Address correspondence and reprint requests to Margaret C. McEntee, at the above address. E-mail: email@example.com
Portal radiography involves the acquisition of images to visualize radiation treatment field(s) using the radiation treatment source. The standard has been the use of film-based systems with improvements over the years in film–screen technology providing near diagnostic quality images. More recent advances have included the development of digital systems with such notable improvements including ability to window/level images to enhance viewing and readability, and significant shortening of the time required to acquire images.
Many steps are involved in the precise and accurate delivery of radiation therapy. One of the most important steps involves the accurate alignment of the radiation beam relative to the target (tumor in the patient) that is accomplished through portal radiography.1 A key component of this process involves patient immobilization techniques and reproducibility of daily positioning for radiation therapy.2–5 Portal imaging, to provide field setup verification, effectively reduces systematic and random setup errors.6 Random errors are those deviations that occur between different fractions, and systematic errors are deviations between the intended patient position and the average patient position over a course of fractionated radiation therapy. Errors in setup can result in insufficient dose coverage of the targeted tumor volume, excessive irradiation of adjacent normal structures, and compromised clinical outcome. Port films are radiographs made using the radiation treatment source, and are used to visualize the radiation treatment field to assess inclusion of tumor and adjacent normal anatomical structures.7
Portal images are created using a single-exposure, double-exposure, or an orthogonal-pair technique. A single-exposure technique may suffice if there are adequate anatomic landmarks within the radiation treatment field for assessment of the treatment field placement. A double-exposure technique is most commonly used and is particularly useful for small treatment fields and those that do not include bony landmarks for visualization and verification of field placement. One exposure is made of the treatment field and then any field shaping blocks are removed and the collimators are opened and a second exposure is made. The treatment field is exposed twice and therefore is definable relative to the surrounding anatomic area. The orthogonal-pair technique involves a dorsal–ventral film and a medial–lateral film or other appropriate orientations perpendicular to each other. Typically port films are evaluated by assessing field placement relative to bony anatomic landmarks. Another option is the implantation of gold seed fiducial markers in the target organ or in close proximity in surrounding normal tissue.8,9 Visualization of fiducial marker location within the treatment field can provide a surrogate for the target organ, and provide an orientation for correction of errors in treatment field placement.
Based on a survey of veterinary radiation oncology facilities in 2001, various port film protocols are routinely used in the majority of facilities before and often during the course of radiation therapy.10 Weekly portal imaging has long been the standard in human radiation oncology facilities, although this is changing with the move to more conformal radiation treatment fields requiring a greater level of precision for patient setup with as frequent as daily imaging.11 In veterinary radiation oncology facilities port films are typically acquired on the first day of treatment to assist in setting up the radiation treatment fields. Port film evaluation is most commonly accomplished through visual comparison of a port film with a reference image. In patients with computed tomography (CT)-based, computer-generated three-dimensional radiation treatment plans, the portal images can be compared with a digitally reconstructed radiograph generated by a three-dimensional treatment-planning system. In patients that have manual treatment field setups the port film assists determination of the adequacy of margins around a tumor or surgical incision, and visualization of adjacent critical normal tissues which may need to be blocked. For manual setups subsequent portal images can be compared with the original definition of the treatment field through the day 1 portal image.
Port films for all patients are considered a permanent part of that patient's medical record and provides information as to the actual area and tissues irradiated, assessment as to whether or not a regional recurrence is within the irradiated volume or is a geographic miss, and in patients that develop an unrelated second neoplasm the assessment can be made as to whether or not the new lesion can be safely irradiated or if there will be overlap of the radiation treatment fields.
There are two steps in routine portal imaging after image acquisition. The first step is to determine the accuracy of treatment field placement and the second step is to decide if there is an error and if a correction should be made. Depending on the treatment field size and margins around the tumor incorporated into the planning target volume, any errors <5 mm may not be corrected.12 Furthermore it has been shown that visual inspection alone is not adequate to identify field placement errors up to 5 mm.12 Some facilities incorporate additional margins around the target volume to compensate for the fact that it may not be possible to detect small systematic variations in field position on the order of 2–3 mm.13 In evaluating port films an estimate is made as to the x and y translations in millimeters that will result in centering of the treatment field appropriately in comparison with that visualized on the digitally reconstructed radiograph or initial portal image. In systems with image matching software, an objective determination can be made as to the changes necessary to appropriately align the treatment field to the area to be irradiated.14 The accurate positioning and alignment of treatment fields is of particular concern as we move to more conformal therapy including investigation of and clinical implementation of intensity modulated radiation therapy (IMRT) in veterinary oncology.15
Portal Screen–Film Technology
When making portal images using a megavoltage radiation source, the high energy photon interactions with the patient are primarily by Compton scattering and therefore provide less contrast than diagnostic X–rays. Some facilities have a diagnostic radiology imaging unit in the radiation treatment room to image a patient's setup which can be mounted to the head of the linear accelerator.10 Fortunately there have been advances made in the portal screen–film technology such as the Kodak EC system* that resulted in enhanced contrast and image quality with megavoltage imaging.16–19 Currently available film–screen systems available for portal imaging provide near diagnostic quality images.1 Film–screen combinations currently available require only a few monitor units to give an optimally exposed image. One potential problem is that some linear accelerators take a few monitor units to stabilize dose output, so the dose received by the film may be unpredictable. Linear accelerators equipped with a film mode setting with a lower output per minute for acquiring portal images will likely provide more uniform film quality.
The Kodak EC film system uses a 1-mm-thick copper front screen to produce electrons. These electrons interact with a gadolinium oxysulfide-intensifying screen to produce light that exposes the film. There is a separate cassette used for localization (EC-L system) vs. verification (EC-V system). Technique charts (based on use in people) are available for the Kodak system, but the body thickness and field size in veterinary radiation oncology patients make use of these charts difficult. Based on clinical experience, there is a more limited range of the number of monitor units used in portal imaging of small animal veterinary radiation oncology patients.19,20
Port films must be developed using a processor. The time required for processing depends on the particular film processor. At our facility this has ranged from 90 s using the radiology department processor to 5½ min using a table top processor that replaced the decommissioned radiology department processor. Depending on the location of the processor, transit time and processing time should be considered. One option is to acquire portal images and process and review the films at a later time but this delays implementation of corrections in treatment field setup. An additional concern with the use of a film processor is maintenance of quality of the images produced if the processor is not being used on a regular basis. The additional time required to maintain a processor, replenish chemicals, etc., should also be taken into account. In veterinary radiology the trend is toward the installation of digital, film less imaging systems decreasing the availability of film processors.
Digital Imaging Options for Portal Radiography
Using digital images, one can manipulate the contrast and brightness of images to optimize viewing and to facilitate comparison of the portal image to the digitally reconstructed radiograph. The portal images are quickly and readily available throughout the hospital using a PACS. Digital images are easier to distribute throughout the hospital and require less physical storage space compared with analog radiographs. Additionally the radiation oncologist can review digital portal images and digitally reconstructed radiographs remotely for assessment of field placement and corrections.
The Kodak system is a Windows-based system that digitizes computed radiography (CR) plates using the Kodak 2000RT CR System desktop scanner, and sends the images to a workstation. The same imaging plate can be used for simulation, localization, or verification imaging. It is possible to view and manipulate images (rotate to match digitally reconstructed radiograph orientation, flip to ensure proper orientation, magnify, change window/level), measure angles and distances, and annotate images. An additional useful function is the ability to designate image status to facilitate work flow relative to image review and approval. This involves having the therapist and/or radiation oncologist review the images although only an oncologist can approve the portal images. Images can be coded as unread, reviewed, and approved by an oncologist, reviewed and rejected by a therapist or an oncologist, or as corrective action has been taken on a rejected image. Of notable benefit is the ability to import and view digitally reconstructed radiographs and portal images side by side for visual assessment of the portal images. The treatment planning computer can be connected to the Kodak system to allow downloading of the digitally reconstructed radiographs (single or orthogonal depending on patient setup). Patient and study information can be created to attach to images for ready identification of individual patient images, and a drag-and-drop feature can be used to move images in the patient list. CR plates are scanned with low resolution typically adequate for portal scanning. The scanned images are DICOM compliant and it is possible to send, query, retrieve, print, and archive image files over a local area network or wide area network. The system is left on during the work week and shut down for the weekend. When the system is brought back on line it automatically goes through a self-test to ensure that the 2000RT meets acceptable performance standards. The only other maintenance of the system that is required is cleaning of the plates once a week using anhydrous alcohol and lint free wipes.
For facilities that have been utilizing the Kodak EC-L cassette system, the cassettes themselves can be used with the CR system. Two screens are necessary for portal imaging to allow orthogonal imaging without processing delay. There is a small decrease in the number of monitor units that are used to expose the screen as opposed to the film system on the order of 1–2 monitor units per exposure with the double-exposure technique. The most common exposure technique used at our facility with the double-exposure technique is 1 monitor unit for the first exposure and then 1–2 monitor units for the second exposure. Port film dose is often not subtracted from the treatment dose and therefore represents additional radiation dose to the patient. This is a consideration particularly with the double-exposure technique wherein potentially a large surrounding area of normal tissue is irradiated during the second exposure.21 It takes 25 s for the reader to scan the reusable plate and display the image on the monitor. After an additional 30 s, the plate is ready for reuse after it has been automatically erased by exposure to fluorescent light. Room lights are dimmed to avoid degradation of image quality as the plate is transferred to the processor, and to optimize ambient illumination for viewing of the portal images on the monitor. Options for manipulation of the image includes the ability to rotate the image, change the contrast level to optimize viewing of the object of interest (e.g., bony anatomical landmarks, metal hemoclips, or other metallic markers in the patient), magnify the image for viewing and side-by-side comparison with the digitally reconstructed radiograph, and the ability to measure structures and distances and annotate the image to create a permanent record of shifts in patient and/or field position made based on the portal image. Compared with analog portal systems, fewer port film repeats are needed using digital systems. Postprocessing of digital images allows the user to enhance portions of the image that may not have been apparent using analog portal systems. Digital images (localization and verification) can be acquired with increased frequency without adding substantial time to the treatment or cost. Although the initial investment in the Kodak CR system is greater than that for the cassettes, film, and processor for the film based system, the ongoing day-to-day costs are less. The cost per film is approximately $3–4. Within our department the policy is to charge on a per port film basis. The application of a charge for imaging with the CR system will offset the expense of acquiring the system.
At the author's facility port films are made at least once a week. For more complex setups and fields with more narrow margins orthogonal double-exposure port films are acquired on day 1, repeated on day 2 or 3 of radiation therapy, and then are made as frequently as is deemed appropriate. For instance if the treatment field shift each time is approximately 1–2 mm this is tolerable and the frequency of portal radiography is decreased. If larger treatment field shifts are noted, 4–5 mm or greater, then the frequency of making portal images is increased to every 2–3 days of treatment. The therapist and/or radiation oncologist reviews port films before treatment and treatment field shifts are made in real time before that day's treatment. Conversely, at some human radiation oncology facilities, there may be a delay in the radiation oncologist reviewing the port films and an additional one or two treatments may be delivered before any changes are instituted. At our institution veterinary radiation therapists are trained to review the port films in real-time and make the necessary changes with oversight provided by the veterinary radiation oncologist. Furthermore compensatory table shifts to correct patient position are facilitated by an XYZ tabletop coordinates. These coordinates can be zeroed and the table unlocked in just one direction to facilitate table position adjustments of 1 mm or greater.
A gradicule is placed in the head of the linear accelerator to help determine the size of shifts required to center the treatment field in the x and y planes. Using the gradicule, crosshairs and markers that are 1 cm apart are projected onto the port film. Rotational patient shifts are more difficult to detect and accurately correct. Reproducible patient positioning is critical. The number of monitor units used to make portal images is documented on the patient's treatment chart and that technique is subsequently used each time portal images are acquired. Once a patient has completed a course of radiation therapy, all portal images (Fig. 1) and the associated digitally reconstructed radiographs (Fig. 2) are sent to the PAC system. This facilitates review of patient radiation treatment field information as patients return to the hospital for recheck appointments.
Using the CR system, making port films is easier and this may be an impetus for increased acquisition of images. The increased patient dose should be considered.21,22 There is some risk to the patient in terms of cancer induction from radiation exposure due to leakage and scatter from the collimated radiation treatment beams, with additional risk potential with radiation from localization and verification of the radiation treatment fields through portal radiography.21 Although information is not available on the potential impact of such exposures relative to tumor induction in veterinary patients, it is advisable to consider the potential sequela from portal radiography and strike a balance between the necessity to verify treatment field position and the risk of cancer induction with frequent imaging. Using double-exposure portal imaging, the potential induction of cancer in sensitive organs outside the treatment volume is increased. Increased frequency of portal imaging and hence increased radiation exposure is expected with conformal or IMRT. In conformal radiation or IMRT the treatment plan entails a narrow margin around the tumor with the intent to escalate dose while minimizing dose to surrounding normal tissues. IMRT requires more frequent imaging to ensure accurate patient positioning and radiation delivery. The experience of the individual setting up the patient for radiation therapy affects the number of portal images needed. More portal images means more patient dose. Any time saved using digital systems can be lost by increasing the number and frequency of portal imaging.
Other Imaging Options
Advances in human radiation oncology include the use of electronic portal imaging devices (EPIDs) incorporated into linear accelerator construction, and kilovoltage and megavoltage cone-beam CT systems used in image-guided radiation therapy.23–26 These imaging options, particularly EPIDs, will likely be slowly integrated into the field of veterinary radiation oncology due to the initial cost involved. The use of EPIDs and cone-beam CT systems requires an additional level of attention to various aspects including acceptance testing, commissioning, and quality assurance. It is also possible to obtain information on patient dosimetry through portal imaging and information on anatomy. The exit dose from the patient can be measured through calibration of the imager and image. These functions can be facilitated more readily with the use of an EPID but can also be carried out using portal imaging film and CR systems. EPID can acquire images with close to real-time display. Improvements in immobilization techniques and automation of the process for correcting systematic errors will allow for design and implementation of tighter target volumes.
Limited veterinary studies have been done to compare film-based portal imaging systems.16 To the authors knowledge there are no veterinary studies comparing the image quality of film-based systems and CR system. Based on initial side by side comparison of the two systems, and limited experience with the Kodak CR system, the CR system provides equal or better image quality with less radiation dose to the patient and the images are acquired quicker. Anatomic detail of small patients (cats) and small body parts do not have much detail using either system, but the detail is slightly better using the film based system. Overall, the CR system with the list of associated improvements and enhancements is preferred over the film-based system.
*Kodak, Rochester, NY.
This work was supported in part by Dr. Peter Malnati, DVM '51 Cornell University.
Disclosure of Conflicts of Interest: The authors have declared no conflicts of interest.