First published online September 27, 2006
British Journal of Radiology (2007) 80, 126-131
© 2007 British Institute of Radiology
doi: 10.1259/bjr/32367104
Design and implementation of an electronic data recording and processing system for physics quality control checks in external beam radiotherapy
I Patel, MSc, CSci, MIPEM
and
M C Kirby, PhD, CSci, MIPEM
North Western Medical Physics, Radiotherapy Department, Rosemere Cancer Centre, Royal Preston Hospital, Preston PR2 9HT, UK
Correspondence: Mr Imran Patel, North Western Medical Physics, Radiotherapy Department, Rosemere Cancer Centre, Royal Preston Hospital, Preston PR2 9HT, UK. E-mail: imran.patel{at}lthtr.nhs.uk
 |
Abstract
|
|---|
Quality control (QC) of external beam radiotherapy equipment ensures that commissioning performance is maintained. Paper based data recording is still used for QC, but this is resource intensive in terms of data calculation and processing. Electronic systems of data recording have many advantages; for example, they facilitate the analysis of data on a regular basis, allowing the user to examine the "health" of each machine; they help review and audit local frequencies of each check and can possibly predict component failure. They also allow for secure calculation of results and automatic charting for routine trend analysis. Initially, data recording at our centre was paper-based for daily, weekly and monthly checks. This paper system has been successfully replaced with an electronic system for QC data recording and processing for linear accelerators and superficial units. The system makes use of personal digital assistants and networked laptops for online recording of data, and networked desktop PCs for offline work. The systems of data recording have been designed using the power of macros within Microsoft Excel, which automatically calculate each QC parameter and charts the data recorded for long-term analysis of trends. As of the beginning of 2006, these systems have been fully implemented. The benefits of implementing such a system are numerous, for example, central storage, backup and archiving of data, for greater security and reducing operator errors in calculations. Other benefits are discussed within the paper. In the future, we hope to develop similar systems of data recording for QC checks on other radiotherapy equipment.
|
Introduction
|
|---|
Quality control (QC) of external beam radiotherapy equipment ensures that commissioning performance is maintained by measuring parameters which affect the geometric and dosimetric accuracy of the applied dose to the patient [1–3]. These parameters can change rapidly due to electronic malfunction, component failure or mechanical breakdown, or more slowly due to deterioration and ageing of components.
Many systems of work for QC still use paper records for recording results. This makes it resource intensive to analyse data trends on a regular basis [4] or extract statistics on the general long term "health" of a treatment machine. This type of analysis is vital for providing evidence for reviewing and justifying the local frequency of QC checks [5] and quantifying the possible consequences of those frequencies and any changes to them [6]. Having the ability to examine data over the short-, medium- and long-term may enable the user to predict possible problems and component failures, thereby leading to a reduction of unplanned downtime. This analysis is made more secure and readily achievable through electronic systems such as spreadsheets, workbooks and databases. Implementation has been facilitated by more accessible mobile computing, such as palmtops (Personal Digital Assistants [PDAs]), laptops and networked workstations.
|
How the Rosemere Cancer Centre systems have evolved
|
|---|
The Rosemere Cancer Centre (RCC) is relatively new, having treated its first patient in February 1997. Expansion since then has been rapid, with the complement of equipment increasing from a simulator, two linear accelorators (linacs) and a superficial X-ray therapy (SXT) unit in 1997, to two simulators, six linacs and an SXT unit in 2005. There are six electronic portal imaging devices (EPIDs), four multi-leaf collimator (MLCs), and Image/Data networks with a full QC programme for all equipment.
Our programme was initiated with check frequencies and tolerances based upon the available literature (e.g. [7]) and prior experience of the responsible physicists. Data recording was paper based for daily, weekly and monthly checks. The department quickly moved to Microsoft (MS) Excel spreadsheets for monthly checks on the linacs, but the data were not connected from month to month, i.e. it was still resource-intensive to review and chart data. The data were summarised annually in greater depth by entering daily, weekly and monthly check data into separate analysis spreadsheets so that overall performance could be examined. This paper focuses on the recent advances made by our department in improving the above systems for entering, calculating, storing, charting and routinely analysing QC data.
|
Methods and materials
|
|---|
Data entry – hardware
Central recording and storage of data was a key factor in designing our electronic systems of work. Desktops, laptops and palmtops were all considered and used as hardware solutions. For frequent data input (such as daily and weekly checks on linacs), part of the system of work was designed using Axim X5 palmtops (Dell, Berkshire, UK). The palmtops operated under the MS Pocket PC 2002 operating system with MS Pocket Excel installed. Data could be synchronised between desktop PC and palmtop using MS ActiveSync software and connecting the palmtop to the PC with a synchronisation cable to the PC's USB port. The palmtops had the advantage of being highly portable, thereby allowing the user to take them into the treatment room for, say, recording geometry check results. For the initial trials, two palmtops were used for the five linacs which were in clinical use at that time.
For less frequent checks (such as monthly/quarterly checks), where the data entry was also greater and more complex, laptops and desktops were used, connected through convenient network points to the hospital's central servers, which are backed up daily.
Data entry – software
For the palmtops, MS pocket Excel provided fundamental spreadsheet tools, such as formulae and functions. However, it did not allow the use of macros or charts. As a result, the software was designed in two parts. First, a workbook was created with a simple user interface for palmtop use to input the results of the daily/weekly checks online and perform the appropriate calculations. Second, a spreadsheet was designed (stored on the central server) that would synchronise with the data on the palmtop, transfer the data using macros, add it to previous data and chart the results automatically. The data recorded directly onto the palmtops included: (1) dosemeter readings; (2) data for energy and symmetry checks; (3) geometric test results (such as crosswire rotation, light field data etc.); and (4) beam parameters (such as gun and steering currents). Data entry was designed for both X-ray and electron checks, and all outputs and energy ratios were calculated within the sheets. Calibration factors required for calculating output are input by the user. These factors are found on data charts controlled within our department's quality management system. This approach was chosen rather than an automated selection process within the spreadsheets as our current system of updating and issuing calibration factors (and indeed other QC reference parameters) follows a clear, well-established and robust method of change control.
For monthly checks, spreadsheets were designed to allow data entry "cumulatively", i.e. divided into columns for each month so that automatic charting would be straightforward. This enabled online entry of data using networked laptops or offline data entry using networked desktops. Calculations were performed automatically on the data, with the added advantage of having the previous month's data in the same spreadsheet for comparison. Data recorded included checks for output, energy and wedge factors, radiation field symmetry, X-ray and light field sizes (including MLC leaves) and EPID resolution and contrast.
One of the main reasons why MS Excel was chosen for the system as opposed to any other software package was because of our familiarity with the functions within the software. This meant that less time was required in designing the spreadsheets. Training was given in the form of demonstrations and one-to-one tutorials in the use of the new systems to all staff members (physicists and engineers) who were involved in the daily, weekly and monthly QC of the linacs. Practical trials were conducted to refine the systems over a period of a few months.
|
Results and discussion
|
|---|
Systems for frequent (daily/weekly) checks
The software for the palmtops (Figure 1
) was easy to design and found to be secure and robust. As anticipated, they were lightweight and simple to use inside and outside the treatment room. Calculations of parameters (such as output, energy ratios etc.) were automatic, reducing the risk of operator error. 2 weeks of daily/weekly check data was stored on the palmtop at any one time so that the user could scroll through it looking for daily changes. Synchronization with the spreadsheets stored on the central server (which is backed up daily) was possible without loss of data. Data could be synchronized on a daily or weekly basis, with an automatic update of the charted data, such as output, beam parameters, light field readings and energy ratios etc. The recording of daily parameters, such as gun current, which is then continuously analysed with the aid of the spreadsheets, can help to pre-empt machine problems. For example, an imminent gun failure could be predicted by looking at the rate of change of gun current over time. This has an obvious advantage of making the component change in a planned, scheduled manner. Similarly, a potential failure of the monitor ion chamber (related to excessive leakage) might be pre-empted in the same manner. Approximately 50% of staff found the palmtops easy enough to use to input data directly whilst performing the checks, without any significant increase in the time taken.
View larger version (87K):
[in this window]
[in a new window]
|
Figure 1. The Dell AXIM X5 palmtop and an example of one of the data input screens(for 6 MV dosimetric daily checks).
|
|
Conversely, about 50% of the staff found the palmtops not so easy to use. The main observations from these staff were:
- The small palmtop screen was sometimes difficult to read and limited in its display of information (Figure 1
)
- Data entry using the stylus and software keyboard or script language was difficult
- Scrolling to different parts of the screen to enter data or compare daily parameters was sometimes awkward
As stated earlier, the pocket version of Excel did not support all the software functionality of the desktop version. So, for example, cells cannot be locked, charts cannot be used in the pocket version, etc. Additionally, some difficulties were encountered with connectivity permissions for individual user accounts on the hospital's networked PCs when trying to synchronize the data. Allowing only one login user account on the hospital network to synchronize the palmtop data with PC data solved this problem. Difficulties were also encountered with having palmtops shared across the linacs; data had to be kept up to date on both devices for all the linacs, especially if no restriction was placed on which device was used on which linac. However, this problem would have been easily resolved if there were dedicated palmtops for each linac.
As a result of the practical trial, the system for entering data electronically for daily/weekly checks has now been modified for our department. It is now more flexible in allowing a wide range of hardware options for data input. For instance, staff may still choose to use the palmtops, synchronizing data daily or weekly. Alternatively, networked laptops can also be used for online data entry and networked desktops for offline data entry. The data can be entered on a daily or weekly basis without any consequence to data handling and transfer. The concept of data entry workbooks separated from workbooks for the cumulative results is still maintained. Simple buttons linked to macro commands are used to format the data entry pages, transfer the data when complete and then reset the workbook ready for the following week's work.
Systems for less frequent (monthly or 3 monthly) checks
The workbooks designed for less frequent checks were also found to be robust and secure. All hardware options worked well (networked laptops and desktops). Calculations and charts were again performed automatically, minimizing the consequences of operator error; results could be easily compared with baseline data and trends observed. However, many staff found the format of data entry (columnar) awkward to use online. This was resolved by designing a two-part system similar to that for the daily/weekly checks, with separate workbooks for data entry and data accumulation/charting. For data entry, a format was designed similar to that used previously, where a workbook contains all the monthly check data for a single machine and for a single month. The columnar form was retained for the cumulative workbook, but buttons were added which were linked to macro commands so that data was handled automatically.
Figure 2a shows an example of one of the data entry sheets for linac QC checks; this one shows part of the page for the dosimetric checks for 6 MV X-rays on a multimodality linac. The tabs for other worksheets are visible for entry of electron, EPID and non-dosimetric data. Once all the check data have been recorded (usually online as the checks are being conducted), the data can be formatted and printed for hardcopy records (which can be archived), transferred to the cumulative workbook for that linac, and then the worksheets reset ready for the next month's data entry. The latter sequence takes place using the simple, clear and user-friendly front worksheet (shown in Figure 2b).
View larger version (68K):
[in this window]
[in a new window]
|
Figure 2. Two worksheets from the data entry workbook for monthly linac QC checks.(a) Part of the worksheet for 6 MV dosimetric checks. (b) The front worksheet of the workbook, designed for printing and transferring data and resetting the workbook.
|
|
The buttons shown are linked to macros programmed for each of the tasks described above. They have been written with data security in mind. For example, the user cannot inadvertently overwrite the previous month's data in the cumulative data workbook during the transfer process. Once transferred, data cannot be altered or deleted since the macros automatically protect the workbook. If data are transferred inadvertently or the wrong data were entered, the workbook can be unlocked by the user and the data transferred again.
A worksheet from the cumulative data workbook for linac monthly QC is shown in Figure 3a. It shows how data have been reformatted from the data entry workbook and appended into each particular worksheet for that machine, month by month. The worksheets are programmed with pre-defined tolerances for key parameters, so that a figure will be highlighted red if beyond that tolerance.
View larger version (51K):
[in this window]
[in a new window]
|
Figure 3. Worksheets from cumulative data workbook for monthly linear accelorator(linac) QC checks. (a) Part of the worksheet for cumulative data for 6 MV dosimetric checks. (b) The chart for percentage change in output updated automatically from the accumulated data.
|
|
Although it is clear that the data can be easily reviewed, the data are also automatically charted so that trends become obvious, as shown in Figure 3b. Charts are used for parameters such as output, energy ratio, wedge factor, EPID spatial resolution and EPID contrast-to-noise ratio.
Practical trials with the monthly check workbooks have proved highly successful. The time taken to report on monthly checks has reduced dramatically since the data are automatically formatted and printed ready for signing, once complete. As a result of their success, new workbooks have also been designed with the same two step process for monthly checks on the superficial therapy unit and no difficulties were found while trialling these.
As of the beginning of 2006, these systems have been released for full use as part of the daily, weekly and monthly checks on the linacs and superficial therapy unit. As the data accrue, they will provide an invaluable source of information. The data can easily be reviewed on a much more regular basis, which could show changes in machine performance and possibly component deterioration. This information might then be used to carry out planned detailed checks, services, preventative maintenance or component changes before any component failure, thereby reducing the probability of unplanned downtime and therefore delays to patient treatments. It also makes the task of accruing local evidence for reviewing and possibly changing frequencies of checks much easier. Our examples show some of the QC checks used at RCC and their frequencies; but our system of work can be applied to any series of QC checks and their specific frequencies, e.g. those suggested by IPEM Report No. 81 [5].
The individual undertaking the checks is clearly identifiable under this system of work. This would act, for example, as a record of the operator in the context of IRMER [8]. Hardcopies of the QC results are signed and dated by the appropriate individual and soft-copies (spreadsheets) have cells for the date and operators' initials. During audit, it is easy to identify operators and verify (by consulting dated authorisation records) that the individual was permitted, at the time, to perform the specific QC checks.
Table 1 compares the advantages and disadvantages of paper based and electronic data systems. The following are additional advantages of using electronic systems of work:
- Maintaining a flexible approach to data recording and analysis inkeeping with staff requirements
- The possibilities of programmed scheduling for checks which vary from weekly to monthly to 3 monthly etc.
- Automatic highlighting of results which are out of tolerance
- Central access, storage, retrieval, backup and archiving of data for greater security
- Formatted hard-copy can still be easily generated for signing purposes
- Data that can be more readily shared with manufacturers and compared with other users
In the future we hope to develop similar systems of data recording and processing for routine checks on dosemeters, simulators and CT scanners.
|
Conclusion
|
|---|
A robust and easy to use electronic system of work has been successfully designed and implemented for recording, processing and storing data for daily, weekly and monthly physics QC checks on linear accelerators and superficial therapy units. The system makes use of palmtops and networked laptops for online recording of data, but data can also be entered offline with the use of networked desktop PCs. The systems of data recording have been designed using Microsoft Excel, which automatically charts the data for long-term analysis of trends. This can lead to a possible reduction of machine downtime (by monitoring parameters and results), and also provides solid evidence for reviewing and/or changing frequencies of QC checks.
Received for publication May 10, 2006.
Revision received July 7, 2006.
Accepted for publication August 15, 2006.
|
References
|
|---|
- World Health Organization. Quality Assurance in Radiotherapy. Geneva, Switzerland: WHO, 1988
- Nath R, Biggs PJ, Bova FJ, Ling CC, Purdy JA, van de Geijn J, et al. AAPM code of practice for radiotherapy accelerators: Report of AAPM radiation therapy task group no. 45. Med Phys 1994;21:1093–121.[CrossRef][Medline]
- Thwaites DI, Scalliet P, Leer JW, Overgaard J. Quality assurance in radiotherapy. Radiother Oncol 1995;35:61–73.[CrossRef][Medline]
- McKay DM, Bulmer N, Fillingham E, Glendinning AG, Harris F, Kirby MC, et al. Quality control of Elekta SL and RTD linacs – an analysis of 5 year results. In: IPEM Safety, QC and Clinical Governance in Radiotherapy Meeting, 2002 November 4; London, UK
- IPEM (Institute of Physics and Engineering in Medicine). Physics aspects of quality control in radiotherapy. IPEM Report 81. York, UK: IPEM, 1999
- IPEM (Institute of Physics and Engineering in Medicine). Balancing costs and benefits of checking in radiotherapy. IPEM Report 93. York, UK: IPEM, 2006
- IPEM (Institute of Physics and Engineering in Medicine). Commissioning and quality assurance of linear accelerators. IPEM Report 54. York, UK: IPEM, 1990
- IR(ME)R 2000 (The Ionising Radiation (Medical Exposure) Regulations 2000). Statutory Instrument No. 1059. London: HMSO, 2000
Top
Abstract
Introduction
