GB 17589-2011Specifications for quality assurance test for computed tomography X-ray scanners (English PDF)

4.1 Diagnosis bed

4.1.1 minimum scale of 1mm, an effective length of 500mm near the straightedge diagnosis fixed bed of moving bed, and to ensure that the ruler Parallel to the direction of movement of the bed and the bed surface can indicate as a ruler tick mark pointer.

4.1.2 ensure that the bed load of about 70kg.

4.1.3 were given the diagnosis of bed "into 300mm" and "back 300mm" instruction.

4.1.4 record forward, back the start and end point values shown on the ruler, measure the positioning error and homing error.

4.2.1 Motif detection

4.2.1.1 detector module having a body with a surface clear alignment marks, embedded within the specific shape of the object, the shape of the object, the position of Should mold surface alignment mark with strict spatial relations.

4.2.1.2 The detection module body is placed in the center line of the radiation field is fixed, the mold body axis perpendicular to the scanning cross section, so that all surfaces of the phantom trimming Marking and locating light coincide.

4.2.1.3 the head of clinical common exposure conditions, the total thickness of less than 3mm imaging collimator axial scanning mode to obtain positioning light Markable layer image, comparing the image with a standard shape and positional relationship between the level of a particular object are the same, if the same, then the alignment light, accurate. 4.2.1.4

4.2.1.3 If the two do not match, the scan back and forth perpendicular to the axis of the level of fine-tuning die body in accordance with

4.2.1.3 Scanning conditions, ultimately consistent with the nominal level of the image, according to the mold body to adjust the distance along the axis, positioning light to determine the degree of deviation.

4.2.2 Film detection

4.2.2.1 Select an unexposed film in a dark room with black paper wrapped tightly and keep it neat.

4.2.2.2 In the paper side of the direction perpendicular to the longitudinal direction draw two straight lines, a straight line distance of 10cm or more, respectively, with a needle through the two straight lines 2 and 3 holes, hole diameter should be as small as possible and should not exceed the maximum diameter of 1mm, hole spacing at least greater than 5cm.

4.2.2.3 The perforated with small holes on the film flat on the examining table, the film coincides with the central axis of rotation and CT center will first have to wear two Linear positioned within the aperture of light projected on the film coincides with the CT machine positioning within the set mode, choose a lower exposure conditions, minimum Nominal thickness, the use of single-axial scan mode to scan.

4.2.2.4 and then pan the film, will wear a straight line with three holes positioned outside the projection of light on the film coincides with the CT machine outside the given set Bit mode, using the same exposure conditions

4.2.2.5 final rinse scanned film were measured inside, positioned between the outer light corresponding scanning line of the image on the film and the line in a spin Turn on the central axis of the pitch, the pitch as an inner and outer locating the light of deviation.

4.3 gantry angle precision

4.3.1 center with a clear mark of rectangular motifs, the phantom center coincides with the center of the fault wild, and the level fixed, according to

4.2 optical positioning accuracy of test results, adjust the position of the mold body to determine the level of the scan, so scan through the phantom center level.

4.3.2 commonly used in clinical use for scanning head scanning conditions.

4.3.3 motif stationary rack tilted at an angle, in accordance with the conditions in

4.3.2 scanned again.

4.3.4 ranging use workstation software, distance between the upper and lower edges of the cross-sectional image measurement phantoms, were recorded as L1 and L2.

4.3.5 using the equation (5) calculate the actual value of the gantry angle, compared with the set value, determines the accuracy of the gantry tilt angle. alpha = arccosL1L2 (5) Where. alpha

--- gantry angle size.

4.4 Reconstruction thickness deviation

4.4.1 for axial slice thickness deviation measured using phantom embedded with a homogeneous background high contrast markers, markers have determined The geometry, through its geometrical position reflects imaging reconstruction thickness; spiral CT machine for measuring the thickness of a sheet deviation or marker Beads, the attenuation coefficient of not less than the marker material is aluminum, to ensure high signal to noise ratio, and the Z-axis direction of the maximum thickness or diameter should be Between 0.05mm ~ 0.1mm.

4.4.2 The phantom axis perpendicular to the scanning level and placed in the scanning field center fixed.

4.4.3 According to the head of exposure conditions, the image reconstruction set the nominal thickness, axial or helical scan.

4.4.4 adjusted according to the mold body image viewing conditions specification window width and position, and for the record, layer thickness measurements reconstruction.

4.4.5 For axial scan reconstruction thickness variation measurement, current measurement methods There are two common motif.

a) Method One. adjust the window width to a minimum, change the window level until the marker image just completely disappeared, record the CT values, namely CTmax, under the window width and position conditions, measured CT value marker near background is CTbackground, the CT value of half-height CT values from the two and a half, denoted CThm, and then re-adjust the window level to CThm, this time measuring markers Length, full width at half maximum (FWHM), and then using a fixed geometric relationship markers, calculated layer thickness measurements reconstruction; this Methods commonly used to set the layer thickness is greater than 5mm thickness deviation detection reconstruction.

b) Method two. labels used in the method of determining the geometric relationship with minute details, by counting the number of details and Combined with the geometric relationship between the details of the calculated thickness measurements reconstruction, the method can be used to detect and set the layer thickness relative to the diameter of details Close rebuilding thickness deviation.

4.4.6 thickness deviation measurement method for the reconstruction of spiral CT machine is. with helical scan mode scanning markers, and a nominal thickness of 1/

10 The image reconstruction interval, and the total width of the Z-axis direction of image reconstruction of at least 3 times the nominal thickness; with an appropriate ROI (e.g. thin markers Tablets, the ROI is set for 2 times the diameter of the sheet, if it is micron beads, then the ROI set point. ) Measurements to obtain a series of helical sweep Image description sheet or the average CT values beads material; record the average CT value for the Z axis as a function of the horizontal curve, and determine the FWHM function curve, that is, as the FWHM of thickness measurements reconstruction.

Note 1. The reconstruction of the thickness of CT machine is referred to the default image reconstruction thickness.

Note 2. For older spiral CT machine, it is difficult to achieve a nominal thickness of 1/10 of the reconstructed image, you can fine-tune the scanning starting point to get multiple sets of image reconstruction, measuring The average amount of CT values for these center of the image and make the curve, determine FWHM.

Note 3. For multi-slice CT, currently limited to the detection of multi-axial scan reconstruction slice thickness, testing phantoms having a defined geometric position markers Z Axis direction should be long enough, if the length does not meet the needs of multi-scanning, can be divided according to the layout of the detector array open, detect, and Z-axis center to ensure phantom markers as part of the array to be detected in the same scan levels.

Note 4. having used to determine the geometry of the high-contrast markers mold body should clearly be able to detect the smallest thickness d, the minimum thickness according to equation (6) Calculation get. d = T/costheta (6) Where. The thickness of the T

--- marker level scanning angle formed.

4.5 CT dose index

4.5.1 The use of human tissue-equivalent material homogeneous cylindrical phantom head die having a diameter of 160mm, the phantom diameter of 320mm, respectively, in the Heart and 10mm from the surface there may be placed at a dose probe hole, relative error dose measuring instrument should be less than 5%, and has been calibrated.

4.5.2 The headform or phantom placed scanning field center, the axis of the cylindrical molded body with an effective level of vertical center of the scanning probe, the scanning probe With its central location level.

4.5.3 respectively, according to the head and body commonly used in clinical conditions axial scan.

4.5.4 Records dosimeter readings and calculated CTDI100 and CTDIw according to formula (1) and (2).

4.6 CT value (water), noise and uniformity

4.6.1 using the homogeneous aqueous cylindrical phantom.

4.6.2 the mold body cylindrical axis perpendicular to the scanning level and in the scanning field center, the middle level of the cylindrical scan.

4.6.3 using the scanning head scanning conditions, and each radiation dose scan mode at the position of the center of the body should not exceed 50mGy.

4.6.4 Select the center of the image is about 500 pixels in size (about one-tenth of the mold body area) ROI, measure the ROI of the average CT Value, standard deviation, wherein the average CT value of water as measured CT value, standard deviation divided by the contrast scale as noise measurement value n, See equation (7). n = sigma CT water air water -CT × 100% (7) Where. standard deviation sigma water

--- water phantom ROI measured; CT Water

--- water measured CT values; Air

--- CT CT value of air measurements. CT Water -CT air

--- contrast ruler.

4.6.5 Also in the image corresponds to the circumference of the watch clockwise 3,6,9,12 point, according to the motif of about 10mm at the edge of the image, select about ROI 500 pixels size, the average CT values were measured four ROI, wherein the center of the image and

4.5.4 ROI average CT Value as the maximum difference value of the uniformity measurement.

4.6.6 High Contrast the special algorithm corresponding Measurement resolution, while

4.5 should press the scanned image reconstruction algorithm is special, according to the method of measuring

4.5 Application of the special reconstruction algorithm After recording the image noise and the noise should be less than 1.5%.

4.8 low-contrast detectability capacity

4.8.1 Details molded body with diameter typically between 0.5mm ~ 4mm, and a background of 0.3% to 20% in the contrast between, And the minimum diameter of not greater than 0.8mm, the minimum contrast ratio shall not exceed 0.5%.

4.8.2 The phantom placed in the center of the scan field, and its axis perpendicular to the scanning level. 4.8.3

4.5.3 with scanning conditions.

4.8.4 The molded body image viewing conditions or manual adjustment to achieve observer status that the details of the most clearly.

4.8.5 The minimum diameter of each recording contrast detail can be observed, and the noise level as amended, normalized to the noise level of 0.50% Details diameter under background conditions, and then multiplied by contrast, mean different contrast details of the product as a low-contrast detectability Ability Detection value.

4.8.6 noise level correction according to equation (8) is calculated. Tsigma2R3 = k1D (8) Where. The T

--- nominal thickness, in millimeters (mm); sigma

--- noise level,%; R

--- the smallest detail can be observed diameter in millimeters (mm);

--- scale factor K, is a constant, regardless of their specific values; D

--- scanning dosage units of milli-Ge (mGy).

4.7 High contrast resolution

4.7.1 uses can be evaluated by direct observation or using phantom image transfer function (modulatedtransfer by calculating modulation function, MTF) evaluate the spatial resolution high-contrast motifs, calculated MTF motif description and corresponding high contrast resolution measurements Methods According to GB/T 19042.5.

4.7.2 for direct observation image motif should be evaluated with periodic detail, the spacing between the periodic structure that should be a single week Equal to the width of the details of their own, effective attenuation coefficient differences in the details of the periodic effective attenuation coefficient and homogeneous background caused by CT values The difference should be greater than 100HU.

4.7.3 The phantom placed in the center of the scan field, and the cylinder axis perpendicular to the scanning level. 4.7.4

4.5.3 with scanning conditions.

4.7.5 The molded body image viewing conditions or manual adjustment to achieve observer status clear that the details of most, but not greater than the window bit thin Section CT CT value and the background value of the difference.

4.7.6 can distinguish count was recorded minimum size or MTF curve corresponding to 10% of the value of the spatial frequency detail as spatial periodicity Discrimination of measuring value.

4.7.7 If a special algorithm to obtain high contrast resolution should first recording method according to

4.9 CT number linearity

4.9.1 using embedded with four or more different CT values phantom module, and the difference between the value of the CT module should be greater than 100HU.

4.9.2 using phantom instructions specified scanning conditions or is used in the clinic were scanning the head and body scanning conditions, respectively.

4.9.3 In the center select different modules of about 100 pixels size ROI, measure the average CT value.

4.9.4 in accordance with label instructions phantom various attenuation module attenuation coefficient in the corresponding ray radiation quality conditions, calculated the various modules The average value is then calculated for each CT module, the nominal value of the resultant CT measurement of the module; the nominal value of CT in this ray radiation quality conditions CT value difference, the difference between the maximum evaluation parameters were recorded as CT values linear.