BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
British Journal of Radiology (2006) 79, S50-S65
© 2006 British Institute of Radiology
doi: 10.1259/bjr/21517185
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kirby, M C
Right arrow Articles by Glendinning, A G
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kirby, M C
Right arrow Articles by Glendinning, A G

Full paper

Developments in electronic portal imaging systems

M C Kirby, PhD, CSci, MIPEM and A G Glendinning, PhD, MIPEM

North Western Medical Physics, Radiotherapy Department, Rosemere Cancer Centre, Royal Preston Hospital, Preston, UK


   Abstract
Top
 Abstract
Introduction
 Technical developments
 Clinical developments
 Technical innovations - the...
 Conclusions
 References
 
Verification of geometric accuracy at the time of treatment delivery has always been a necessary part of the radiotherapy process. Since the introduction of conformal and intensity-modulated radiotherapy, the consequences of patient positioning errors are more serious. Portal imaging has played a large part in fulfilling the need for improved geometric accuracy. This review examines how portal imaging has progressed through the development and evolution of electronic portal imaging devices (EPIDs). Changes in technology, including the current commercial systems, and how image quality has changed are presented. The clinical usage of EPIDs and the technological innovations being devised for further improvements in image quality and systems are considered.


   Introduction
Top
 Abstract
 Introduction
Technical developments
 Clinical developments
 Technical innovations - the...
 Conclusions
 References
 
Technological developments in radiotherapy have been driven by the fundamental requirement to maximize dose to the intended target volume whilst minimizing dose to healthy surrounding tissue. Advances in collimation have enabled accurate, precise and more reproducible beam shaping and placement. Continuing accelerator and treatment planning system (TPS) development has further improved dose conformality around the target volume through intensity-modulated radiation therapy (IMRT). In addition, improvements in diagnostic imaging and their combination through different modalities are showing clinicians more precisely the anatomical areas that must define the high dose volume.

However, such efforts can be made ineffective by poor patient immobilization, set-up and verification at the time of treatment delivery [1]. The "chain" of the radiotherapy process must include a link for treatment verification. The overall effectiveness of radiotherapy is dependent upon all links in the chain.

Portal imaging, or taking images at the time of treatment delivery, is a key element of that verification. Its main use has been for geometric verification of field placement, but its dosimetric applications are also being developed. However, verifying the position of a soft tissue target volume with megavoltage energy X-ray beams is not trivial. Bony landmarks have to act as a surrogate for the target volume because, at megavoltage energies, the inherent subject contrast is naturally poor [2]. Soft tissue markers can be used for certain tumour sites with the advent of more advanced imaging detectors [3, 4]. Kilovoltage imaging could be used to improve the contrast of soft tissue and bony detail, and is the subject of current developments in image-guided radiation therapy (IGRT) [5]. See also the paper by Moore et al in this issue.

This review paper will examine the developments in portal imaging systems over the last 30 years, and is divided into three main areas: (1) how the technology and methods have evolved; (2) clinical developments over that time; and (3) technical innovations being researched for future generation of EPIDs.


   Technical developments
Top
 Abstract
 Introduction
 Technical developments
Clinical developments
 Technical innovations - the...
 Conclusions
 References
 
Portal film
Radiographic film was the traditional medium for acquiring images for treatment verification and is still used alongside EPIDs. Early images were acquired with industrial, direct exposure films, usually in cardboard holders, and required special processing techniques [6, 7]. In 1974, Kodak introduced film in "ready-packs", which could be processed in automatic, rapid processors [7].

In conventional film for high energy X-rays, the media is sandwiched between metal plates (usually Cu about 1 mm thick) [6–9] or the rear plate may be made of a thermoplastic [7, 10]. The front metal plate acts as a build-up layer producing high-energy electrons that in turn expose the film. It will also filter incident scattered radiation. The rear plate produces back-scattered electrons and helps maintain a tightly packed cassette.

For the more recent, advanced film systems, additional phosphors are used either side of the film. For example, Kodak have a commercial film, known as enhanced contrast localization (EC-L), which provides much improved image quality through the use of a fine grain, high gamma film combined with two phosphor screens and a 1 mm Cu front metal screen [7, 11]. Electrons produced by the front plate interact with both phosphor screens, producing optical photons that expose the film. The detection efficiency of the X-rays is increased by a factor of 2 [10]. The total noise within the system is reduced due to the smaller size and distribution of the very fine grain microcubic crystals. The film also has a higher gamma (around 6) [11] giving higher display contrast – ideal for the inherent low contrast of objects imaged at megavoltage X-ray energies [2, 10, 12].

Advantages include high quality imaging, using simple, relatively lightweight cassettes, allowing for single/multiple exposures and with good, well-defined dosimetric properties. Disadvantages include the required processing, a fixed dynamic range and digitization required to enhance or manipulate the images. This makes film impractical for on-line imaging. They are also consumables and require significant storage solutions.

Computed radiography systems (photostimulable phosphors)
Computed radiography (CR) systems were initially developed for diagnostic imaging [13] and used for radiotherapy from the 1980s [1, 6, 14, 15]. Film is replaced by a flexible plate, about 1 mm thick, coated with europium-activated fluorohalide compounds in crystalline formation embedded in an organic binding material. The photostimulable phosphor acts as an energy trap when exposed to ionizing radiation, producing a latent image. When scanned with a red (633 nm) laser, the energy is released as visible light, which is then converted to electrical signals via a photomultiplier tube [1, 14]. Commercial systems are available from Kodak (Figure 1Go), Agfa (www.agfa.com/healthcare) and Fuji (www.fujimed.com).


Figure 1
View larger version (89K):
[in this window]
[in a new window]
  		Figure 1. The Kodak CR System for radiotherapy verification.(Images from Kodak website www.kodak.com/go/oncology).

 
These systems are used for diagnostic and therapy applications (e.g. by reversing the imaging cassette). Image quality is good and compares well with that from modern EPIDs (Figure 2Go).


Figure 2
View larger version (76K):
[in this window]
[in a new window]
  		Figure 2. Simple comparison of images acquired of an anthropomorphic type phantom using(a) an amorphous silicon (a-Si) imaging system (Elekta iViewGT) and (b) a CR system (Kodak 2000RT). Both images acquired using 6 MV X-rays. The a-Si image was acquired with 100 MU and the CR with 20 MU. (CR image courtesy of Steve Weston, Cookridge Hospital, Leeds, UK).

 
They have very similar advantages and disadvantages to those of film. However, they are reusable and the readout mechanism (effectively equivalent to film processing) produces immediate digital images in multiple formats.

First generation EPIDs
Electronic systems for portal imaging were first proposed in the late 1950s. Reviews by Boyer, Langmack and Antonuk chart their progress very well [2, 6, 10]. A variety of technologies have emerged over the years which may be classed broadly into (i) camera based systems, (ii) ionization chamber matrix devices and (iii) scanning array and other systems. Devices in classes (i) and (ii) were successfully commercialized.

The motivation for electronic devices was simple: to obtain high quality images rapidly, with the versatility and flexibility afforded by computer digitization. With this premise, systems were designed to acquire quick digital images, making on-line enhancement, analysis and set-up correction a real possibility.

Camera based EPIDs
The first electronic systems developed were camera based systems [2, 6, 10], three of which became commercially viable products for manufacturers Siemens ("Beamview") [16], Infimed ("Theraview") [17] and Philips/Elekta (SRI-100/"iView") [18]. Two systems are shown in Figure 3Go.


Figure 3
View larger version (75K):
[in this window]
[in a new window]
  		Figure 3. Two commercial camera-based electronic portal imaging devices (EPIDs). (a) The Elekta iView system attached to the gantry of an SL Linac. (b) The Infimed (Cablon) Theraview system in its retracted position. Also shown in (a) is a typical portal film cassette (Kodak EC-L system) in its portal film holder.

 
The principle of detection is similar to that described previously for film. The X-ray conversion stage is a metal plate (typically about 1–2 mm copper, steel or brass) bonded to a rare-earth phosphor (usually gadolinium oxysulphide) of various thickness (100–400 mg cm–1). X-rays incident on the metal plate produce high-energy electrons. The metal plate also filters some of the low-energy scattered radiation (X-rays and electrons), which would otherwise reduce the contrast of an image [2]. The phosphor converts the high-energy electron energy (and also that of residual high-energy X-rays) into light. The light diffusing through the phosphor on the camera side is reflected off a series of mirrors into a lens and camera. This enables the camera, and its radiation sensitive components, to be placed outside the primary X-ray beam, usually behind lead shielding [18, 19]. The lens and camera capture only a very small proportion of the light, which is one of the inefficiencies of such devices. The camera converts the image into a video signal that is fed into computer hardware for digitization, processing, enhancement and display.

The spatial frequency dependent detective quantum efficiency (DQE(f)) is a widely used index of X-ray imaging performance [2, 10, 20]. It characterizes the efficiency of information transfer from the input to the output stages of a system. It can also be expressed as the ratio of the squared output signal to noise ratio (SNR) to the squared input SNR. The detectability of an object is dependent upon SNR [2, 12]. DQE depends upon the spatial resolution, gain and noise transfer properties of the system [2, 10, 20]. It is maximized by increasing the efficiency of detection of incident radiation, improving spatial resolution and decreasing system noise.

For camera-based systems, the noise components are numerous [21], but the major inefficiency is light output and collection. Increasing phosphor thickness improves light output, but at the cost of spatial resolution [17, 22]. Generally, only about 10–20% of the light emitted by the phosphor escapes. More important limitations are the poor light collection efficiency and the electronic noise within the camera chain [21, 22]. Light is emitted from the phosphor isotropically but only photons emitted within a small cone subtended by the camera lens generate a signal in the camera. Between 0.1% and 0.01% of the light emitted reaches the camera [10]. If this light subsequently produces a small signal, then it may be swamped by noise generated in the camera and associated frame processing system [21, 22]. Attempts to improve light collection have included modifying the phosphor screen [23, 24] and using large aperture lenses. However, the latter introduce image distortions and aberrations, among other effects, which can reduce spatial resolution and cause non-uniformities across the image [22]. Camera and electronic noise can be reduced by different integration strategies and using cooled, low noise target chips on CCD type cameras [10]. The maximum DQE achievable has been about 1%.

The main advantages of camera-based systems are that the whole image is viewed simultaneously, imaging is very fast (video rate), there is good spatial resolution and the system is relatively cheap to service and maintain. A practical downside is their "bulkiness" when in use, making it difficult to work around them for patient set-up.

Liquid ionization chamber matrix EPID
This device was designed and first used at the Netherlands Cancer Institute in the 1980s and commercialized rapidly by Varian Medical Systems in the early 1990s [25, 26]. Its operational design and characteristics are well reviewed in the literature [2, 6, 10, 22].

Briefly, the ionization chamber is formed by two planes of electrodes, separated by a 0.8 mm gap. The gap is filled with an organic fluid (iso-octane or 2,2,4-trimethylpentane), which acts as the ionization medium during exposure. Ions produced in the liquid are collected by the electrodes, of which there are 256 in each plane spaced 1.27 mm apart. The planes are oriented orthogonally forming a 2562 matrix. Primary X-rays incident on the front surface of a 1 mm thick plastoferrite plate produce high-energy electrons, which (in addition to non-interacting X-rays) produce ionization within the liquid medium. High voltage is applied to each electrode individually, but multiplexing electronics are used so that the image is readout sequentially row upon row. The detector and peripheral electronics can be tightly compartmentalized so that the EPID can retract fully (under motor control) into the gantry.

A full resolution image is acquired and processed in just over 5 s; faster acquisition was initially made possible by reading pairs of electrodes, but at the cost of a lower spatial resolution. The most recent systems use a higher voltage and a shorter readout time per electrode, bringing full resolution image acquisition down to about 1.25 s [22].

For the commercial system, its greatest advantages were its compact, practical design making it easy to use, its lack of geometric distortion, but also its associated software [2, 22, 27]. Its main disadvantages were the sensitive nature of the control electronics surrounding the active detection area and, most significantly, the under-utilization of incident X-ray quanta compared with true area detector EPIDs. Whilst about 1.5% of the incident X-rays interact in the plastoferrite plate and liquid ionization medium, generating measurable signal [22], the DQE of the system is only about 0.5% due to signal loss in sampling [10]. This means that the dose for image formation is higher than true area detectors.

Other EPIDs used in the development of portal imaging
Two other designs of EPID have greatly helped to progress the clinical developments of electronic portal imaging [2, 10]. The first was a variation of the camera based EPID [28], which used a construction of fibre-optic bundles to "pipe" the light directly from phosphor to camera. This made the device less bulky and more practical than the standard camera-based construction. However, image distortion and non-uniformity was a problem because of minor irregularities in alignment and shape of the fibre bundles at the output stage, and variations in light transmission through the bundles [2, 28].

The second was a scanning system that used a linear array of scintillation crystals optically coupled to photodiode detectors [2, 10, 29]. Zinc tungstate, bismuth germanate and caesium iodide were all used as scintillators in variants of the device. High quality images were produced, but scanning times were relatively long, especially compared with video based EPIDs. However, many clinical studies were performed using the device, as well as dosimetric and megavoltage CT investigations [6, 10, 30, 31].

Active matrix flat-panel imagers (AMFPIs)
Initially conceived in 1987 by research scientists at the University of Michigan, Xerox PARC and elsewhere [10, 32–34], indirect detection active matrix flat panel imagers (AMFPIs) now form the main focus of developments in portal imaging. All manufacturers offer this technology and, because of their enhanced image quality and potential for dosimetric uses, they constitute an effective second generation of EPIDs. These devices became commercially available in 2000; two devices are shown in Figure 4Go.


Figure 4
View larger version (60K):
[in this window]
[in a new window]
  		Figure 4. Active matrix flat panel imagers(AMFPIs) from two of the major radiotherapy equipment manufacturers; (a) Elekta and (b) Siemens. Electronic portal imaging devices (EPIDs) of this type are highly practical devices, being fully motorized and retractable to the gantry of the linear accelerator when not required for use. The AMFPI from Varian uses similar gantry mounting technology.

 
AMFPI technology and process
The technical details for such EPIDs are covered in recent review papers [2, 6, 10, 34]. In brief, AMFPIs consist of a number of subsystems. These are (i) a large area, pixelated array, (ii) an X-ray converter, (iii) a control and acquisition system and (iv) a computer for controlling acquisition and processing, enhancing and archiving the images.

The array consists of a ~1 mm thick glass substrate upon which the electronic circuits reside. Each pixel in the active matrix-imaging array incorporates a thin-film switch connected to a sensor. The pixels are arranged into a two-dimensional grid with the conductivity of each switch regulated by varying the voltage of control lines, where each control line is connected to all the switches in a particular row. During irradiation, the switches are kept non-conducting so that radiation generates electrical signals integrated in the capacitative element of each pixel. At the time of readout, the switches are made conducting, one row at a time. At this point, the charge in each pixel is transported to external electronics by means of the data lines. Each data line is connected to all the pixels in a given column, so a line of data is read [10].

There are two general methods for the initial conversion of incident X-ray energy into charge stored in each of the capacitative elements of the AMFPI. These are Indirect or Direct detection methods [10, 34]. For Indirect detection, a metal plate/phosphor screen combination is used as the X-ray converter. High-energy X-rays and electrons produce light in the phosphor or scintillator, which is positioned directly over the photosensor of the array. Optical photons are converted into electron-hole pairs in the photosensors, which also act as the capacitative element in each pixel until readout takes place using the pixel switches [6, 10]. For Direct detection, the X-ray converter consists of a metal plate and a photoconductor. The photoconductor is electrically coupled to a separate capacitor built into each pixel. Electron-hole pairs are produced in the photoconductor, which are then stored in each capacitor before readout [10, 35, 36].

At present, all commercial AMFPIs use the indirect detection method [10]. Image acquisition can be fast (up to 10 frames per second), with dynamic ranges up to 16 bits deep and pixel resolutions up to 10242 in matrices up to 41 cmx41 cm at the detector [10, 37]. Readout is usually synchronised to occur between linac pulses. Image correction and processing is similar for all the commercial devices, consisting of the use of gain (flood field) and offset correction as well as a method of filtering for reducing or eliminating the influence of bad pixels or line defects [10, 38–40]. Being a flat panel design, there is no spatial distortion present in the resulting images [41].

Image quality – qualitative comparisons
Resultant image quality is very good, but this is to be expected since (i) there is good, close contact between the phosphor and the photodiode array (improving dramatically the light collection efficiency), (ii) the areas of the photodiodes pixels and the phosphor are similar, (iii) there is a high conversion efficiency for optical photons into electron-hole pairs and (iv) the readout of signals from the pixels is also highly efficient.

The noise characteristics are considerably improved over previous systems, which means that both the SNRs and contrast to noise ratios (CNRs) are better than all the first generation EPIDs. This results in a higher DQE for the flat panel imagers, and shows that the devices are now quantum noise limited. Pixel size for all commercial systems is smaller than previous EPIDs, with most devices having arrays which are 10242. Spatial resolution is better although this depends upon the phosphor thickness and not solely on the pixel size [10, 34, 37].

The improvements can be seen qualitatively in Figure 5Go. These images have been acquired with Elekta equipment, but the comparisons are similar for all commercial systems. They compare first generation technology (iView, a camera-based EPID) [42] with a newer AMFPI (iViewGT) [10, 37].


Figure 5
View larger version (53K):
[in this window]
[in a new window]
  		Figure 5. Images comparing, qualitatively, the image quality obtainable with a camera based first generation EPID(iView) and a new AMFPI system (iViewGT). The subject is the neck region of an anthropomorphic type ("Rando") phantom. All images have been acquired with 6 MV X-rays. (a) iView, 2 MU, (b) iViewGT 2 MU, (c) iViewGT 100 MU. From Kirby et al 2002 [41].

 
Images of an anthropomorphic phantom are ideal for assessing comparative image quality between different EPID technologies. Images of the head and neck region (particularly of cervical vertebrae) have always been an excellent visual indicator of the improvements in EPID image quality. Bony detail in the chest (such as the lumbar spine and ribs) and in the pelvis was visible with the earliest EPIDs. However, fine detail, such as the posterior spinous processes in the cervical vertebrae, have always been hard to identify.

The expected gains in terms of reduced noise and improved spatial resolution are clearly visible. The graininess is much reduced for the AMFPI images and there is a considerable increase in the visibility of low contrast, high spatial frequency bony detail, which improves with exposure (Figure 5Go). A simple comparison of structures outlined on the 2 MU exposure images from the two different technologies shows that there can be differences of up to 2–3 mm in identifying the edges of structures [41].

Improvements in SNR are easily seen and, since the quantity of noise is greatly reduced, image processing can be applied to the images with greater benefit. GoFigures 6 and 7Go show examples of this comparing the Varian as500 AMFPI and liquid ionization chamber matrix EPID. Figure 6Go shows an enhanced image for a lateral pelvic treatment. It illustrates how enhancement for the ionization chamber EPID image does show bony detail like AMFPI images, but that the noise within the image is equally enhanced by the processing. Figure 7Go shows how the reduced noise characteristics of the AMFPI permits the use of more aggressive image processing techniques.


Figure 6
View larger version (44K):
[in this window]
[in a new window]
  		Figure 6. Images comparing, qualitatively, the image quality obtainable with(a) an ionization chamber matrix electronic portal imaging device (EPID) (PortalVision) and (b) a new active matrix flat panel imager (AMFPI) system (as500). Both images show bony detail for lateral pelvic treatment fields after image enhancement. (Images courtesy of Margaret Bidmead, Royal Marsden Hospital, London, UK).

 

Figure 7
View larger version (37K):
[in this window]
[in a new window]
  		Figure 7. Images showing the visibility of bony detail(a) before and (b) after image processing and enhancement. Image acquired with the Varian as500 AMFPI. (Images courtesy of Margaret Bidmead, Royal Marsden Hospital, London, UK).

 
Image quality – quantitative comparisons
Quantitatively, the improvements in spatial resolution and SNR are clearly demonstrable. Using contrast detail type-phantoms (such as those described in Kirby et al, Hermann et al, Dong et al, Harrison et al and Low et al [19, 22, 43–45]), AMFPIs demonstrate improved detectability of small-scale (1 mm) low contrast (1.2%) objects at 6 MV [41]. This is similar to that of EC-L film [6]. These thresholds are exposure dependent (as expected) and show that phantoms like the Las Vegas [19, 22] are no longer suitable for assessing AMFPIs; for certain exposures and energies, all objects are now visible [41].

Phantoms like the PIPS QC3V (www.standardimaging.com) can give quantitative analysis of spatial resolution and noise characteristics (such as CNR) [46]. This phantom has been used in studies to examine and compare the image quality of AMFPIs from different manufacturers, the same manufacturer (over a protracted period of time) and also compared with first generation EPIDs [37, 41, 47].

Clements et al [37] examined indices of CNR, 50% (f50) and 30% (f30) relative modulation transfer function (RMTF) for three early examples of commercial AMFPIs. Their results indicated that some systems demonstrate better spatial resolution whilst one showed better noise characteristics. All examples were considerably better than first generation EPIDs.

These results were also observed in a multicentre study (Table 1Go), which examined 10 iViewGT systems and 7 iView systems from four radiotherapy centres in the UK, over a 12–18 month period (from first installation for the iViewGT systems) [47].


View this table:
[in this window]
[in a new window]
  		Table 1. Data from the QC3V phantom examining the contrast to noise ratio(CNR), f50 and f30 values of 10 active matrix flat panel imagers (AMFPIs) (iViewGT systems) and 7 camera based electronic portal imaging devices (EPIDs) (iView systems). f50 and f30 values are in lp mm–1. Data are a "snap-shot" of values taken in 2003 across the 4 UK radiotherapy centres for various exposures. All data are for 6 MV X-rays, 15 cmx15 cm collimator setting, with the phantom in contact with the EPID (approx. 153 cm focus to surface distance)

 
The data show that spatial resolution can be improved by over 50% compared with camera based EPIDs. The CNR is also improved at the longer exposures (by a factor of 4), as expected.

The data in Table 1Go are a single "snap-shot" of the data collected from the EPIDs across the radiotherapy centres during the early part of 2003 [47]. They do not necessarily reflect results expected from newly installed AMFPIs, or represent the best values obtainable. The study showed clearly that there is greater variation in results between examples of the same type of AMFPI compared with older technology. Some of the AMFPI EPIDs (those between 12 months and 18 months old) showed that the image quality was fairly stable; most failures in these early examples were due to wholesale changes (e.g. failures in the acquisition electronics) rather than degradation in the active detection matrices.

Note that the examples shown for the camera-based EPIDs are not as good as the most recent systems. Improvements have been made in terms of light collection efficiency [10, 22] and noise characteristics of the camera [10, 48–50].

Problems with commercially available AMFPIs
As the use of AMFPIs grows, data are being accumulated showing current imaging problems and long-term stability of the devices. Tolerance to radiation damage was proven in the early prototype versions, with the active imaging area able to withstand doses of about 104 Gy without significant changes in performance [34]. In the absence of radiation, pixels accumulate signal due to "dark current" effects resulting from array and sensor leakage currents [34, 38]. These currents are small, which is a requirement if they are not to contribute significantly to the additive noise within the system as a whole. The dark current is stable over long periods of time [34], although there is a dependence upon temperature which may be accounted for dynamically [40].

Ghosting or image lag is another recognized problem [38, 39, 51–53]. Here, charge trapping within the photodiodes (post-irradiation) manifests itself as a latent image on subsequent frames, or alters the gain (or sensitivity) of the a-Si layer itself [38, 39, 51]. The magnitude of the effects is dependent upon exposure and acquisition time (frame rate) [52, 53].

A number of other artefacts have been noted in the clinical setting. These are all types of structural (or fixed-pattern) noise since they are produced in a non-random manner [21]. They significantly affect the visual image quality of the clinical images; more so than random noise components (such as X-ray quantum noise and electronic noise). Some examples are shown in Figure 8Go. They include:


Figure 8
View larger version (90K):
[in this window]
[in a new window]
  		Figure 8. Examples of some typical artefacts that have been observed with commercial active matrix flat panel imager(AMFPI) electronic portal imaging devices (EPIDs). (a) Component of the offset correction image present in clinical images at low exposures (linear, streaking artefact which has been ringed). (b,c) Examples of line artefacts arising from pulsing of the linac. (d) Example of loss of signal in rows of pixels in a complete segment caused by radiation damage to the associated electronics (ringed area).

 

The most common failure modes appear to be associated with damage to the external acquisition and readout electronics around the array [34, 40, 54]. Once affected, whole sections of the panel may cease to function, or rows of individual pixels begin to accumulate no data at all [40, 54], as shown in Figure 8dGo. The timescale for these types of failure is still quite variable, ranging from a few months to years. The failure may be catastrophic (e.g. losing a whole panel section so there is no image data), or more gradual (Figure 8dGo). For the latter, the artefact becomes more predominant over time, eventually making it impossible to adequately assess clinical images. Panels have been changed for this type of problem after about 2–3 years.

One difficulty with AMFPIs, at present, is their maintenance/replacement costs. Currently, the best first generation EPIDs (such as cooled charge coupled device (CCD) camera systems) can produce images as good as the AMFPIs, for certain applications. Whilst this is the case, the user is faced with the dilemma of procuring a mature, well-established system with minimal running costs or a system with more advanced technology and improved imaging characteristics, but with much higher maintenance costs and a variable lifetime.


   Clinical developments
Top
 Abstract
 Introduction
 Technical developments
 Clinical developments
Technical innovations - the...
 Conclusions
 References
 
Clinical implementation and image acquisition
A number of published guidelines are available on clinical implementation [22, 56, 57]. Changes in working practices may be extensive and depend upon the verification strategy being used. Strategies for image acquisition, analysis and set-up correction are varied and can depend on anatomical site [22, 57]. Implementation is best achieved using a multidisciplinary approach and the full range of technologies and strategies available [12].

EPID acquisition modes are varied and image quality is now better than film for certain applications. It is in the flexible use of digital information and availability of processing and analysis software that EPIDs have shown their real value [6, 58, 59]. Good images are possible from both short exposures (for on-line applications) and also from the entire duration of a treatment field exposure (Figure 9Go).


Figure 9
View larger version (64K):
[in this window]
[in a new window]
  		Figure 9. Images acquired with an active matrix flat panel imager(AMFPI) for a lateral pelvic treatment (6 MV X-rays). (a) An image acquired with a reticule present to indicate central axis position, orientation and verify scaling. A 4 MU exposure was used. (b) An image acquired with the full treatment exposure (approximately 80 MU). Distinct differences can be seen in image quality between the two, with (a) having a poorer signal to noise ratio (primarily due to the increased presence of both structured and random noise components).

 
Double exposures or single open field images may be used where treatment fields alone show insufficient anatomy for analysis. This type of acquisition requires high quality imaging with very short exposures, which is achievable with both film and EPIDs. However, even with the latest AMFPIs, the image quality may be dependent upon exposure (Figure 9Go). EPIDs also enable modes of image acquisition that are only possible with digital technology [60–62]. These include movie loop acquisition and display. Individual frames within the loop could be acquired during a single exposure or may be images from different fractions. Images may be acquired from the individual segments of a complex technique, summed or manipulated digitally and then displayed.

Evaluating field placement errors
The main goal of portal imaging is verification of the geometric placement of the irradiated high dose volume [6, 22]. Deviations between the actual and planned geometric positions are termed field placement errors (FPEs) or set-up errors. The FPE will often vary between fractions (interfraction variation) and sometimes during a treatment fraction/field (intrafraction variation). When analysed over the course of treatment, it may reveal systematic and random variations [31, 57, 63]. These may vary between different techniques and anatomical sites [31].

There are various strategies that could be used to evaluate and correct for these FPEs, for example, on an individual image basis in which case both random and systematic components are eliminated. Alternatively, a series of images over a number of fractions may be acquired, analysed and a protocol used to evaluate and correct for the systematic component. Analysis of the random component may be used to determine the margin used in treatment planning to account for motion (both organ motion and set-up error) [63–65].

In most cases, bony anatomy must be used for assessment [22]. However, bony landmarks may not always represent the true position of the PTV. This may be due to relative movement between the respective tissues, normal physiological functions (such as bladder and rectal filling), respiratory movement etc. For certain sites, radio-opaque seed markers may be introduced into the soft tissue within the planning target volume (PTV). Acceptable images are possible with EPIDs (especially AMFPIs) showing the position of the markers, and studies have shown that there can be differences between FPE evaluated through bony anatomy compared with implanted markers [57].

For any evaluation of FPE, some reference image must be available. This may be a verified simulator radiograph or a digitally reconstructed radiograph (DRR) from the original CT planning data. Bony anatomy is compared with the linac coordinate system for both reference and treatment images. The linac coordinate system may be identified using field edges [6, 22, 57] or through an indication of the position of the isocentre (achieved during EPID calibration [66] or with a reticule, as shown in Figure 9aGo). The reticule is easy to use, but has practical disadvantages in that operators need to enter the treatment room to fit it, and it limits the exposure. However, a simple imaging strategy helps minimize the disruption. For example, two orthogonal images could be acquired with the reticule, and then it may be removed prior to treatment. Alternatively for a treatment that includes, say, an anterior and lateral treatment fields, imaging may be performed at the end of one field and at the beginning of the next. If auto (assisted) set-up is used, then the operator need only enter the room twice rather than four times. Identifying the field edges by automatic methods was developed in first generation EPID software. It may also be used for double exposures, whereby contrast is improved for the open field portion by simply overlaying the field edges of the treatment field [6, 57].

FPE evaluation may be performed using simple measurement (using digital rulers etc.). Alternatively individual points or anatomical outlines may be used (outlined manually or semi-automatically) and software used to register images providing a quantitative index of FPE. Much work has been done to automate this [6, 22, 59, 67], but fully automated analysis is still not possible for all anatomical sites. When quantitative analysis is performed on multiple images (usually orthogonal) then both 2D and 3D information may be obtained for the overall set-up error [6, 59, 68].

Once the FPE has been identified, corrections to set-up may be required. If the imaging, evaluation and analysis are performed using short exposures at the start of a single fraction, then it is possible to correct patient set-up prior to delivery of the full treatment. This on-line strategy addresses both systematic and random error components simultaneously, although possibly at the cost of patient throughput [56]. In addition, it needs to be performed every fraction in order to be effective (for patients with a significant systematic error), with possibly higher concomitant dose to normal tissues.

More common is analysis of a series of images from different treatment fractions and a protocol used to identify the systematic and random errors. This is an off-line strategy [57]. Correction of patient set-up is made for the systematic component for the rest of the treatment course with or without further imaging. Data from a population of patients for a particular anatomical site helps to validate the treatment technique itself and quantify the margins required for treatment planning [63, 65].

For both on-line and off-line strategies, integration of the EPID analysis with the record and verify or network system is highly desirable [69]. This verifies the set-up correction itself and also makes full analysis of individual and population data easier. A number of commercial developments are available for this type of networking.

Effective clinical use can only be maintained through appropriate quality assurance for the EPID and its clinical use [70, 71]. This includes quality control for the hardware (to maintain optimum image quality) and also the software (for ensuring that image evaluation and analysis is correct) [57, 72, 73]. The clinical implementation and results should also be periodically reviewed [31, 56, 57].

Volume imaging
Volume imaging is achievable through various methods; in-room CT systems [74], kilovoltage cone-beam CT (CBCT) equipment [4, 5, 75] and megavoltage CBCT techniques [76–79]. Much research is being conducted into all these methods, their clinical implementation and the issues of concomitant dose which they deliver [80]. See also papers by Thieke et al, Moore et al and Chen et al in this issue.

Non-geometric clinical uses
Although the primary use of portal imaging is to verify geometric field placement, dosimetric verification is also possible under special circumstances. For any dosimetric application the EPID response must be a well-defined, quantitative function of the dose delivered, it must be stable and reproducible [53, 55, 81, 82]. All commercial EPIDs have been investigated and a number of applications are possible [83].

Verification of patient dose in vivo
The verification of patient dose in vivo is one of the main dosimetric applications [30, 50, 83–89]. One approach is to predict the dose at the plane of the EPID from the treatment plan and compare it with EPID measurement. Another is to use the two-dimensional information within the portal image to determine the dose in the patient at a point or within a plane. By combining this information with the pre-treatment CT dataset, a volumetric model of the dose distribution can be produced. Ultimately combining EPID dosimetry with treatment time volume imaging will produce a model of the actual delivered dose distribution [76].

To date, the accuracy achieved using first generation EPIDs is between 2% and 5% [22]. Initial studies suggest that AMFPIs are better suited to dosimetry (the calibration and modelling is simpler) and accuracy might improve [39, 40, 81].

IMRT verification
Verification of IMRT delivery is achievable with the EPID in a dosimetric mode [83]. Images of individual segments (for step and shoot IMRT) or full dynamic delivery may be acquired, and MLC leaf positions computed and compared with those predicted [90–92]. Alternatively, the data may be used to compare predicted with actual delivered fluence in one or two dimensions [6, 22, 82–84, 88, 93–95]. One group has used the compact nature of AMFPIs in an ingenious way to perform in-phantom dosimetry for IMRT verification [96]. Recently, measured pre-treatment fluence data have been used as the input to a TPS to compute dose distributions [97]. Some groups are also examining the extraction of geometric information for patient set-up during the actual delivery of IMRT [95, 98].


   Technical innovations – the future for EPIDs
Top
 Abstract
 Introduction
 Technical developments
 Clinical developments
 Technical innovations - the...
Conclusions
 References
 
A number of advances are being made to improve the performance of EPIDs for both planar and volume imaging. Ultimately, imaging of both kinds requires optimum image quality whilst minimizing concomitant radiation doses to the patient. There are three main areas of development. These are (i) looking at further improvements to current EPID technology, (ii) examining ways in which the radiation source itself could make a difference to image quality and (iii) examining completely new technologies for EPIDs.

EPID modifications for improving image quality
All current improvements in EPID design focus on enhancing the quantum efficiency of the detector.

Changes to camera based EPIDS
For camera based EPIDs, attempts are being made to increase efficiency by changing the phosphor, adapting the light collection mechanism and improving the camera.

Light collection efficiency can be improved with large aperture lenses (although possibly with consequent spherical aberrations [10, 22]), and larger format camera target chips. Noise within the system can be reduced by using Peltier cooling mechanisms for CCD based cameras [48, 94, 95].

Changes in the phosphor include adaptations of current materials, but in different geometrical configurations (such as "grooved" screens [23]). Conventional phosphors are rare earth based materials (such as gadolinium oxysulphide Gd2O2S), but their quantum efficiency is only a few percent for 6 MV X-rays. This can be increased 5–6 fold by using CsI crystals. Mosleh-Shirazi et al [77] have used segmented 3 mmx3 mmx10 mm CsI(Tl) crystals, changing the X-ray quantum detection efficiency from 2–3% to about 18%. Each crystal is isolated optically by introducing titanium dioxide powder into the 0.3 mm gap between them. A 3 mm thick aluminium plate acts as the initial X-ray converter, bonded to the crystal layer.

Sawant et al [99] have used a large-area array of CsI in a novel geometric arrangement. Here, incident X-rays pass through the mirror before striking a transparent lead-glass plate (about 1 cm thick), which is coupled (source side) to a transparent CsI(Tl) scintillator layer 1.25 cm thick. Thus, the camera-lens-mirror mechanism is located source side of the X-ray converter and scintillator. For this system the X-ray quantum detection efficiency is about 25%.

Changes to AMFPIs
Indirect detection AMFPIs are also being investigated with different phosphors or phosphor configurations. Seppi et al [78] have used individual 0.38 mmx0.38 mmx8 mm CsI(Tl) crystals. Each crystal was coated (on five sides) with a reflective powder/expoxy resin mixture, with the uncoated end in contact with the flat panel sensor array.

Sawant et al [20] have investigated a design using a two-dimensional matrix of cells dimensionally matched to the pixels of the active array. Each cell is optically isolated from its neighbours and the inside of each cell is reflective to maximize the light reaching the photodiodes. The cells are packed with Gd2O2S:Tb powder, with 3–7 µm grains. The results are promising, but not as good (in terms of DQE) as expected. This is attributed to the strongly depth-dependent light escape efficiency of the phosphor – clear scintillator materials promise further gains.

High QE AMFPI detectors using the direct-detection mechanism are also being investigated [10, 36, 100]. For these devices, a continuous amorphous selenium (a-Se) layer is deposited over the active array. The thick a-Se layer, coupled with a build-up layer, converts X-rays into secondary electrons which are directly converted into electron-hole pairs in the energy sensitive a-Se layer. The holes are collected on the pixel electrodes, stored on the pixel capacitors and read out through the matrix.

Even for these devices, attempts are being made to couple their high efficiency with improved spatial resolution. One group is proposing a design which uses a photolithographic process to create a large number of packed, aligned and focused micro-structured plates [101]. The cavities between the plates are filled with an ionization medium (such as a-Se) surrounded by high-density materials.

Linac beam-line modifications
Some research groups are examining changes in the linac conditions that would favour verification imaging. For example, using a low Z target with low MV energy X-ray or electron beams gives the imaging beam a greater component of kV energy X-rays [102–104]. Low Z materials have been used in place of primary electron scattering foils [105], in place of the main X-ray target itself [102, 105] and also in the secondary electron scattering foil holder [104]. Some configurations remove the X-ray flattening filter to improve X-ray output and maintain a softer beam spectrum [105]. Some clinical results are now available (using the new beam lines with radiographic film) for head and neck patients showing better overall contrast for anatomical features than with film [106].

New designs for EPIDs
One of these novel approaches is the use of kinestatic charge detection methods [10, 107]. These were originally designed for diagnostic applications and use a scanning technique. The detector has separate volumes for detecting the incoming X-rays and collecting the subsequent signal. The detection volume is continuous, but the collection volume consists of a linear array of detectors. A narrow, fan beam of X-rays is used so that when the X-rays interact, a line of ionization charge is produced. An electric field is applied across the volume which drives the charge towards the linear array. Scanning the detector in the opposite direction to that of the driven charge, at a velocity equal but opposite to the mean signal charge drift velocity and perpendicular to the X-ray beam, produces a 2D image. Using a gas (such as Xenon) as the ionization and detection media gives a predicted X-ray quantum detection efficiency of about 36% [107].

Another device is also based upon a gaseous detection medium. It uses gas electron multipliers (GEMs) with the potential for dual energy imaging [10, 108]. A single gas converter overlying a GEM might be used for diagnostic X-rays and a combination of solid converters and GEMs used to increase the detection efficiency for megavoltage imaging.

Expected gains from modified and new EPIDs
All the innovations described aim to increase the efficient use of the X-rays incident upon the detector. The absolute contrast of anatomical objects will inherently be limited whilst megavoltage energy X-ray beams are used. However, improved X-ray quantum detection efficiency and noise reduction throughout the system will increase SNR and thus the overall DQE. This would enable improvements in image quality for low dose imaging.

Improvements in SNR can help in other areas; most notably in allowing the use of more aggressive image processing techniques, making automatic feature detection and extraction algorithms more achievable and improving the detection of implanted markers. It would also improve image quality for volume imaging, reducing the concomitant dose still further and thereby allowing their more frequent use.


   Conclusions
Top
 Abstract
 Introduction
 Technical developments
 Clinical developments
 Technical innovations - the...
 Conclusions
References
 
Great advances have been made in the technical development and clinical use of electronic portal imaging systems. Image quality has improved and, with AMFPIs, promises to progress still further. Cone beam CT techniques have become a practical reality and pave the way for more precise IGRT. Although artefacts, longevity and replacement costs are still being addressed, AMFPI is the technology of the future for EPIDs. The best first generation EPIDs are probably close to their optimum image quality achievable, whereas flat panel technology is still in its development. However, there is still a role for film and other commercial EPIDs and verification devices within a modern radiotherapy department. Clinical implementation and use of EPIDs has reached maturity and protocols show the way to full use. The advent of complex radiotherapy techniques has demonstrated portal imaging is vital. The systems and software now available make it a real, integrated part of the verification process for improving patient positioning and perhaps, in the future, dosimetrically verifying and fine-tuning the actual treatment plan.

Received for publication August 18, 2005. Revision received March 6, 2006. Accepted for publication March 9, 2006.


   References
Top
 Abstract
 Introduction
 Technical developments
 Clinical developments
 Technical innovations - the...
 Conclusions
 References
 

  1. Webb S. The Physics of three-dimensional radiation therapy. Bristol, UK: IOP Publishing, 1993
  2. Boyer AL, Antonuk L, Fenster A, van Herk M, Meertens H, Munro P, et al. A review of electronic portal imaging devices (EPIDs). Med Phys 1992;19:1–16.[CrossRef][Medline]
  3. Henry AM, Stratford J, Davies J, McCarthy C, Swindell, Sykes J, et al. An assessment of clinically optimal gold marker length and diameter for pelvic radiotherapy verification using an amorphous silicon flat panel electronic portal imaging device. Br J Radiol 2005;78:737–41.[Abstract/Free Full Text]
  4. Sykes JR, Beck J, Sharrock PJ, Simpson PD, Stratford J. An integrated tool for on-line image guided radiotherapy using radiographic markers. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:52–3
  5. Sonke J-J, Remeijer P, Betgen A, Zijp L, Smitsmans M, Jaffray DA, et al. Cone-beam kV-CT for image guided radiation therapy. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:12–5
  6. Langmack KA. Portal Imaging. Br J Radiol 2001;74:789–804.[Abstract/Free Full Text]
  7. Haus AG. Film systems for radiotherapy imaging. In: Hazle JD, Boyer AL, editors. Imaging in radiation therapy. Madison, WI: Medical Physics Publishing, 1998:179–206
  8. Hammoudah MM, Henschke K. Supervoltage X-ray beam films. Int J Radiat Oncol Biol Phys 1977;2:571–7.[Medline]
  9. Droege RT, Bjarngard BE. Influence of metal screens on contrast in megavoltage x-ray imaging. Med Phys 1979;6:487–93.[CrossRef][Medline]
  10. Antonuk LE. Electronic portal imaging devices: a review and historical perspective of contemporary technologies and research. Phys Med Biol 2002;47:R31–R65.[CrossRef][Medline]
  11. Haus AG, Dickerson RE, Huff KE, Monte S, et al. Evaluation of a cassette screen-film combination for radiation therapy portal localisation imaging with improved contrast. Med Phys 1997;24:1605–8.[CrossRef][Medline]
  12. Munro P, Kirby MC. The physics and technology of portal imaging. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:138–41
  13. Sonoda M, Takano M, Miyahara J, Kato H. Computed radiography utilising scanning laser stimulated luminescence. Radiology 1983;148:833–8.[Abstract/Free Full Text]
  14. Wilenzick RM, Merritt CR, Balter S. Megavoltage portal films using computed radiographic imaging with photostimulable phosphors. Med Phys 1987;14:389–92.[CrossRef][Medline]
  15. Scheck RJ, Wendt T, Panzer M. Digital storage phosphor radiography for treatment verification in radiotherapy. Br J Radiol 1993;66:801–6.[Abstract/Free Full Text]
  16. Shalev S, Lee T, Leszczynski K, Cosby S, Chu T, Reinstein L, et al. Video techniques for on-line portal imaging. Comput Med Imaging Graph 1989;13:217–26.[CrossRef][Medline]
  17. Munro P, Rawlinson JA, Fenster A. A digital fluoroscopic imaging device for radiotherapy localization. Int J Radiat Oncol Biol Phys 1990;18:641–9.[Medline]
  18. Visser AG, Huizenga H, Althof VGM, Swanenburg BN. Performance of a prototype fluoroscopic radiotherapy imaging system. Int J Radiat Oncol Biol Phys 1990;18:43–50.[Medline]
  19. Kirby MC, Atherton S, Carson P, McDonagh C, Sykes JR, Williams PC. The performance of a fluoroscopic electronic portal imaging device modified for portability. Br J Radiol 1999;72:1000–5.[Abstract]
  20. Sawant A, Antonuk LE, El-Mohri Y, Li Y, Su Z, Wang Y, et al. Segmented phosphors: MEMS-based high quantum efficiency detectors for megavoltage x-ray imaging. Med Phys 2005;32:553–65.[CrossRef][Medline]
  21. Kirby MC. The consequences of fixed-pattern noise and image movement on electronic portal images. Phys Med Biol 1996;41:2345–56.[CrossRef][Medline]
  22. Herman MG, Balter JM, Jaffray DA, McGee KP, Munro P, Shalev S, et al. Clinical use of electronic portal imaging: Report of AAPM Radiation Therapy Committee Task Group 58. Med Phys 2001;28:712–37.[CrossRef][Medline]
  23. Wowk B, Shalev S, Radcliffe T. Grooved phosphor screens for on-line portal imaging. Med Phys 1993;20:1641–51.[CrossRef][Medline]
  24. Partridge M, Evans PM, Symonds-Tayler JR. Optical scattering in camera-based electronic portal imaging. Phys Med Biol 1999;44:2381–96.[CrossRef][Medline]
  25. Meertens H, van Herk M, Weeda J. A liquid ionisation detector for digital radiography of therapeutic megavoltage photon beams. Phys Med Biol 1985;30:313–21.[CrossRef][Medline]
  26. Meertens H, van Herk M, Bijhold J, Bartelink H. First clinical experience with a newly developed electronic portal imaging technology. Int J Radiat Oncol Biol Phys 1990;18:1173–81.[Medline]
  27. Meertens H, Bijhold J, Strackee J. A method for the measurement of field placement errors in digital portal images. Phys Med Biol 1990;35:299–323.[CrossRef][Medline]
  28. Wong JW, Cheng AY, Binns WR, Epstein JW, Klarmann J, Perez CA. Development of a second-generation fiber-optic on-line verification system. Int J Radiat Oncol Biol Phys 1993;26:311–20.[Medline]
  29. Morton EJ, Swindell W, Lewis DG, Evans PM. A linear scintillation-crystal photodiode detector for radiotherapy imaging. Med Phys 1991;18:681–91.[CrossRef][Medline]
  30. Hansen VN, Evans PM, Swindell W. The application of transit dosimetry to precision radiotherapy. Med Phys 1996;23:713–21.[CrossRef][Medline]
  31. Hurkmans CW, Remeijer P, Lebesque JV, Mijnheer BJ. Set-up verification using portal imaging; review of current clinical practice. Radiother Oncol 2001;58:105–20.[CrossRef][Medline]
  32. Street RA, Nelson S, Antonuk LE, Perez Mendez V. Amorphous silicon sensor arrays for radiation imaging. Mater Res Soc Symp Proc 1990;192:441–52.
  33. Antonuk LE, Boudry J, Huang W, McShan DL, Morton EJ, Yorkston J, et al. Demonstration of megavoltage and diagnostic x-ray imaging with hydrogenated amorphous silicon arrays. Med Phys 1992;19:1455–66.[CrossRef][Medline]
  34. Antonuk LE, El-Mohri Y, Jee KW. Active matrix flat-panel imagers (AMFPIs) for electronic portal imaging. In: Hazle JD, Boyer AL, editors. Imaging in radiation therapy. Madison, WI: Medical Physics Publishing, 1998:371–92
  35. Zhao W, Rowlands JA. X-ray imaging using amorphous selenium: feasibility of a flat-panel self-scanned detector for digital radiology. Med Phys 1995;22:1595–604.[CrossRef][Medline]
  36. Pang G, Lee DL, Rowlands JA. Investigation of a direct conversion flat panel imager for portal imaging. Med Phys 2001;28:2121–8.[CrossRef][Medline]
  37. Clements R, Luchka K, Poulliot J, Sage J, Shalev S. Initial comparison of three am-Si EPIDs using the QC3 phantom. In: Proceedings of the 7th International Workshop on Electronic Portal Imaging – EPI2k2; 2002 June 27-29; Vancouver BC, Canada. University of British Columbia, Vancouver, Canada, 2002:108–9
  38. Wischmann H, Luijendijk H, Meulenbrugge H, Overdick M, Schmidt R, Kiani K. Correction of amplifier non-linearity, offset, gain, temporal artefacts and defects for flat-panel digital imaging devices. Proc SPIE 2002;4682:427–37.[CrossRef]
  39. McDermott LN, Louwe RJW, Sonke J-J, van Herk MB, Mijnheer BJ. Dose-response and ghosting effects of an amorphous silicon electronic portal imaging device. Med Phys 2004;31:285–95.[CrossRef][Medline]
  40. Louwe RJW, McDermott LN, Sonke J-J, Tielenburg R, Wendling M, van Herk MB, et al. The long-term stability of amorphous silicon flat panel imaging devices for dosimetry purposes. Med Phys 2004;31:2989–95.[CrossRef][Medline]
  41. Kirby MC, Shentall GS, Bulmer NC, Atherton S, Glendinning AG. Commissioning an amorphous silicon flat panel electronic portal imaging device (EPID) – initial findings. In: Proceedings of the IPEM Biennial Radiotherapy Physics Meeting, 2002 July 16-17, Southampton, UK. IPEM, York, UK, 2002:28–30
  42. Kirby MC, Bulmer NC. Accepting and commissioning an upgraded fluoroscopic electronic portal imaging device (EPID). In: Proceedings of the 6th International Workshop on Electronic Portal Imaging, 2000 June 5-7; Brussels, Belgium. Free University of Brussels, Brussels, Belgium, 2000:28
  43. Dong L, Boyer A. An objective method for evaluating electronic portal imaging devices. Med Phys 1994;21:755–60.[CrossRef][Medline]
  44. Harrison RM, Lambert GD. A method and contrast detail phantom for the quantitative assessment of radiotherapy portal imaging systems. Br J Radiol 1994;67:384–8.[Abstract/Free Full Text]
  45. Low DA, Klein EE, Maag DK, Umfleet WE, Purdy JA. Commissioning and periodic quality assurance of a clinical electronic portal imaging device. Int J Radiat Oncol Biol Phys 1996;34:117–23.[CrossRef][Medline]
  46. Rajapakshe R, Luchka K, Shalev S. A quality control test for electronic portal imaging devices. Med Phys 1996;23:1237–44.[CrossRef][Medline]
  47. Bulmer N, Budgell G, Clements R, Glendinning A, Kirby M, Sage J, et al. Evaluation of iView and iViewGT EPIDs using the QC3V phantom – a multicentre study. In: Proceedings of the PIPSpro User Meeting, 2003 April 3: Manchester, UK
  48. Drake DG, Jaffray DA, Wong JW. Characterisation of a fluoroscopic imaging system for kV and MV radiography. Med Phys 2000;27:898–905.[CrossRef][Medline]
  49. Pang G, Rowlands JA. Electronic portal imaging with an avalanche-multiplication based video camera. Med Phys 2000;27:676–84.[CrossRef][Medline]
  50. De Boer JCJ, Franken EM, Heijmen BJM. A general framework for portal dosimetry with a fluoroscopic EPID based on direct imaging of position dependent crosstalk kernels. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:74–5
  51. Siewerdsen JH, Jaffray DA. A ghost story: spatio-temporal response characteristics of an indirect-detection flat-panel imager. Med Phys 1999;26:1624–41.[CrossRef][Medline]
  52. McDermott LN, Sonke J-J, Wendling M, van Herk MB, Mijnheer BJ, Loot M, et al. A comparison of ghosting effects for two commercial a-Si EPIDs. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:32–3
  53. Fielding AL, Evans PM, Partridge M. Investigation of the dosimetric properties of an a-Si flat panel EPID. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:28–9
  54. Kirby MC. Developments in electronic portal imaging. In: Proceedings of 2D & 3D Radiotherapy Treatment Verification Symposium, 2005 March 2-3, Manchester, UK
  55. Chin PW, Lewis DG. The Varian as500 EPID as a dosemeter: preliminary investigations. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:30–1
  56. Herman MG, Kruse JJ, Hagness CR. Guide to clinical use of electronic portal imaging. J Appl Clin Med Phys 2000;1:38–57.[CrossRef][Medline]
  57. Herman MG. Clinical use of electronic portal imaging. Semin Radiat Oncol 2005;15:157–67.[CrossRef][Medline]
  58. Dong L. Portal image correlation and analysis. In: Hazle JD, Boyer AL, editors. Imaging in radiation therapy, AAPM Medical Physics monograph No. 24. Madison, WI: Medical Physics Publishing, 1998:415–44
  59. Eilertsen K, van den Heuvel F. Image processing and image registration. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:152–6
  60. Kirby MC, Kane B, Williams PC. Clinical applications of composite and realtime megavoltage imaging. Clin Oncol (R Coll Radiol) 1995;7:308–16.
  61. Kubo HD, Shapiro EG, Seppi EJ. Potential and role of a prototype amorphous silicon array electronic portal imaging device in breathing synchronised radiotherapy. Med Phys 1999;26:2410–4.[CrossRef][Medline]
  62. Kron T, Lee C, Perera F. Evaluation of intra- and inter-fractional motion in breast radiotherapy using electronic portal cine imaging. Technol Cancer Res Treat 2004;3:443–50.[Medline]
  63. McKenzie AL, editor. Geometric uncertainties in radiotherapy. Defining the planning target volume. London, UK: British Institute of Radiology, 2003
  64. Prisciandaro JI, Frechette CM, Herman MG, Brown PD, Foote RL, Garces YI. EPID-based methodology to evaluate CTV-PTV margins. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:38–9
  65. Van Herk M. Errors and margins in radiotherapy. Semin Radiat Oncol 2004;14:52–64.[CrossRef][Medline]
  66. Kirby MC. A multipurpose phantom for use with electronic portal imaging devices. Phys Med Biol 1995;40:323–34.[CrossRef][Medline]
  67. Gilhuijs KG, van de Ven PJ, van Herk M. Automatic three-dimensional inspection of patient setup in radiation therapy using portal images, simulator images, and computed tomography data. Med Phys 1996;23:389–99.[CrossRef][Medline]
  68. Remeijer P, Geerlof E, Ploeger L, Gilhuijs K, van Herk M, Lebesque JV. 3-D portal image analysis in clinical practice: an evaluation of 2-D and 3-D analysis techniques as applied to 30 prostate cancer patients. Int J Radiat Oncol Biol Phys 2000;46:1281–90.[CrossRef][Medline]
  69. Eilertsen K, Tveit J, Strom D, Guttormsen E. Electronic portal imaging workflow and analysis integrated in a record and verify system. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:114–5
  70. Mayles WPM, Evans PM. In vivo dosimetry and portal verification. In: Mayles WPM, Lake R, McKenzie A, Macauley EM, Morgan HM, Jordan TJ, et al, editors. IPEM Report 81 Physics aspects of quality control in radiotherapy. York, UK: IPEM, 1999:229–40
  71. Luchka K, Rajapakshe R. Quality assurance of electronic portal imaging devices. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:142–3
  72. Kirby MC, Dowling AR, Bulmer NC. A portal imaging test object for accurately assessing image registration algorithms. In: Proceedings of the VIth International Conference on Medical Physics, 1999a September: Patras, Greece. Phys Med 1999;XV:204
  73. Bulmer NC, Kirby MC. Quality control of treatment verification imaging software. In: Proceedings of the BIR Meeting on Treatment Verification: the radiotherapy team approach, 2001 June 11: London, UK
  74. Court L, Rosen I, Mohan R, Dong L. Evaluation of mechanical precision and alignment uncertainties for an integrated CT/LINAC system. Med Phys 2003;30:1198–210.[CrossRef][Medline]
  75. Jaffray DA, Siewerdsen JH, Wong JW, Martinez AA. Flat-panel cone-beam computed tomography for image-guided radiation therapy. Int J Radiat Oncol Biol Phys 2002;53:1337–49.[CrossRef][Medline]
  76. Pouliot J, Bani-Hashemi A, Chen J, Svatos M, Ghelmansarai F, Mitschke M, et al. Low-dose megavoltage cone-beam CT for radiation therapy. Int J Radiat Oncol Biol Phys 2005;61:552–60.[CrossRef][Medline]
  77. Mosleh-Shirazi MA, Evans PM, Swindell W, Webb S, Partridge M. A cone-beam megavoltage CT scanner for treatment verification in conformal radiotherapy. Radiother Oncol 1998;48:319–28.[CrossRef][Medline]
  78. Seppi EJ, Munro P, Johnsen SW, Shapiro EG, Tognina C, Jones D, et al. Megavoltage cone-beam computed tomography using a high-efficiency image receptor. Int J Radiat Oncol Biol Phys 2003;55:793–803.[CrossRef][Medline]
  79. Pouliot J, Chen J, Morin, Svatos M, Ghelmansarai F, Mitschke M, et al. Dose-guided radiation therapy with megavoltage cone-beam CT. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:16–9
  80. Aird EGA. Second cancer risk, concomitant exposures and IRMER(2000). Br J Radiol 2004;77:983–5.[Free Full Text]
  81. Greer PB, Popescu CC. Dosimetric properties of an amorphous silicon electronic portal imaging device for verification of dynamic intensity modulated radiation therapy. Med Phys 2003;30:1618–27.[CrossRef][Medline]
  82. Van Esch A, Depuydt T, Huyskens DP. The use of an aSi-based EPID for routine absolute dosimetric pre-treatment verification of dynamic IMRT fields. Radiother Oncol 2004;71:223–34.[CrossRef][Medline]
  83. Visser AG, Evans PM. Dose verification and IMRT verification with EPID. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:148–9
  84. Kirby MC, Williams PC. The use of an electronic portal imaging device exit dosimetry and quality control measurements. Int J Radiat Oncol Biol Phys 1995;31:593–603.[CrossRef][Medline]
  85. Pasma KL, Kroonwijk M, de Boer JC, Visser AG, Heijmen BJ. Accurate portal dose measurement with a fluoroscopic electronic portal imaging device (EPID) for open and wedged beams and dynamic multileaf collimation. Phys Med Biol 1998;43:2047–60.[CrossRef][Medline]
  86. Bogaerts R, Van Esch A, Reymen R, Huyskens D. A method to estimate the transit dose on the beam axis for verification of dose delivery with portal images. Radiother Oncol 2000;54:39–46.[CrossRef][Medline]
  87. Glendinning AG, Bonnett DE. Dosimetric properties of the Theraview fluoroscopic electronic portal imaging device. Br J Radiol 2000;73:517–30.[Abstract]
  88. Van Esch A, Vanstraelen B, Verstraete J, Kutcher G, Huyskens D. Pre-treatment dosimetric verification by means of a liquid-filled electronic portal imaging device during dynamic delivery of intensity modulated treatment fields. Radiother Oncol 2001;60:181–90.[CrossRef][Medline]
  89. Wendling M, Louwe RJW, McDermott LN, Sonke J-J, van Herk MB, Mijnheer BJ. Accurate measurement of dose profiles using EPID dosimetry. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:72–3
  90. James HV, Atherton S, Budgell GJ, Kirby MC, Williams PC. Verification of dynamic multileaf collimation using an electronic portal imaging device. Phys Med Biol 2000;45:495–509.[CrossRef][Medline]
  91. Partridge M, Evans PM, van Herk M, Ploeger LS, Budgell GJ, James HV. Leaf position verification during dynamic beam delivery: a comparison of three applications using electronic portal imaging. Med Phys 2000;27:1601–9.[CrossRef][Medline]
  92. Vieira SC, Dirkx MLP, Pasma KL, Heijmen BJM. Fast and accurate leaf verification for dynamic multileaf collimation using an electronic portal imaging device. Med Phys 2002;29:2034–40.[CrossRef][Medline]
  93. Pasma KL, Dirkx ML, Kroonwijk M, Visser AG, Heijmen BJ. Dosimetric verification of intensity modulated beams produced with dynamic multileaf collimation using an electronic portal imaging device. Med Phys 1999;26:2373–8.[CrossRef][Medline]
  94. Bolt RA, Pasma KL, Mestrom LHM, Huizenga H, Bal M, Spies L, et al. Dosimetric pre-treatment IMRT verification with a new CCD camera based EPID. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:98–9
  95. Dirkx MLP, Vieira SC, Heijmen BJM. Integrated verification of the patient geometry and the delivered IMRT fluence using electronic portal imaging for daily 3D dose reconstruction. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:104–5
  96. Moran JM, Roberts DA, Nurushev TS, Antonuk LE, El-Mohri Y, Fraass BA. An active matrix flat panel dosemeter (AMFPD) for in-phantom dosimetric measurements. Med Phys 2005;32:466–72.[CrossRef][Medline]
  97. Steciw S, Warkentin B, Rathee S, Fallone BG. Three-dimensional IMRT verification with a flat-panel EPID. Med Phys 2005;32:600–12.[CrossRef][Medline]
  98. Fielding AL, Evans PM, Clark CH. Verification of patient position and delivery of IMRT by electronic portal imaging. Radiother Oncol 2004;73:339–47.[CrossRef][Medline]
  99. Sawant A, Zeman H, Samant S, Lovhoiden G, Weinberg B, DiBianca F. Theoretical analysis and experimental evaluation of a CsI(Tl) based electronic portal imaging system. Med Phys 2002;29:1042–53.[CrossRef][Medline]
  100. Pang G, Rowlands JA. Development of high quantum efficiency flat panel detectors for portal imaging: intrinsic spatial resolution. Med Phys 2002;29:2274–85.[CrossRef][Medline]
  101. Pang G, Rowlands JA. Development of a novel high quantum efficiency direct-conversion flat panel detector for megavoltage cone beam CT. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:86–7
  102. Ostapiak OZ, O'Brien PF, Faddegon BA. Megavoltage imaging with low Z targets: implementation and characterisation of an investigational system. Med Phys 1998;25:1910–8.[CrossRef][Medline]
  103. Flampouri S, Evans MP, Verhaegen F, Nahum AE, Spezi E, Partridge M. Optimisation of accelerator target and detector for portal imaging using Monte Carlo simulation and experiment. Phys Med Biol 2002;47:3331–49.[CrossRef][Medline]
  104. Flampouri S, Evans PM, Fielding AL, Partridge M, Verhaegen F. Investigation of megavoltage beams and detectors for electronic portal imaging. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:116–7
  105. Zheng Z, Ghelmanserai F, Hristov D, Svatos M. Simple beamline modifications for high performance portal imaging. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:60–1
  106. McNair HA, Evans PM, Flampouri S, Partridge M, Donovan EM, Nutting CM. Evaluation of an optimised portal imaging system for patients receiving radiotherapy for head and neck cancer. In: Proceedings of the 8th International Workshop on Electronic Portal Imaging – EPI2004; 2004 June 29-July 1; Brighton, UK. Institute of Cancer Research/Royal Marsden Hospital, London, UK, 2004:62–3
  107. DiBianca F, Samant S, Laughter J, Rasmussen J, Rodriguez C. Use of a kinestatic charge detector for megavoltage portal imaging. Proc SPIE 1997;3032:195–201.[CrossRef]
  108. Ostling J, Wallmark M, Brahme A, Danielsson M, Iacobaeus C, Fonte P, et al. Novel detector for portal imaging in radiation therapy. Proc SPIE 2000;3977:84–95.[CrossRef]



This article has been cited by other articles:


Home page
 Br. J. Radiol.Home page
I PATEL, T NATARAJAN, S S HASSAN, and M C KIRBY
The use of computed radiography for routine linear accelerator and simulator quality control
Br. J. Radiol., October 1, 2009; 82(982): 827 - 838.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Radiol.Home page
R M HARRISON
Doses to organs and tissues from concomitant imaging in radiotherapy: a suggested framework for clinical justification
Br. J. Radiol., December 1, 2008; 81(972): 970 - 974.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kirby, M C
Right arrow Articles by Glendinning, A G
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kirby, M C
Right arrow Articles by Glendinning, A G

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS