliv- ered by a linear accelerator is dependent on several parameters including the dose calibration, percent- age depth dose, and other dose ratios used in cal- culating the dose distribution and machine monitor unit settings, off-axis beam characteristics, wedge and block factors, multileaf collimator (MLC), cali- bration, etc. All parameters must be carefully deter- mined when the therapy machine is installed during machine commissioning. Quality assurance proce- dures must then be implemented to ensure the accu- racy and reproducibility of the dose delivered by the linear accelerator.
Table 16.3 Typical QA instrumentation/equipment needed for a modern radiation oncology clinic
Secondary standard dosimetry system Field use dosimetry system
Parallel plate ionization chamber
Standard and field use barometers and thermometers Plastic phantom for output calibration constancy checks Small water phantom with movable fixed ion chamber holder
Array detector device to monitor beam symmetry Polystyrene, solid water, substitute tissue heterogeneity stack phantoms
Specialized phantoms and instrumentation for treatment- planning systems
Specialized phantoms and instrumentation for simulator, CT simulator, and linac onboard imaging systems Anthropomorphic phantom
Film densitometer system Beam data scanning system
In vivo dosimetry system (Diodes, MOSFETs, and/or TLD)
Fig. 16.2 Example of an array detector (MapCHECK) used for radiation therapy QA measurements. (Courtesy Sun Nuclear Corporation)
Fig. 16.3 Plastic photon output calibration constancy phan- tom. (Courtesy Valiant Instruments)
A QA program for linear accelerators requires certain key ingredients if it is to be successful; these include: (a) a commitment by the staff to QA; (b) participation of radiation, electronic, and mechani- cal specialists; (c) active radiation therapists partici- pation; (d) regularly scheduled QA and preventive maintenance inspections of the linac; (e) agreed on QA machine performance tests and acceptance cri- teria; (f) adequate test instrumentation; (g) accurate and complete documentation of the linac, and (h) commitment to good maintenance and radiation calibration record keeping (bound archival records or computer database). It should be recognized that a QA program for treatment machines is very much a team effort and will at some stage call on the exper- tise of the physicist, dosimetrist, therapist, mainte- nance technician, and radiation oncologist.
A typical linac QA program is designed to pro- vide testing at different levels and frequencies; these typically include: (a) daily checks, performed each morning by the radiation therapist who normally operates the machine; (b) weekly checks performed by a physicist, dosimetrist, therapist, or a physics
resident; (c) monthly checks performed by a dosime- trist, physics resident, or physicist, and preventive maintenance inspections performed by an accel- erator maintenance technician (radiation oncology clinical engineer); and (d) annual full calibration performed by a qualified medical physicist.
The responsibility of performing the various tasks is divided among physicists, physics residents, dosimetrists, maintenance technicians, and thera- pists, but the exact distribution is not critical. What is essential is that each individual competently per- form and record the results of their tests on a reg- ular basis, and that the overall responsibility for a machine QA program be assigned to one individual, generally the medical physicist.
All measurements should be recorded chrono- logically in bound notebooks or in a computer data- base. Such records and the data contained therein are a valuable resource in maintaining the treatment machine. All parties involved should receive peri- odic reports on the QA measurement results.
The QA acceptance criterion should be estab- lished for each of the constancy checks performed.
Table 16.4 Quality assurance of simulators. (From AAPM TG 40)
Frequency Procedure Tolerance
Daily Localizing lasers 2 mm
Distance indicator (ODI) 2 mm
Monthly Field-size indicator 2 mm
Gantry/collimator angle indicators 1q
Cross-hair centering 2-mm diameter
Focal spot-axis indicator 2 mm
Fluoroscopic image quality Baseline Emergency/collision avoidance Functional Light/radiation field coincidence 2 mm or 1%
Film processor sensitometry Baseline Annual Mechanical checks
Collimator rotation isocenter 2-mm diameter Gantry rotation isocenter 2-mm diameter Couch rotation isocenter 2-mm diameter Coincidence of collimator, gantry 2-mm diameter Couch axes and isocenter 2 mm
Table-top sag 2 mm
Vertical travel of couch Radiographic checks
Exposure rate Baseline
Table-top exposure with fluoroscopy Baseline kVp and mAs calibration Baseline High and low contrast resolution Baseline The tolerances mean that the parameter exceeds the tabulated value (e.g., the measured isocenter under gantry rotation exceeds 2-mm diameter)
Table 16.5. Quality assurance of medical linear accelerators (from AAPM TG 40)
Frequency Procedure Tolerance
Daily Dosimetry
X-ray output constancy 3%
Electron output constancy 3%
Mechanical checks
Localizing lasers 2 mm Distance indicator (ODI) 2 mm Safety Interlocks
Door interlock Functional Audiovisual monitor Functional Monthly Dosimetry
X-ray output constancy 2%
Electron output constancy 2%
Backup monitor constancy 2%
X-ray central axis dosimetry parameter (PDD, TAR) constancy
2%
Electron central axis dosimetry parameter constancy (FDD)
2 mm at therapeutic depth X-ray beam flatness constancy 2%
Electron beam flatness constancy
3%
X-ray and electron symmetry 3%
Safety interlocks Functional Emergency off switches Functional Wedge, electron cone
interlocks
Functional Mechanical checks
Light/radiation field coincidence
2 mm or 1%
on a side Gantry/collimator angle
indicators
1q
Wedge position 2 mm (or 2%
change in transmission factor)
Tray position 2 mm
Applicator position 2 mm Field-size indicators 2 mm Cross-hair centering 2-mm
diameter Treatment couch position
indicators
2 mm/1q Latching of wedges, blocking
tray
Functional
Jaw symmetry 2 mm
Field light intensity Functional
Frequency Procedure Tolerance
Annual Dosimetry
X-ray/electron output calibration constancy
2%
Field-size dependence of X-ray output constancy
2%
Output factor constancy for electron applicators
2%
Central axis parameter constancy (PDD, TAR)
2%
Off-axis factor constancy 2%
Transmission factor constancy for all treatment accessories
2%
Wedge transmission factor constancy
2%
X-ray output constancy vs gantry angle
2%
Electron output constancy vs gantry angle
2%
Off-axis factor constancy vs gantry angle
2%
Arc mode Manufactur-
er’s specifica- tions Safety interlocks
Follow manufacturers test procedures
Functional Mechanical checks
Collimator rotation isocenter 2-mm diameter Couch rotation isocenter 2-mm
diameter Coincidence of collimetry,
gantry, couch axes with isocenter
2-mm diameter Coincidence of radiation and
mechanical isocenter
2-mm diameter Table-top sag
Vertical travel of table
2 mm 2 mm
The tolerances listed should be interpreted to mean that if a parameter either (a) exceeds the tabulated value (e.g., the mea- sured isocenter under gantry rotation exceeds 2 mm diameter), or (b) the change in the parameter exceeds the nominal value (e.g., the output changes by >2%), then an action is required.
The distinction is emphasized by the use of the term constancy for the latter case. Moreover, constancy, percent values are + the deviation of the parameter with respect its nominal value;
distances are referenced to the isocenter or nominal SSD. All electron energies need not be checked daily, but all electron energies are to be checked at least twice weekly. Whichever is greater should also be checked after change in light-field source. Jaw symmetry is defined as difference in distance of each jaw from the isocenter. Most wedges’ transmission factors are field size and depth dependent
The frequency of each QA procedure to be performed depends primarily on the stability of the parameter tested, based on one’s own experience. Tables 16.4 and 16.5, adapted from the AAPM’s TG 40, lists the recommended QA tests with frequency and tol- erance values for simulators and medical linacs, respectively (Kutcher et al. 1994).
For some tests, only a very quick “observation- only” type test is required. For example, a weekly
“light-radiation field congruence” test result can be analyzed by simply looking at a radiograph showing the light field and radiation field edges and observ- ing whether or not the agreement appears reason- able. Other tests require a more careful quantitative analysis. For example, a quantitative “light-radia- tion field congruence” test would require that the film be analyzed using a film densitometer and that the results be carefully plotted on graph paper in order to determine the edge agreement precisely.
The QA tests should be designed to be quick and reproducible checks on key parameters, if they are to be accepted and performed faithfully. The discus- sion provided below is for general guidance and the actual QA tests required at a particular institution must be developed by that institution.
Recommended daily checks are listed in Table 16.5. The manufacturer’s instructions for start-up and operation of the accelerator should be followed and readings of the various meters, dials, and gauges recommended for monitoring should be recorded. The daily readings should be maintained in a logbook or computer database. These data pro- vide performance trends of a particular component which are helpful in isolating faults and may even alert one to a developing problem before full com- ponent failure occurs.
Daily treatment room checks include testing the functionality of the treatment room door interlock, intercom and closed circuit monitor system, and radiation warning lights. Treatment checks include the accuracy of the optical distance indicator (ODI), the alignment of the laser localization lights, and the radiation output calibration constancy for all photon treatment modes.
The photon-beam radiation output calibration constancy for each of the photon energies used should be checked daily for the reference geometry (e.g., 10u10 cm, 100 cm SSD or SAD). This can be accomplished efficiently using an ion chamber in a simple plastic phantom which contains a fitted hole at a standard depth from the top surface for the ion chamber, and which attaches to the accelerator at the standard SSD. A cylindrical ion chamber (e.g.,
Farmer type) or other type of ion chamber can be used for the test. The ion chamber reading should be corrected for temperature and pressure and con- verted to dose using predetermined factors and the output value should be compared with the value established at the time of the last full calibration.
The electron beam radiation output calibration constancy for each of the electron energies used for the reference geometry applicator (e.g., 14u14 cm) should be checked once or twice weekly. This can be accomplished efficiently using a plastic stack phan- tom and ion chamber dosimetry system (Fig. 16.4).
The plastic constancy phantom used for the photon beam output check is also convenient for checking the ODI and the laser localization lights.
A visual inspection of where the ODI indicator image strikes the output constancy phantom sur- face generally suffices as a daily check. Tests using a mechanical front pointer can be performed when a more quantitative test is needed. The vertical and sagittal lines of the laser localization lights should pass through the central axis of the beam. This may also be checked using marks on the top surface of the plastic constancy phantom and observing the intersection of the laser lines with the image of the cross hair.
Light field radiation congruence and radiation field symmetry is typically checked monthly but, in some cases, may need to be checked more often. This test can be performed by exposing a film placed per- pendicular to the central axis of the beam. The film is aligned with metal markers placed on the edges of the light field (or alternatively pressure marks from a sharp device or pin prick) to mark the position of the light field. A plastic sheet of adequate thickness to provide electronic build-up is placed over the film. The developed film can be analyzed visually
Fig. 16.4 Plastic electron output calibration constancy phan- tom. (Courtesy Valiant Instruments)
for the weekly test and with a film densitometer for a monthly quantitative check.
The alignment of the intersection of the cross hairs with the center of the light and radiation field should be checked regularly. This is easily accom- plished by placing a film at the isocenter perpendic- ular to the beam and marking (e.g., pin prick) the intersection of the cross-hairs on the film. The rela- tive motion of the cross-hairs around the pin-prick should be observed as the collimator is rotated r90q.
The center of the radiation field can be determined from the exposed film and compared with the posi- tion of the pin prick.
Symmetry can be checked using a device con- taining an array of detectors (usually four or five) and compared with the values obtained at the last full calibration. Symmetry can also be checked with a film in a plastic phantom placed perpendicular to the beam central axis for a large field. It should be noted that for “bent-beam” linear accelerators, where the electron scattering foil and the photon flattening filter are moved in and out of position, more frequent checks of beam symmetry may need to be performed.
The QA checks discussed above require about 15
20 min per treatment machine to complete.
Additional monthly checks for the linac are listed in Table 16.6. For example, the monthly preventa- tive maintenance program for each machine should include regular safety checks for all electrical and mechanical interlocks. All “emergency off” switches on the machine should be checked periodically.
Collimator and gantry angles, field size indica- tors, and the mechanical distance indicator should be checked at least monthly. Using a simple level and rotating the gantry, the gantry angle indicators can be checked at the vertical and horizontal posi- tions. The collimator angle indicators and the field size indicators at selected field sizes can be checked using graph paper placed on the treatment couch at the reference distance. Collimator rotation can be quickly checked by observing the movement of the image of the cross hair as the collimator is rotated
± 90o.
Photon beam energy can be checked by measur- ing depth dose at two specified depths, although a more sensitive test is to measure the beam profile in a plastic phantom at a specified depth. For electrons, the relative ionization measured at two depths is usually a sufficient check.
In addition, the radiation output calibration should be checked using a different dosimetry system than that used for the daily output constancy
check. This serves as a redundancy check on both the treatment machine monitor chamber calibra- tion (cGy/MU) and the daily system used to check the output calibration constancy.
Accessories such as wedge filters, electron beam applicators, and blocking tray assemblies --includ- ing the mounting slots and micro-switches -- should be examined for any cracks and potential malfunc- tions.
If onboard imaging is performed, additional checks related to mechanical (collision, interlocks, readout accuracy, centering, etc.), imaging quality (contrast, sensitivity, constancy, etc.), and related software, such as magnification accountability, should be checked on a monthly basis. These tests are discussed in more detail in a later section.
Also, one should ensure that an up-to-date machine operator’s manual is located at the treat- ment machine console. In addition, complex treat- ment techniques require that detailed, unambiguous written procedural instructions also be available at the control console. One should also ensure that the posting of radiation warning signs and emergency instructions have not been removed.
Table 16.6 Recommended monthly QA checks
General conditions of treatment-unit checks: (key switch, monitors, machine movements, pendant, accessories, treat- ment aids, audio-visual/communication, room condition) Review of daily/weekly check logs (daily machine operation check log, daily photon/electron calibration constancy check log, weekly machine maintenance inspection log)
Safety-features checks [operating instructions at machine, emergency instructions displayed, radiation warning sign, treatment-room door interlock, treatment-room operability, beam condition indicator lights (door), beam condition indi- cator lights (monitor), emergency offs]
Mechanical checks (gantry rotation angle, mechanical and digital; rotation arc check of MU/degree; collimator rotation angle indicator check; mechanical and digital; cross-wire deviation)
Radiation/light-field check (visual inspection for ±2-mm tol- erance for light/radiation field congruence and central plane overlap/gap)
Photon beam energy/off-axis factor check
Radiation output calibration constancy check (treatment unit monitor chamber, assigned dosimetry system)
Electron beam energy check (ionization depth ratio) Electron beam output calibration constancy check MLC QA for IMRT
EPID mechanical and imaging checks
Onboard imaging mechanical and imaging checks
The monthly checks may take from 1 to 2 h depending on the number of tests performed and the number of modes and energies available. These times clearly show that a QA program for treatment machines designed to insure the delivery of dose can be implemented in any size radiation therapy clinic at an acceptable cost in terms of time, staff, and equipment. When such a program is neglected and problems are not fixed as they arise, the treatment machines will inevitably deteriorate and the quality of patient treatments will be compromised.
16.4.1
Linac Annual QA Tests
A full calibration of the treatment machine should be performed annually. Suggested tests are listed in Tables 16.5 and 16.7. The basic calibration should be performed in a water phantom using an ion chamber according to an appropriate protocol, e.g., AAPM (AAPM 1983; Almond et al. 1999). The stability of the dose per monitor unit and the beam symme- try should be checked at different gantry angles.
Verification of the output factors and central axis depth dose should be done for several different field sizes. In addition, current values for off-axis factors,
monitor linearity, monitor end effect, all wedge and tray factors, and bolus and comp filter attenuation factors should be verified.
In addition, various mechanical alignments should be checked annually. For example, the mechanical isocenter can be checked by observing the position of the front pointer tip in relation to a 2-mm-diameter rod as the gantry is rotated through 360q. A “star pattern” is sometimes produced to check radiation isocenter, i.e., a film is placed par- allel to the radiation beam and one set of collima- tor jaws is closed to a narrow slit and exposures are made at different gantry angles. All couch move- ments and table-top sag underload should also be evaluated.
Continuing education lectures on the machine operation, safety, and QA should be presented to the staff on an annual basis. It is important that emer- gency procedures be reviewed periodically with the staff to ensure proper interpretation and under- standing. A thorough hands-on training period for all therapists is essential following instruction about the operation of the equipment and prior to assum- ing treatment responsibilities. Written instructions should be provided to guide therapists as to a safe response when equipment malfunctions or exhibits unexpected behavior, or after any component has been changed or readjusted.
With a good QA and preventive maintenance program in which the parameters are measured and adjusted on a regular basis as outlined herein, treat- ment machines can be kept running in good operat- ing condition. These QA checks and adjustments are generally simple to learn and easy to implement.
16.5
Linac Advanced-Technology QA
Exciting technical developments for improving dose delivery have occurred over the past two decades as a result of the development of 3D con- formal radiation therapy (3D CRT) and intensity- modulated radiation therapy (IMRT; Purdy et al.
2001). These developments are based on advanced computer hardware and software technology and include 3D treatment planning systems (3D TPS), computer-controlled treatment machines, asym- metric collimators, dynamic wedge, multileaf col- limators (MLC), beam-intensity modulation, imag- ing devices (MV electronic portal imaging devices (EPIDs), and mV and kV cone-beam CT) for treat-
Table 16.7 Recommended annual QA checks Emergency off switches and interlocks
Mechanical and digital indicators (gantry, collimator, field size, couch)
Inspection of mechanical parts of accelerator including blocking tray and treatment aids
Machine alignment (isocenter check) Light-radiation field congruence
Radiation beam symmetry for all treatment modalities Monitor chamber linearity and end effect
Dose calibration (cGy/monitor unit) for all treatment modalities
Output field size dependence for all treatment modalities Percent depth doses for several field sizes for all treatment modalities
Wedge factors for all treatment modalities Tray factors for all treatment modalities Off-axis factors for all treatment modalities MLC checks
Special procedure modes (Arc therapy, TBI, TSEI, IMRT) Onboard mechanical and imaging checks
ment verification. Peripheral to the clinic are other daily localization systems. These advanced tech- nologies provide for radiation therapy techniques that will likely improve therapeutic ratios through the use of conformal physical dose distributions that cannot be achieved using 2D planning, delivery, and verification methods; however, ensuring that the radiation therapy process is safe and accurate when these advanced technologies are used is much more difficult than ensuring the 2D process that uses a treatment machine with simple electromechanical controls, 2D planning systems, standard treatment- aid devices (alloy blocks, physical wedges, etc.), and weekly port-film treatment verification. In many cases, the complexity of interactions between hard- ware and software in a near real-time environment makes it virtually impossible to demonstrate with certainty that the operation of the advanced technol- ogy systems is correct and that all possible failure modes have been eliminated. Exhaustive testing of all possible combinations of inputs, in all possible sequences, and from all possible sources, cannot be realistically accomplished; therefore, it is essential that a well-planned and rigorous approach to QA tests and safety procedures be practiced when these advanced technologies are implemented in the clinic (Klein et al. 1996).
16.5.1
Linac Computer Control System
Modern medical linear accelerators utilize computer control systems. Incidents in the past have shown that such accelerators have the potential for massive overdoses to the patient as a result of software flaws (Joyce 1986; Karzmark 1987). This poses a major problem for the radiation therapy community since standard QA tests on accelerators are not designed to catch software flaws. We caution the physicist to scrutinize carefully the computer operation of the linac during the acceptance testing period, paying particular attention to verifying what happens when beam setup parameters are edited.
An AAPM task group report discuss the safety considerations stemming from the increased use of computer logic and microprocessors in the con- trol systems of treatment units (Purdy et al. 1993).
It suggests how procedures and operator responses can be improved to reduce or obviate risks associated with hardware and software failures in radiation therapy equipment. Two other publications address testing and QA of computer-controlled accelerators
(Weinhous et al. 1990; Rosen and Purdy 1992).
Recommendations from these three publications are summarized in this section.
Acceptance testing procedures for new software updates and/or new computer-control features should be designed specifically to test the software and control aspects of the system. Safety interlocks and new functionality should be tested rigorously after review of all vendor documentation and test- ing information provided by the vendor. The reader should also note that while it is much easier to test safety interlocks in service mode, such tests do not necessarily properly predict the accelerator’s behav- ior in clinical modes; therefore, to ensure safe opera- tion, it is important that interlocks be tested in the clinical modes used to treat patients.
Routine updates of software for a computer-con- trolled machine should be treated as if it includes the possibility of major changes in system operation. All vendor information supplied with the update should be studied carefully, and a detailed software/control system test plan should be created. All safety inter- locks and dosimetry features should be carefully tested, regardless of the scope of the changes implied by the update documentation. All tests suggested by the manufacturer to confirm correct operation of the new software should be performed. Treatment beam parameters which may be affected by the soft- ware changes should be verified. Near full-accep- tance testing may be necessary depending on the nature and extent of the software changes.
Safety interlocks also may have to be tested follow- ing non-trivial repairs. Because software and hard- ware are intimately linked in a computer-controlled machine, even minor changes in hardware can pro- duce aberrations in the operation of the machine if there is a flaw in the software design or implementa- tion. Integrity of software and data should be veri- fied using appropriate tools supplied by the manu- facturer. If repairs are extensive or involve critical components, near full-acceptance testing may again be necessary to ensure proper operation.
Computer-assisted setup features should be verified. If possible, return of the machine to a safe condition in the event of a computer or computer- related hardware failure should be verified. If power conditioning and isolation for the computer is not used, the computer and machine operation should be carefully monitored for any adverse effects of occasional power transients.
Routine scheduled maintenance and testing
should be performed to minimize hardware mal-
functions that can occur over time due to normal
wear of components and environmental stresses such as radiation damage. In computer-controlled machines, hardware changes may also affect correct software operation by corrupting essential data or software; therefore, even though the software may have passed acceptance testing without demonstra- ble errors, latent bugs may appear as the hardware changes with age. Even minor changes in hardware can produce aberrations in the operation of the machine if there is a flaw in the software design or implementation; therefore, constant vigilance is necessary. In particular, safety interlocks may have to be tested following non-trivial repairs. If repairs are extensive or involve critical components, full acceptance testing may again be necessary to ensure proper operation.
In the case of software updates, the integrity of all safety interlocks and the software and database should be verified following installation. Treatment beam parameters that may be affected by software changes should be verified. Full-acceptance test- ing may be necessary depending on the nature and extent of the software changes. To assist users in properly verifying new versions of software, the documentation for software updates should include:
(a) reasons for all changes, including bug fixes; (b) details of modifications made; (c) details of planned or expected operational changes following installa- tion of the update; (d) effects on site-dependent and user-accessible data and/or software; (e) suggested procedures for testing operations affected by the update; (f) revised design specifications, support documentation, and/or operations manuals; and (g) results of beta tests.
Despite extensive in-house and field testing by the manufacturer, new problems are occasionally discovered by users in the field. Manufacturers should provide procedures for reporting such prob- lems. These procedures should clearly describe the information to be submitted with the report. Manu- facturers should respond to problem reports with a written acknowledgment, followed by a timely response evaluating the severity of the problem, a recommended temporary solution or a recommen- dation to suspend treatments, and a proposed per- manent solution with time schedule for implemen- tation. Dissemination of significant problem reports to all users should be done in a timely manner.
16.5.2
Asymmetric Jaws (Independent Collimation) All modern medical accelerators have collimator jaws that move independently. For example, the Varian linear accelerators, the Y-Jaws (upper), can be move independently 10 cm beyond isocenter, whereas the X-jaws (lower) can move independently 2 cm past isocenter. Independent jaw capability allows the isocenter to be positioned at locations other than the treatment field center. This flexibility allows simplified patient positioning and improved safety by avoiding overlapping field abutments without the necessity of using heavy beam-splitting blocks. For example, breast irradiation techniques commonly use the independent jaw feature (Klein et al. 1994).
Monitor unit calculations are only slightly more complex for independent jaws than for symmetric jaws (Slessinger et al. 1993). An off-axis correction factor can be used that depends only on the distance from the machine’s central axis to the center of the independently collimated open field. The influence of backscattered photons into the monitor chamber may influence the output for very elongated sym- metric fields (Palta et al. 1988). The dose distribu- tions of asymmetric fields defined by jaws are quite similar to those defined by alloy blocks. Only slight differences are seen at the field edges due to the dis- tance between the patient and the jaws.
Treatment record forms that denote each jaw setting (Y
1, Y
2, X
1, X
2) should be used. We also rec- ommend that the four jaws be identified by labels placed on the treatment machine, simulator colli- mator, and the block trays. This is especially useful when treatment techniques call for collimator rota- tion or involve the use of MLC or dynamic wedge, which are oriented in a particular direction along an independent jaw set (i.e., MLC - X
1or X
2, dynamic wedge Y
1or Y
2).
The QA checks include a monthly check of each
independent jaw by comparing jaw setting vs the
light field position vs 50% radiation value (edge)
for fields designed as quadrants (two non-divergent
edges). We recommend specifying the jaw position
accuracy to an accuracy of 1 mm for setting vs light
field at all positions (note that this exceeds TG 40
recommendations) and a specification of 0.5 mm
for light vs radiation field. An effective QA test is to
irradiate a film superimposing each quadrant of the
field separately. Ideally, the composite film should
exhibit no distinct regions of overlapped regions or
gaps. On a monthly basis, the simulator’s asymmet-
ric jaw (wire) settings accuracy should be checked for the 0.0- and 10.0-cm settings. An accuracy of 1 mm is expected, which is the same criteria used for symmetric fields.
Annually, the same quadrant test should be performed at the four cardinal gantry angles. An isodensity scan of the superimposed film will show dose homogeneity across all intersections. A dose inhomogeneity of d r5% (when compared with a point away from the junction) over a distance of d2 mm is acceptable. Annually, OAFs (used for asymmetric jaw monitor unit calculations) should be spot checked to ensure agreement within 0.5%
of the tabulated OAFs. The output at d
maxoff-axis when the Y-jaw is 10.0 cm beyond isocenter should be checked annually to ensure that backscattering to the monitor chamber that can cause an eventual decrease in output is not occurring. This measure- ment (at the respective d
maxdepths) is compared with calculations using the appropriate OAF. Annu- ally, corresponding quadrant films are also taken on the simulator.
16.5.3
Dynamic Wedge
Dynamic wedge technology takes advantage of asymmetric jaw technology in conjunction with control of the dose rate over the course of one treat- ment (Leavitt et al. 1990). An initial field is set along with a desired isodose angle (wedge angle) with a particular wedge direction (heel to toe). After a specific number of monitor units have been deliv- ered, the designated collimator jaw begins to move with a varying speed while the dose rate is varied simultaneously. This type technology as generally replaced physical wedges. The authors’ experience is based primarily on the Varian system and will be used for discussion purposes, but the recommended testing should be applicable to other manufactur- ers dynamic wedge systems. The variations in jaw position and dose rate are driven by computer files called segmented treatment tables (STTs), which are unique for each energy, wedge angle, and field size.
This customization of each dynamic wedge angle for each field size yields excellent-wedged isodose distributions when compared with physical wedges.
The obvious practical advantages of dynamic wedg- ing include no lifting of heavy wedges over patients, no blocking of light field during setup, and larger wedge field sizes (up to 30 cm field for 60° wedges).
Dosimetric advantages include no beam harden-
ing, less scatter outside of the field, slightly shorter treatment times, lower-intensity hot spots (in most cases), no wedge tray “play,” and improved wedged isodoses for all field sizes.
The QA program appropriate for dynamic wedge technology have been reported in the litera- ture and are summarized here (Klein et al. 1995, 1998). Periodic checks on the accelerators should include spot checking the dynamic wedge dosim- etry. These checks include verification of wedge fac- tors for selected field sizes and wedge angles. The wedge factors should be checked for each wedge orientation (Y1-in and Y2-out). Spot checks of the isodoses (or profiles) should be performed to ensure that the dose distributions have not changed.
The ability of the dynamic wedge to be completed during interruptions should be tested by terminat- ing and restarting the beam during a dynamic wedge run. Patient-specific checks include a diode check on the patient during the first fraction. A dual- diode system works best, particularly for wedge treatments, with one placed at the central axis on the patient’s skin surface and the other at an off- axis point toward an anatomically noted direction that corresponds to the “heel or toe” of the wedge.
The diode electrometer reading should be corrected by factors that depend on SSD, field size, and diode response for that day. The corrected reading corre- sponds to the dose at d
max. Deviations of 5% or less are considered acceptable due to the diode’s system- atic limitations in terms of spatial resolution and placement by the therapists. A second reading and investigation should be performed for larger devia- tions. Therapists should also be instructed to illumi- nate the light field at the end of each fraction to con- firm that the remaining light field strip corresponds to the “toe” of the wedge. An R&V system provides a direct check of the dynamic wedge angle and ori- entation and a visual check need only be confirmed during the first fraction; however, there should be an independent check confirming the R&V entry of the wedge angle and direction before the first frac- tion. On the first fraction, the therapist still should check the indicated position and light field strip after treatment. The R&V system provides ongoing checks for the remaining fractions.
16.5.4
Multileaf Collimation
Multileaf collimators have now become the state-
of-the-art method for generating irregularly shaped
fields for photon-beam radiation therapy (Boyer et al. 1992, 2001; Klein et al. 1995). The authors’
QA experience is based on use of a Varian MLC system (Klein et al. 1995, 1996; Klein and Low 2001). In that system, the leaf settings for each field are sent to a dedicated MLC computer, interfaced to the treatment machine, which drives the leaves via a controller system. The leaf settings may be obtained by two different methods. Firstly, computer software and hardware (called the “Shaper”) is provided with which the user can digitize a portal shape drawn on a simulation film or using a 3D treatment-planning system. In either case, the patient’s MLC configura- tion files are sent over a local area network to the MLC computer.
Rigorous QA is essential when MLC is clinically implemented. When first implemented in a clinic, all MLC fields should be checked visually (light field vs skin marks) and imaged using film or an electronic portal imaging device (EPID) on a daily basis.
Specific QA tests performed include the daily run- ning of a sampling of actual clinical fields during the morning checkout of the accelerator. In addition, the following periodic (quarterly) checks are rec- ommended: (a) testing of MLC settings vs light field vs radiation field for selected gantry and collimator angles; (b) network testing; (c) check of active patient files; and (d) interlock checks (carriage under jaw, leaf spread, leaf movement during electron and/or port film modes, etc.). Particular attention must be paid to the testing of MLC files sent over a computer
network. When first implemented, it is prudent to test the network by illuminating every MLC field configuration onto the original simulation film or digitally reconstructed radiograph. The AAPM TG- 50 report on basic applications of MLCs describes the QA program of recommended patient and quar- terly and annual checks. The annual check of accel- erators equipped with MLC include the above-listed monthly MLC tests and also film scans to review interleaf leakage, abutted leaf transmission, pen- umbra dependence on leaf position, and a review of procedures with clinical staff. A summary of the QA checks recommended is given in Table 16.8.
16.5.5
Online Electronic Portal Imaging
Online portal imaging systems consist of a suit- able radiation detector, usually attached through a manual or semi-robotic arm to the linac and capable of transferring the detector information to a com- puter that will process it and convert it to an image.
These systems use a variety of detectors, all produc- ing computer-based images of varying degrees of quality. Currently these systems include: (a) fluoro- scopic detectors; (b) ionization chamber detectors;
and (c) amorphous silicon detectors. The QA issues for this technology can be separated into five cat- egories: (a) physical operation and safety; (b) spatial and contrast resolution; (c) image storage, analysis,
Table 16.8 Multileaf collimation QA. (From Klein et al. 1996 and AAPM TG-50 report). DRR digitally reconstructed radio- graphs
Frequency Test Tolerance
Patient specific Check of MLC-generated fields vs simulator film/DRR before each field is treated
2 mm
Double check of MLC field by therapists for each fraction Expected field Online imaging verification for patient on each fraction Physician’s discretion Port-film approval before second fraction Physician’s discretion Quarterly Setting vs light field vs radiation field for two designated patterns 1 mm
Testing of network system Expected fields over network
Check of interlocks All must be operational
Annual Setting vs light field vs radiation field for patterns over range of gantry and collimator angles
1 mm
Water scan of set patterns 50% radiation edge within 1 mm
Film scans to evaluate interleaf leakage and interleaf, abutted leaf transmission
Leakage <3%, abutted leakage <25%
Review of procedures and in-service with radiation therapists
