Evaluation of a test device to assess x-ray phototimers.
The Mammography Quality Standards Act of 1992, a federal law, requires that the phototimer be tested extensively during annual inspections of mammographic units. However, there is no similar federal requirement for other types of radiography units. Several states regulate the testing of phototimers on radiographic systems, but most require only a check of minimum/maximum response times (mAs) and verification of a back-up tinter.
Film-screen systems used in mammography have a much narrower exposure latitude, which explains why they are subject to a more rigorous, mandated set of standards. However, phototimers are a fundamental feature of most clinically installed radiographic systems and their performance should be evaluated on a regular basis, whether mandated or not. Poorly calibrated phototimers result in incorrectly exposed radiographs, leading to retakes, wasted film and unnecessary radiation exposure to patients.
A well-calibrated phototimer tracks both thickness and kVp. Essentially, this means that the phototimer keeps the radiation exposure (film optical density) within an acceptable range when the thickness of the material being radiographed changes but kVp remains constant and, conversely, when kVp changes but the thickness of the material remains constant.
Proper evaluation of the phototimer on a radiographic system requires the production of 5 to 10 images, depending upon whether or not a table Bucky and chest Bucky are used. A complete assessment might also include a check of the cell balance in the field to ensure that the left, center and right sensors are producing the correct radiation exposure.
For fluoroscopic units, the phototimer for the spot film device should be checked for thickness/kVp tracking and reproducibility of exposure in different image formats.
For mammographic units, phototimer testing requires tip to 25 films, all of which must be processed and measured for equal density. This includes production of films with the following parameters:
* Tracking at three clinical kVp stations with 4 cm of phantom material.
* Tracking at 2 cm, 4 cm and 6 cm or more of phantom material at the same kVp.
* Producing images that use a small grid, a large grid, a large focal spot with no grid and magnification mode with a small focal spot and no grid.
* Producing images at all density selector settings (eg,-5 to +5).
* Producing images using all automatic modes.
Completing all these tests takes a substantial amount of time and requires a large number of films, resulting in a significant cost in labor and material. This article investigates whether using a commercially-available test device can reduce the time, labor and expense involved in assessing the performance of phototimers on x-ray systems.
The test device evaluated was the Automatic Exposure Control Analyzer manufactured by Diagnostic Imaging Specialists Corporation, St. Malo, Manitoba. The AEC device consists of a digital, liquid-crystal display readout meter, an electronic test cassette and a 9-foot connecting cable. Separate test cassettes are available for testing mammographic systems. (See Fig. 1.)
[Figure 1 ILLUSTRATION OMITTED]
The radiographic test cassette contains a sensor that duplicates the energy response of green and blue rare earth, two of the most commonly used intensifying screen phosphors. The sensor converts the incident radiation into an electrical signal that is proportional to the light emitted by the intensifying screens as the radiation is absorbed. The electrical signal is sent to the meter, where it is integrated and converted to a digital value for display.
Specifications for the radiographic analyzer are:
* kVp range: 60 kVp to 130 kVp.
* Minimum exposure time: 2 msec.
* Maximum exposure time: 0.4 sec at 800 speed, 0.8 sec at 400 speed and 1.6 sec at 200 speed.
* Energy response: +/-5% (as compared to film).
* Power: two 9-volt batteries.
* Cassette size: 26.7 x 26.7 x 1.3 cm.
* Meter size: 20.5 x 16.2 x 2.7 cm.
* Weight: 1.7 kg.
* Screen compatibility: rare earth green and blue light.
The manufacturer of the device recommends that conversion/calibration curves be produced for each film-screen and processing system to determine the value of the digitally displayed number that corresponds to the measured optical density. To determine if the AEC test device would provide reproducible and accurate measurements from different generators, calibration curves were produced for two radiographic systems at 60 kVp, 80 kVp and 100 kVp. The two radiographic systems used were a Philips system (Philips Medical Systems, Shelton, Conn) and a Toshiba system (Toshiba America Medical Systems, Tustin, Calif).
The radiographic calibration curves were created under the following conditions:
* Lanex regular film-screen cassette, 400 speed, 24 x 30 cm.
* 20 cm Lucite phantom in 2.5 x 30 x 30 cm slabs.
* Field size autocollimated to Bucky, with radiation field within perimeter of the phantom.
* The central cell was used as the phototimer domain.
* The optical density range was 0.5 to 2.2.
* Tube currents used were 320 mA to 500 mA, large focus.
* A source-to-image distance of 100 cm was used for all measurements.
The films were processed on a Kodak Multiloader model 700 system (Eastman Kodak Company, Rochester, NY) and read with a Victoreen densitometer, model 07-443 (Victoreen Inc. Cleveland, Ohio).
Data was collected in the following manner:
* With the test cassette in the Bucky, individual measurements were made by varying the position of the density control selector switch from its highest setting to its lowest:.
* Three readings from the AEC meter were taken and averaged at each density selector setting.
* Identical exposures were made with the test cassette removed from the Bucky tray and replaced with the regular film-screen cassette.
* The film was processed immediately and its optical density was measured. Thus, the AEC test cassette was alternated with the film-screen cassette in the Bucky tray without delay during the recording of data. (See Fig. 2.)
[Figure 2 ILLUSTRATION OMITTED]
Exposure measurements were made randomly during the production of the images using an ion chamber to assure that the exposures were consistent. The kVp accuracy was checked for each generator during the investigation using a noninvasive kVp meter. Deviations in kVp accuracy were never greater than +/-3%.
The data for 60 kVp, 80 kVp and 100 kVp for both the Philips and the Toshiba generators were very similar. (See Figs. 3 and 4.) There was a slight deviation at 60 kVp where the optical density values on the films were slightly lower than those at 80 kVp and 100 kVp, particularly for the high-density selector positions. As expected, the use of 20 cm phantom material at 60 kVp resulted in high exposure times. The resulting drop in optical density values probably was a result of film reciprocity failure.
[Figure 3 and 4 ILLUSTRATION OMITTED]
At 80 kVp, the data from the Philips and Toshiba generators were almost identical. (See Fig. 5.) When the AEC numbers were entered into a spreadsheet, the data generated curves that best fit second-order polynomials with the resulting R-squared values close to 1.00, where the value of y represents the optical density and x represents the AEC number. The closer the value of R-squared is to unity, the more linear the relationship of optical density to AEC number and the more faithfully the meter readings can be substituted for optical density values.
Another example of the excellent correlation between measured optical densities and AEC readings at both 80 kVp and 100 kVp is shown in Table 1. The numbers in Table 1 represent data selected at random from all the measurements made during the process of constructing the calibration curves. The range of AEC numbers recorded was between 2.2 and 9.6, for a corresponding optical density range between 0.5 and 2.4. (The wide range in numbers for both the AEC device and optical densities is due to the methodology of collecting the calibration data at all density settings of the phototimer, from lowest to highest.)
Table 1 AEC Readings vs Measured Optical Densities(*) Toshiba Philips AEC Optical kVp Reading Density Reading Density 80 3.1 0.79 3.0 0.81 5.0 1.37 4.6 1.33 7.7 2.06 7.2 2.00 100 4.4 1.26 4.0 1.15 6.0 1.67 5.4 1.63 7.8 2.07 7.0 2.00
(*) A 20 cm Lucite phantom was used.
As Table 1 shows, reproducibility -- which can be gauged by comparing the AEC reading for the two systems at similar optical density levels -- is high. The AEC readings also track well with the optical density as it increases or decreases.
The data in Table 2 were obtained from a Philips high-frequency generator and phototimer at a second location using a different processor but the same film-screen speed (if 400. The AEC numbers generated were so close to those obtained at the first h)cation that the optical densities that could be predicted from these numbers from the initial calibration curves differed by only an average of 5% from the measured values.
Table 2 Thickness and kVp Tracking With Philips Generator at Second Location Thickness Optical AEC kVp (cm) Density Reading 60 20 0.80 3.0 80 20 1.06 3.7 100 20 1.20 4.0 80 10 0.80 2.9 80 20 1.06 3.7 80 30 1.08 3.7
[Figure 5 ILLUSTRATION OMITTED]
Measurements were taken with the chest phototimer in one radiographic room at 90 kVp, 105 kVp and 120 kVp using a Lanex 400-speed system. The predicted optical densities then were compared to those taken at 80 kVp in the table Bucky in two other radiographic rooms, also using a 400-speed system. The correlation was so close that the average deviation from the predicted optical density was only 4.4%.
The Mammography Quality Standards Act requires that film density values be reported in assessing phototimer performance. Because more phototimer test films are required in mammography than radiography, there is great interest in a test tool that can be used in place of film for mammography. Such a tool would save money by reducing the amount of time and film involved in assessing phototimer performance.
We examined the performance of a mammography test cassette from Diagnostic Imaging Specialists Corporation, the same company that manufactures the AEC device. The mammography test cassette is smaller than its radiographic counterpart (12.5 cm vs 26.7 cm) to make it compatible with various manufacturers' cassette sizes. Rather than a toggle switch to change from one screen phosphor to another on the radiographic cassette, it uses inserts to mimic the various film-screen combinations.
Three mammographic screen inserts were evaluated. The first, the Min-R-Medium, produced erratic data or simply did not respond, especially with 6 cm of phantom material. The other two screen inserts tested, the Min-R-2000 and Min-R-Regular, produced more meaningful, reproducible data. Figs. 6 and 7 show the calibration curves of the AEC readings vs optical density values for two commonly used mammographic systems. The differences in AEC values in the two graphs are the result of a change in design to notably increase the gain of the system.
[Figure 6 and 7 ILLUSTRATION OMITTED]
Although effective in most applications, the AEC analyzer device has several limitations. For film-screen combinations at speeds other than 400 in wall Buckys, another calibration curve must be generated. This became obvious when an AEC meter reading of 5.5 for both a 400-speed system and a 250-speed system in a chest Bucky with the same technique and phantom resulted in measured optical density values of 1.62 and 1.20 respectively, a difference of 35%.
The phototimer in the spot film device in R/F rooms could pose another limitation for the AEC device. In these situations, the usefulness of the AEC test cassette depends upon the mechanical design of the x-ray unit and the manual dexterity of the operator. Care must be taken so the cable connecting the test cassette to the display meter does not get caught inside the spot film loader. Holding onto the cable with one hand, depressing the load/unload button with the other and guiding the test cassette and cable into and out of the spot film changer usually overcomes this problem. On some traits, the latch to the cassette loader door must be propped open. However, units that will not accept the test cassette will require the assistance of a service engineer. For these types of units, evaluation of the spot film phototimer is best accomplished by the direct film method.
Another precaution is that the measurements produced by the AEC device are based upon incident radiation rather than optical density. Therefore, new calibration data must be generated if there is any change in the parameters, including the processor, processor quality control and image receptor speed.
For this study, the emphasis was on phototimers in table Buckys on radiographic systems. For film-screen systems with the same rated speed, phototimers in wall Buckys and table Buckys should respond the same way. Thus, the AEC values for table Buckys produced in this study can be used to assess the performance of phototimers in chest boards, as long as the same film-screen system is used.
The close correlation of the AEC numbers at the 80 kVp station and the small differences in AEC numbers at all three of the kVp stations lends credibility to using the 80 kVp curve as the sole reference in calibrating and evaluating phototimers. This also was the conclusion reached by Rossi/who evaluated the device but only reported data at 80 kVp. Rossi stated that establishing the relationship between measured image optical density and AEC readings at 80 kVp (lid not result in any significant discrepancies between measured optical density and predicted optical density when used at other kVp stations.
An example of the ability of the device to handle different thicknesses of phantom is illustrated by Table 3. Obviously, this phototimer is well calibrated, since the range of measured optical densities is only 0.11 units from lowest to highest, while the corresponding AEC numbers have a range of only 0.8.
Table 3 Thickness Tracking Performance At 80 kVp (Toshiba Generator) Thickness Optical AEC (cm) Density Reading 10 1.51 5.1 15 1.55 5.5 20 1.62 5.8 25 1.62 5.9 30 1.58 5.9
For an investigation with so many possible variables, it is remarkable not only that reproducible data was obtained, but also that the margin of error of predicted density values from one location to another was so small. As a follow-up to this evaluation, future studies ((mid attempt to confirm the reproducibility of the radiographic AEC values with different processors. To date, the reproducibility of the data using the same film-screen speed with (different processors has been acceptable. It has been helpful to take at least (me control film at each processor location at an average kVp and phantom thickness to compare the measured optical density with that predicted from the calibration curves.
At our institution, we made a calibration/conversion chart for the in-house service engineers to use in adjusting phototimers in the table Buckys. The medical physics department has started to use the AEC device in routine, annual quality assurance physics inspections and for acceptance testing of new equipment.
The Automatic Exposure Control Analyzer is an effective test device for assessing the performance of phototimers in radiographic systems. It can be used in acceptance testing, annual physics inspections and routine quality control spot checks. It also has the potential to save film and labor costs involved in the annual inspections of mammography systems.
[1.] Bushberg J, Seibert J, Leidholdt E, Boone J. The Essential Physics of Medical Imaging. Baltimore, Md: Williams & Wilkins; 1994:96-98,429.
[2.] Rossi RP, Lin PJ, Rauch PL, Strauss KJ. Performance specifications and acceptance testing for x-ray generators and automatic exposure control devices. AAPM Report No. 14. New York, NY: AAPM; 1985:1-65.
[3.] Strauss KJ. Radiographic equipment and components: technology overview and quality improvement. In: Gould RG, Boone JM, eds. A Categorical Course in Physics. Oak Brook, Ill: Radiological Society of North America; 1996:39-43.
[4.] Rossi RP. Evaluation of a device for indirect assessment of automatic exposure controls. Medical Physics. 1994;21 (1):141-143.
Gerald L. Buchel, M.S., the principal investigator of the research reported in this article, is a medical physicist in the Medical Physics Department at Baylor University Medical Center, Dallas, Texas. Mircea N. Sabau, Ph.D., and T. Lyle Wilson, M.S., are medical physicists at Baylor University Medical Center, Dallas, Texas. John J. Sadler, B.S., is a health and radiation safety technologist at Baylor University Medical Center.
Reprint requests may be sent to the American Society of Radiologic Technologists, Communications Department, 15000 Central Ave. SE, Albuquerque, NM 87123-3917.
[C] 1998 by the American Society of Radiologic Technologists.
|Printer friendly Cite/link Email Feedback|
|Author:||Buchel, Gerald L.; Sabau, Mircea N.; Wilson, T. Lyle; Sadler, John J.|
|Date:||Jul 1, 1998|
|Previous Article:||Radiologic technologists' responses to elderly patients.|
|Next Article:||Work settings of newly certified R.T.(R)s.|