Acceptance Testing of Computed Radiography (CR) Systems_



Acceptance Testing of Computed Radiography (CR) Systems

Ehsan Samei

samei@duke.edu

September 17, 1997

The acceptance testing of the CR imaging systems requires an x-ray source and a display device (soft- or hard-copy) previously evaluated for satisfactory performance.

All the imaging plates should be cleaned and erased prior to executing the testing procedures. Keep the delay time between exposing the plates and reading them within a limited range (1-15 min). Except for the test No. 12, no image processing algorithms should be applied to the images. For Kodak systems, select the PATTERN body part, and “raw data” and “no edge enhancement” selections, with the level set to the obtained value for the Exposure Index and the window set to 1024, unless otherwise is specified. For Fuji systems, the default setup for the body parts that have been specified in this document should generate a raw image (GA = A, RE = 0.0).

Equipment Needed:

1. A calibrated x-ray source (with its compliance report).

2. A calibrated hard/soft-copy display (with its compliance report).

3. A densitometer (if a hard-copy display is to be used).

4. A measuring tape.

5. A flash-light.

6. A role of masking tape.

7. Copper and aluminum filters (1 mm Al, 1 in Al, 2.5 mm Cu,).

8. An ion-chamber (RadCal).

9. A stand for the ion chamber.

10. Screen cleaning solution and cloths.

11. A metric steel ruler (for laser-beam function test).

12. A distortion (beads) phantom (optional).

13. Line-pair phantoms (at least 2) up to 5 lp/mm frequencies.

14. The Leeds phantom (TO.12).

15. A screen-contact wire-mesh pattern.

16. A screen-contact fine wire-mesh pattern (mammography).

17. A small lead block (> 3mm thick).

18. An anti-scatter grid (10:1 or 12:1, 103 ln/in) (if the x-ray system does not have one).

19. Anthropomorphic phantoms (foot, hand, pelvis, chest, etc.).

20. A timer.

21. A lead sheet or lead apron (to reduce backscatter)

22. Two lab jacks.

0. Document

Document the systems specifications as outlined in the worksheet.

1. Physical Inspection/Inventory

Check all the components of the system against the Purchase Order. Make sure they are free from physical defects and properly installed.

2. Dark Noise

Erase an imaging plate and read it without exposing it.

For Kodak systems, read the plates using the PATTERN body part. Set the level to 512 and the window to 1024. The image should be uniform without any artifacts except some bands in the plate-scan direction (collector profile). The Exposure Index should be less than 100 and 400 for GP and HR plates, respectively.

For Fuji systems, read the plates using TEST, SENSITIVITY body part with a fixed EDR (S set to 2000). The image should be uniform without any artifacts. Measure the average pixel value in the image and its standard deviation. The average pixel value should be higher than 1020 and the standard deviation less than 0.5. If using a hard-copy display, the optical density (OD) should be less than 0.05 plus base fog.

3. Uniformity (Imaging Plate Test)

This test is applied to all the imaging plates. Visually inspect all the plates for physical defects. Verify that the cassette label matches the type of screen inside. Expose the plates to 10 mR entrance exposure using 80 kVp, 1 mm Al/0.5 mm Cu filter with 180 cm source-to-image distance (SID). Do not use the bucky for the exposures; rather place the cassette either on the bucky or on an elevated platform above the floor (to avoid backscatter). Use the above specified reading conditions, except for the Fuji system, in which case use a Semi EDR. The image should be uniform without any artifacts. Measure the average pixel value in the image and its standard deviation. 2 x SD should be within 10% of the average pixel value. If using a hard-copy display, the optical density at the center of each quadrant should be within 10% of the average value. Plate-to-plate variation should also be within 10%.

4. Exposure Indicator Calibration

Use multiple IPs (2-3) of a given size/type. Expose the plate to approximately 1 mR entrance exposure using 80 kVp and 1 inch of AL filter at the collimator. Use a precise delay of 10 before reading the plates.

For Kodak systems, each time read the plate using the PATTERN body part. Record the Exposure Index and average pixel value for each image. Verify that for 1 mR (0.9 - 1.1) entrance exposure, the average Exposure Index of multiple is within the 1975-2025 range for each type and size of the imaging plates. The Exposure Index from any single plate should fall within 1950-2050 range.

For Fuji systems, each time read the plate using TEST, SENSITIVITY (L = 1) body part with Semi EDR. Record Sensitivity and the average pixel value in each image. Verify that for 1 mR entrance exposure, the average Sensitivity of multiple plates is within the 190-210 range for each type and size of the imaging plates. The Sensitivity from any single plate should be within 180-220 range.

It should be noted that currently no universally accepted beam quality is accepted by industry for exposure indicator calibration of CR systems. Kodak recemmends an 80 kVp beam with 1 mm Al/0.5 mm Cu filtration and Fuji recommends an 80 kVp beam with no filtratoion. This test may be repeated with these specifications or with kVps and filtration more typical of applications commonly performed at the facility. It should be recognized, however, that the use of 1 inch Al filter minimizes the dependency of the results on the x-ray source, as the filter eliminates the soft spectra which is responsible for tube-to-tube variations. It also makes the spectrum a more accurate representative of those incident on the imaging plate after attenuation through the patient body in clinical applications. The results may, therefore, be used for consistency check in annual compliance inspections. However, the accepted range of exposure indicator for any clinical application should be specified. The test may also be repeated using various kVp and filtration and the results plotted as a function of kVp (for a fixed filtration) and as a function of filtration (for a fixed kVp). Minimal beam quality dependency should be observed.

5. Sensitivity Response

This test is only applicable to Fuji systems. Use a single ST 14x17 imaging plate (multiple IPs may also be used if the plate-to-plate variations in the previous test were found minimal). Expose the plate to approximately 0.1, 1, and 10 mR entrance exposures in a sequence of three exposure-reading cycles using 80 kVp, 1 mm Al/0.5 mm Cu filter, and 180 cm SID. Each time read the plate with a constant delay time using TEST, AVE 4.0 body part (Latitude = 4) with a Semi EDR. Record the Sensitivity values and plot them as a function of exposure on a log-log scale. The result should be more or less a straight line, depending on the accuracy of the exposure measurement. Pixel value variations (or the OD in the case of hard-copy prints) should be within 10%. Repeat this test using TEST, CONTRAST (L = 2) and the same exposure conditions.

6. Linearity and Slope

Use a single imaging plate (multiple IPs may also be used if the plate-to-plate variations in the first test were found minimal). Expose the plate to approximately 0.1, 1, and 10 mR entrance exposures in a sequence of three exposure-reading cycles using 80 kVp, 1 mm Al/0.5 mm Cu filter, and 180 cm SID. Use a constant delay time.

For Kodak systems, each time read the plate using the PATTERN body part. Record the Exposure Index and average pixel value for each image. Plot the average pixel value (or the OD in the case of hard-copy prints) as a function of exposure on a linear-log scale. The result should be a straight line with a slope not more than 10% deviation from the expected value (1000).

For Fuji systems, each time read the plate using TEST, AVE 4.0 body part (L = 4) with a Fix EDR (S set to 200). Record the average pixel value (or the OD in the case of hard-copy prints) in each image and plot them as a function of exposure on a linear-log scale. The result should be a straight line with a slope of not more than 10% deviation from the expected value (-256). Repeat this test using TEST, CONTRAST (L = 2) and the same exposure conditions. Verify that the relationship is linear and the slope is within 10% of the expected value (-511).

7. Laser Beam Function

Place a steel ruler roughly perpendicular to the laser-scan direction on a standard-resolution IP. Expose the IP to about 5 mR entrance exposure using a 60 kVp beam without any filtration (SID = 180 cm). Process the plate using TEST, SENSITIVITY, Semi EDR setup for Fuji systems and PATTERN for the Kodak systems. Examine the edges of the ruler on the image for laser beam jitters (under- or over-shoot of the scan lines in light to dark transitions). Ruler edges should be straight and continuous. Use a 10X magnification in various areas of the image to verify uniform scan line spacing (scan line dropouts are detectable as lucent straight lines in the open field).

8. High-Contrast Resolution

This test should be done for each type and size of the imaging plates. Use a 60 kVp, unfiltered x-ray beam (SID = 180 cm). Place two line-pair pattern devices on the IP, each at 5 degrees from the laser-scan and the plate-scan directions. Expose the plate with an exposure of about 5 mR. Process the plate using TEST, SENSITIVITY, Semi EDR setup for Fuji systems and PATTERN for the Kodak systems. Set the window and level to optimally visualize the line-pair patterns using 2-3X magnification. The limiting resolution should be within 10% of the expected value in both the laser-scan and the plate-scan directions. In order to examine the MTF of the system at a lower spatial frequency, acquire of an image of a fine wire mesh (mammography screen-film contact test tool) using the techniques and processing settings given above. The wire mesh should be equally visible over the entire image.

8. Noise/Low-Contrast Resolution

This test should be done for each type and size of the imaging plates. Use a 75 kVp beam with 1 mm of Cu (SID = 180 cm). For each plate type, acquire three images of the Leeds TO.12 low-contrast phantom using 0.1, 1, and 10 mR exposures to the plates. Use a constant delay time of 10 minutes before reading each of the plates. For Kodak systems, each time process the plate using the PATTERN body part with window set to 512 and level to 4096-(Exposure Index) for GP plates, and 4096-(Exposure Index)-300 for HR plates. For Fuji systems, each time read the plate using TEST, CONTRAST body part with Semi EDR.

Record the contrast threshold for the row D (4.0 mm diameter) objects as a representative of the contrast-detail behavior. For each type of plate, the contrast threshold should be proportional to the square root of the exposure, with higher contrast thresholds for standard-resolution plates.

9. Distortion

Place a regular wire-mesh screen-film contact test tool over a standard resolution IP. Expose the IP to about 5 mR entrance exposure using a 50 kVp beam without any filtration (SID = 180 cm). Process the plate using TEST, CONTRAST, Semi EDR setup for Fuji systems and PATTERN for the Kodak systems. Examine the uniformity of the grid spacing. Repeat the acquisition with two steel rulers in the vertical and the horizontal directions (or a beads phantom with known spacing between the beads). Verify that the measurements in both directions on the image (soft-copy or hard-copy, taking into account the magnification/minification) are within 2% of the actual spacing.

10. Erasure Thoroughness

Place a thick lead block at the center of a 14 x 17 standard-resolution IP and expose the plate to about 100 mR using a 50 kVp x-ray beam without any filtration (SID = 180 cm). Erase the plate (in some systems, the plate can only be erased after being read), and expose it a second time to 1 mR entrance exposure without the lead object using the same beam quality collimated about 2 inches from each side of the plate. Process the plate using TEST, SENSITIVITY, Semi EDR setup for Fuji systems and PATTERN for the Kodak systems. Using a narrow window setting, verify absence of a ghost image of the lead block from the first exposure in the obtained image.

11. Anti-Aliasing

This test should be performed for each type and size of the imaging plates that will be commonly used. Place the plate in a bucky that contains an anti-scatter grid (a grid may also be placed on the plate instead) so that the grid lines are parallel to the laser-scan direction. Make sure the grid movement is disabled. Expose the plate to 1 mR using an 80 kVp beam filtered with 1 mm Al/0.5 mm Cu and a SID according to the specification of the grid. Process the plate using TEST, CONTRAST, Semi EDR setup for Fuji systems and PATTERN for the Kodak systems. Repeat, placing the plate perpendicular to the laser-scan direction. Verify that the moiré pattern is only apparent when the grid lines are parallel to the laser-scan direction. If a moving grid is to be used for some applications, repeat the tests with the grid moving, and verify that no moiré pattern is seen when the plate is placed in either direction.

12. Positioning and Collimation Errors

Using standard radiographic techniques, acquire images of various anthropomorphic phantoms with various combinations of collimation and positioning, utilizing the appropriate prescribed algorithms of the system. Verify that in each case the image is properly tone-scaled and is free from unexpected positioning and collimation errors. It is advised to interact with the radiologists to verify the acceptability of the images.

13. Throughput

Process a number of unexposed 14x17 plates (4 to 10) as fast as possible. Check the throughput (plates per hour) against system’s specifications. The deviation should not be more than 10%. For systems for which the throughput may also be a function of exposure level (through variations in the erasure time) (i.e., Kodak systems), this measurement can be repeated for plates exposed to about 2 mR (80 kVp).

This concludes the technical acceptance testing of a CR system. However, prior to a full clinical implementation, the default tone-scaling and frequency processing parameters of the system for all the applicable algorithms (body parts) should be customized by the application specialists of the manufacturer in consultation with the radiologists. It is also important to assure that the default image processing parameters are consistent for CR readers that are installed in the same clinic.

APPENDIX:

Procedure for measuring the exposure

1. Use a vertical x-ray beam with the focal spot at about 205-210 cm from the floor (a cross-table beam may also be used if the imaging plate can be placed in a reproducible position and at least a meter away from the room walls).

2. Lay a lead sheet (or lead apron) on the floor.

3. Make a stand for the imaging plate on the lead sheet with two lab jacks to allow the plate to be positioned more than at least 25 cm from the floor (the plate is positioned perpendicular to the tube axis).

4. Verify the focal spot-to-imaging plate distance is 180 cm

5. Place a calibrated ion chamber at the center of the beam (indicated by the light collimator) at 127.3 cm from the focal spot. The distance should be measured to the center of the chamber.

6. Collimate the x-ray beam to only cover the chamber with no more than 1 inch margins.

7. Make the exposure without the imaging plate in place and divide the measured exposure by 2 in order to determine the exposure that the plate would receive. Change the mAs setting to obtain the desired exposure.

8. To expose the plate, remove the chamber from the beam, place the imaging plate on the lab jacks, and open the collimator to fully cover but not exceed the imaging plate by more than 1/2 inch.

[pic]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download