Section 9: Computerized Tomography
THE HARP ACT EXEMPTS CT DEVICES FROM MANY STANDARDS. IN 2006, THE PROVINCIAL GOVERNMENT ASSESSED THE REGULATORY NEEDS FOR OPERATION OF CT IN ITS DIAGNOSTIC IMAGING SAFETY COMMITTEE. THAT COMMITTEE MADE A REPORT BUT AS OF OCTOBER 2010 - NO STANDARDS OR SPECIFIC REGULATIONS PERTAINING TO CT ARE ON THE BOOKS OTHER THAN THE DESIGNATION REQUIREMENT. THE FOLLOWING IS WHAT XRIS WOULD HAVE RECOMMENDED AS A STANDARD IF ANYONE HAD THE FORESIGHT TO SEE THEY WERE CLOSE TO 20 YEARS AHEAD OF THE REST OF THE WORLD.
Introduction
The X-ray Inspection Service does not specifically test computerized tomography units nor are any calibration standards specified under the HARP Regulation. However, it is expected that routine quality control or preventive maintenance be performed at least every 6 months ensuring that the manufacturers standards are met. Detailed records of these test results must be kept on site.
Inspection Procedure
In the course of an inspection the following information will be verified:
- The installation matches plans approved by the XRIS.
- X-ray operators are qualified and currently registered with the College of Medical Radiation Technologists of Ontario.
- Written safety procedures are in place.
- An estimate of potential exposure to personnel (estimate may be obtained by taking scatter readings usually in the control booth and at the foot end of the x-ray table).
- Visual Indicators: warning sign(s), Line "ON", X-ray "ON", kVp and mA selectors. All controls must be clearly identified, emergency "OFF" button.
- Ensure that quality assurance program is in place and that records are kept. The records should include the results of the tests listed below.
Every six months, or more often depending on manufacturer's instructions and standards, check for:
- a. Sensitivity Profile
- b. kVp Accuracy
- c. Spatial Resolution
- d. Plane Indication and Alignment
- e. Scan Increment Accuracy
- f. Laser Alignment
- g. Thickness of Cut
- h. Position Indicator Accuracy
- i. Scaling (Distance) Accuracy
- j. Noise/Resolution
Maintain records on file which document:
- a. The CTDI Index
- b. Log Book for Repairs
- c. Procedural Manual. This manual should include the following:
- List the scan protocols to be used
- Policies on the use of oral contrast
- Use of intravenous contrast
- Slice thickness, slice spacing, anatomical description of the extent of the field for the scanogram
- The technical factors selected for the examination
- Patient and staff shielding policy
- Restraining policy
- Pregnancy policy
- Darkroom standards (if applicable).
- Condition of patient shielding.
Acceptance Tests Requirements
For any particular unit, it is important to establish the level of performance during acceptance testing of the unit. An appropriate quality control program will ensure maintenance of this performance level over time. At acceptance testing, the manufacturer is required to furnish specific dose information obtained by using the CT dosimetry phantom and imaging performance information.
Following the installation of new or used CT scanners, acceptance tests must be performed and submitted to the X-Ray Inspection Service within sixty days of the installation. Acceptance tests should include, but not necessarily be limited to, the following information:
- Installation conformity to approved plans for radiation safety.
- Actual Patient Entrance Exposure (Patient Entrance Exposure) measurements for examinations most commonly performed (i.e., head and abdomen):
- These measurements should be compared to the dose index profile (CTDI) provided by the manufacturer.
- Results should include measurements taken at 3 different MAS values for 3 different slice thickness available (if applicable).
Note:
The RPO of the facility should decide on how these patient dose measurements are to be done as long as the test procedure can be duplicated at a later date to verify consistency in output.
- Results for all the tests listed under Inspection Procedure in item 6.
Additional Tests for Computerized Tomography
- Calibration
For a unit with CT numbers expressed in Hounsfield Units, the unit should be so calibrated that air yields an average CT number of -1000 +/- 3 and water yields an average CT number of 0 +/- 1.5.
- Constancy
The CT number for water, and its standard deviation over a water-filled phantom, should be verified (daily).
- Uniformity
In an image of the 20 cm diameter phantom filled with a uniformly attenuating medium, the mean CT number of any 100 pixels should not differ by more than 5 from the mean CT number of any other 100 pixels. Some CT units may experience difficulty in meeting this specification; for these units, a more relaxed criterion may be appropriate. The uniformity image can be very helpful in identifying the presence of image perturbations such as beam hardening artifacts, detection nonuniformity rings, etc. (to be done at least monthly).
- Dependence on Slice Thickness
For the 20 cm diameter phantom filled with a homogeneous medium, the mean CT number averaged over 100 pixels at the centre should agree to within +/- 3 CT numbers for all selection thicknesses (to be done semi-annually).
- Dependence on Phantom Size
The CT number of water should change by less than +/- 20 as the phantom diameter changes from 5 to 30 cm, provided that reconstruction techniques and beam shaping filters appropriate for the phantom size are used. (Phantoms as small as 5 cm in diameter are useful if limb studies are performed with the unit or if the unit is used for pediatric patients ( to be done semi-annually.)
- Dependence on Phantom Position
The mean CT numbers should vary by less than +/- 5 CT numbers if the phantom is centred in the gantry, or +/- 15 CT numbers, irrespective of the position of the phantom (to be done annually).
- Dependence on Algorithm
The mean CT number at the centre of the phantom should vary by less than +/- 3 CT numbers irrespective of the algorithm chosen for the reconstruction, provided that the appropriate phantom size is used for each algorithm (to be done annually).
- Hard Copy Output and Visual Display
For a standard image, the hard copy image should not differ noticeably in luminance and contrast from the image displayed on the video monitor, as determined by visual evaluation.
- Low Contrast Resolution
For a midline exposure of 0.5 to 0.8 mC/kg (2 to 3 R), pins of 0.5 cm diameter should be visible in the image when the x-ray attenuation of the pins (intrinsic contrast) differs by 1 per cent or more from that of the surrounding medium ( to be done at least monthly).
- High Contrast Resolution
Pins of 0.1 cm diameter should be visible in the image when the x-ray attenuation of the pins (intrinsic contrast) differs by 1- per cent or more from that of the surrounding medium (to be done at least monthly).
- Video Monitor and Hard Copy Distortion
For a series of holes positioned 1 cm apart in a high contrast test object, the holes in the video or hard copy image should coincide with the location of the holes in the object to within +/- 1 mm in the central 50 per cent of the image when the image is projected to life size. In the periphery of the image, the coincidence may not satisfy the +/- 1 mm specification. (To be done at least monthly; more frequently if images are used for treatment planning or for other quantitative applications.)
Table
- Position Indicator Accuracy
Over repeated positioning of the examination table using the bed index control, the true position of the bed should agree to within +/- 2 mm of the indicated position (to be done semi-annually).
- Backlash
Location of the examination table should agree to within +/- 1 mm when approached from opposite directions (to be done semi-annually).
- Indexing Accuracy
With an appropriately loaded bed, the bed should move the nominal index amount (+/- 0.5 mm) for each increment. This accuracy can be verified by a ruler or by a series of exposures with prepackaged film taped to the bed. When the CT unit is used to produce projection radiographs from which measurements ( e.g., leg length) are obtained, tests to verify the localization accuracy and the accuracy of the quantitative measurements are necessary (to be done semi-annually).
- Sensitivity Profile (Slice Width or Section Thickness)
The section thickness should agree to within +/- 1 mm of the nominal thickness for slice thicknesses between 5 and 15 mm, and to within 0.5 mm for slice thicknesses less than 5 mm, when measured at the full width at half maximtun (FWHM) with a wire oriented at 45 degrees in a test object (to be done semi-annually).
- Noise Characteristic
The standard deviation of CT numbers should vary approximately mAs-1/2 over the region where the image is dominated by quantum mottle (to be done semi-annually).
- Accuracy of Distance Measurements (Caliper Accuracy)
Distances measured in a CT image should agree with actual distances to within +/- 1 mm (to be done annually, more frequently if data are used quantitatively).
- Patient Dosimetry
Exposures should agree within +/- 20 per cent among measurements obtained at different time (to be done semi-annually).