Catphan 500 and 600 Manual March 2006

3

T h e

P h

a

n

t

o

m

L a

b

o

r

a

t

o

r

y

Catphan® Manual Contents Warranty

1

Introduction

5

Multi-Slice CT by David Goodenough, Ph.D. Initial phantom positioning

6

8

Illustration of Catphan® models

9

Incremental phantom modules positioning

10

Phantom position verification 11 CTP401 and CTP404 modules 12 Patient alignment system check

13

Scan slice geometry (slice width)

14

Scan incrementation Circular symmetry

15 16

Spatial linearity of pixel size verification Spherical acrylic contrast targets

16

Sensitometry (CT number linearity)

17

CTP591 Bead Geometry Module

16

18

CTP528 High resolution module with 21 line pair per cm gauge and point source Bead point source for point spread function and MTF 19 Use of automated scanner MTF programs 20 Bead point source (slice sensitivity profile)

21

21 Line pair per centimeter high resolution gauge

22

CTP515 Low contrast module with supra-slice and subslice contrast targets CTP486 Image uniformity module

25

Installation and removal of test modules 27 Optional phantom annuli

28

Optional phantom housings

29

Dose Phantoms 30 Sample quality assurance program

31

Automated computer analysis program Bibliography

32

31

23

19

5 Introduction The Phantom Laboratory and physicist, David J. Goodenough, Ph.D., are continually developing and researching new tests and modifications for the Catphan® phantoms. The test objects that make up the current Catphan® models embody more than a quarter century of scientific evaluation and field experience. This manual outlines the applications of each module contained in the Catphan® 500 and 600 phantoms. We do not make specific recommendations on the content of your quality assurance program, because each medical imaging facility has its own unique set of requirements. A sample program is provided to give you ideas for possible program content. We suggest a review of local governing regulations, manufacturers’ specifications and the needs of your radiologists and physicists before developing your CT quality assurance program. The Catphan® instructional video, which illustrates the phantom setup and scanning of the different Catphan® sections, is also available.

If you have any additional questions please contact The Phantom Laboratory at: Phone: 800-525-1190 or 518-692-1190 Fax: 518-692-3329 email: [email protected] Additional product information is available at: www.phantomlab.com

6 Multi-Slice CT by David J. Goodenough, Ph.D. At the request of The Phantom Laboratory I have put together this summary for physicists who are familiar with CT image performance measurements and have not had experience with multi-slice CT scanners. Multi-slice uses the same basic approach to image reconstruction as axial single slice CT. Both modalities use the data from the detectors (positioned 360° around the patient) to reconstruct the axial patient images. The key difference between axial single slice CT and multi-slice is the axial image produced by single slice CT is developed from a single row of detectors, and the axial image made from a multi-slice scanner uses segments from several rows of detectors. With a multi-slice scan, as the patient moves through the gantry and the tube rotates around the patient, the detector rows utilized change as the patient and gantry move (see sketch on the next page). Additional variables in reconstruction result as the patient slice moves from one row of detectors to the next and the scanner reconstructs the images based on weighted averages between the relevant rows. In this way, multi-slice CT is analogous to spiral or helical single slice CT, but where the reconstruction is obtained from the combined slices rather than the interpolation between the readings of a single moving slice. Now add in focal spot variables and a little scatter to define in more detail the challenges and variables included in the reconstruction of a multi-slice image. Because in spiral mode each multi-slice image is reconstructed from an ensemble of data taken in different positions across the beam and from different detector rows, the overall image quality differences between images are minimal. In the spiral mode each slice represents data as seen from all detector rows in a sense a kind of averaging of detector row positions. However, if you use a multi-slice “step and shoot” mode, where each of the slices may be created from a single detector row (or rows depending on the selected slice thickness) with a consistent collimation, the differences between the slices will be evident. Step and shoot mode in a multi-slice CT scanner is operated like a conventional axial scanner by imaging with a fixed table position and then moving the table to the next position before imaging the next section of the phantom with a fixed table position. For example, with a step and shoot 8 slice scan it is expected that the middle slice numbers 4 and 5 will have better uniformity than outer slice numbers 1 and 8 because of the scanner x-ray beam geometry. However, if 1 and 8 or 4 and 5 are not similar, this may indicate a problem with the scanner. When assessing a scanner with a step and shoot mode, it is important to cover the full detector width with the selected test objects. If the test object is narrower than the slice, the table will need to be incremented between scanning sequences so the object can be scanned by all active rows of the detector. I recommend scanning through the entire phantom using different multi-slice spiral protocols for performance evaluations, as well as using the step and shoot approach for the bead ramps where slice geometry and the MTF can be measured for each slice and uniformity section where the signal to noise and uniformity of each slice (detector row) can be evaluated.

7

4 3 21 Gantry rotation ABCD

Detector rows

4 3 21

ABCD

4 3 21 ABCD

ABCD

4 3 21

This simplified illustration of a multi-slice sequence shows how the slices are reconstructed with information for different detector rows. The imaging sequence of the first selected slice (slice 1) of the patient begins when slice 1 moves over detector row A. As the tube continues to rotate and the patient continues to move through the gantry, slice 1 is picked up by the detectors in row B. At the same time slice 2, which was outside the detector view, is picked up by the detectors in row A. This sequence continues until the last selected region of the patient has passed through all the active detector rows.

8 Initial phantom positioning The Catphan® phantom is positioned in the CT scanner by mounting it on the case. Place the phantom case on the gantry end of the table with the box hinges away from the gantry. It is best to place the box directly on the table and not on the table pads. Open the box, rotating the lid back 180°. If you are using an annulus, additional weight will need to be placed in the box to counterweigh the phantom. The patient straps can be used for additional stability. Remove the phantom from the box and hang the Catphan® from the gantry end of the box. Make sure the box is stable with the weight of the phantom and is adequately counterweighed to prevent tipping. Use the level and adjusting thumb screws to level the Catphan®. Once the phantom is level, slide the phantom along the end of the box to align the section center dots on the top of the phantom with the x axis alignment light. Use the table height and indexing drives to center the first section’s (CTP401 or CTP404, Slice Geometry) alignment dots on the side and top of the phantom with the scanner alignment lights.

Center dots Adjusting thumb screws Counterweight if needed

Gantry

180° Level

Lateral height dot CTP401 Section one

The z axis scan alignment position can be selected from the localizer scan, by centering the slice at the intersection of the crossed wire image created by the slice width ramps. Scan the first section (CTP401 or CTP404) and check the image for proper alignment as illustrated in the Phantom position verification section.

9 Illustration of Catphan® 500 and 600 models

Catphan® 500 110mm 70mm

30mm

2.5mm 10mm CTP401

±.010"

CTP528 ±.010"

CTP515

±.010"

CTP486 ±.010"

Catphan® 600

160mm 110mm 70mm 32.5mm

CTP404

2.5mm 10mm ±.010"

CTP591 ±.010"

CTP528

CTP515

CTP486

±.010"

10 Incremental phantom module positioning The Catphan® phantoms are designed so all test sections can be located by precisely indexing the table from the center of section 1 (CTP401 or CTP404) to the center of each subsequent test module. This design eliminates the need to remount the phantom once the position of section 1 (CTP401 or CTP404) has been verified. The indexing distances from section 1 are listed below. Additional illustrations on the preceeding page show the test modules and their index spacing. Phantom position and alignment verification is described on the next page. Catphan® 500 test module locations: Module Distance from section 1 center CTP401 CTP528, 21 line pair high resolution 30mm CTP528. Point source 40mm CTP515, Subslice and supra-slice low contrast 70mm CTP486, Solid image uniformity module 110mm Catphan® 600 test module locations: Module Distance from section 1 center CTP404 CTP591 Bead geometry 32.5mm CTP528, 21 line pair high resolution 70mm CTP528. Point source 80mm CTP515, Subslice and supra-slice low contrast 110mm CTP486, Solid image uniformity module 150mm

11 Phantom position verification By evaluating the scan image of section 1 (CTP401 or CTP404) the phantom’s position and alignment can be verified. The section contains 4 wire ramps which rise at 23° angles from the base to the top of the module. The schematic sketches below indicate how the ramp images change if the scan center is above or below the z axis center of the test module. The use of the scanner’s grid image function may assist in evaluation of phantom position.

Correct alignment In this image the x, y symmetry of the centered ramp images indicates proper phantom alignment.

Clockwise ramp skew When the ramps are evenly rotated clockwise from center, the phantom is too far into the gantry.

Counter-clockwise ramp skew When the ramps are evenly rotated counter-clockwise from center, the phantom needs to be moved toward the gantry.

Non symmetrical ramp images Poor alignment with the z axis is indicated when the ramps are not symmetrical in lenghts and rotation.

If misalignment is indicated by the scan image, the phantom should be repositioned to obtain proper alignment and then rescanned. If the images of the repositioned phantom duplicate the original misalignment indications, the scanner’s alignment lights may require adjustment (contact your local service engineer). Once correct alignment has been established, you can proceed with the tests.

12 CTP401 Module with slice width, sensitometry and pixel size

23° ramps

Air Sensitometry samples

LDPE

Teflon

50mm spaced air and Teflon rods

10, 8, 6, 4, 2mm acrylic spheres

Acrylic

CTP404 Module with slice width, sensitometry and pixel size

23° ramps

Polystryene

Acrylic

50mm spaced air and Teflon rods Sensitometry samples

LDPE

Delrin™

Teflon

PMP

Air

10, 8, 6, 4, 2mm acrylic spheres

13 Patient alignment system check The laser, optical, and mechanical patient alignment system can be checked for accuracy. Align the white dots on the phantom housing with the alignment lights as discussed in Initial phantom positioning. The scanned image should show good alignment as discussed in Phantom position verification.

A A

A

A

For measuring the z axis alignment accuracy, measure from the center of the ramp image to the part of the ramp which aligns with the center of the phantom and sensitometry samples. Multiply the distance A by 0.42 to determine the z axis alignment light accuracy. To evaluate x and y accuracy, measure from the center of the phantom to the center of the scan field by use of the grid function or knowledge of the central pixel location. The accuracy of the localizer, pilot or scout view can be checked. To check this function perform a localization scan of the phantom. Align an axial scan at the crossing point of the wire ramps. Scan this axial cut and check the misalignment as discussed above.

14 Scan slice geometry (slice width) Section 1 has two pairs of 23° wire ramps: one pair is oriented parallel to the x axis; the other pair to the y axis. These wire ramps are used to estimate slice width measurements and misalignment errors as previously discussed.

Y

FWHM

FWHM

X Z

FWHM Measuring slice width with the 23° wire ramps. The ramp angle is chosen to offer trigonometric enlargement of 2.38 in the x-y image plane. To evaluate the slice width (Zmm), measure the Full Width at Half Maximum (FWHM) length of any of the four wire ramps and multiply the length by 0.42: (Zmm) = FWHM * 0.42 To find the FWHM of the wire from the scan image, you need to determine the CT number values for the peak of the wire and for the background. To calculate the CT number value for the maximum of the wire, close down the CT “window” opening to 1 or the minimum setting. Move the CT scanner “level” to the point where the ramp image just totally disappears. The CT number of the level at this position is your peak or maximum value. To calculate the value for the background, use the region of interest function to identify the “mean” CT number value of the area adjacent to the ramp. Using the above CTvalues, determine the half maximum: First calculate the net peak...

(CT # peak - background = net peak CT #)

Calculate the 50% net peak...

(net peak CT # ÷ 2 = 50% net peak CT #)

Calculate the half maximum CT number... (50% net peak CT # + background CT # = half maximum CT #)

15 Now that you have determined the half maximum CTnumber, you can measure the full width at half maximum of the ramp. Set the CT scanner level at the half maximum CT value and set your window width at 1. Measure the length of the wire image to determine the FWHM. Multiply the FWHM by 0.42 to determine the slice width.

L1

L2

Scan incrementation Schematic illustration of two sequential 5mm scans superimposed. L1 is the center point on the 23° ramp in the first scan image and L2 is the center point on the 23° ramp on the second image. Scan incrementation Use the wire ramps to test for proper scanner incrementation between slices, and for table movement. Scan section 1 using a given slice width, (e.g. 5mm). Increment the table one slice width (e.g. 5mm) and make a second scan. Establish the x and y coordinates for the center of each ramp image. Calculate the distance between these points and multiply by the 23° ramp angle correction factor of 0.42. 0.42(L1 - L2) = scan incrementation This test can also be used to test table increment accuracy. Scan the section and increment the table 30mm in and out of the gantry and scan again. The ramp centers should be the same on both images. 0.42(L1 - L2) = 0

16 Circular symmetry of display system The circular phantom sections are used to test for circular symmetry of the CT image, including calibration of the CT display system. If an elliptical image is produced, the x-y balance of the image display system should be adjusted. 150mm 50mm

X 150mm

50mm

Y

Measuring spatial linearity in x and y axes. Spatial linearity of pixel size verification This section has four holes (one with a Teflon pin). These 3mm diameter holes are positioned 50mm on center apart. By measuring from center to center the spatial linearity of the CT scanner can be verified. Another use is to count the number of pixels between the hole centers, and by knowing the distance (50mm) and number of pixels, the pixel size can be verified. The Teflon pin is used for identification and orientation only. The ability to change the Teflon pin position enables organizations with more than one Catphan® phantom to identify their phantoms by images of the first section. Spherical acrylic contrast targets The section has five acrylic spheres located in a 30mm diameter circular pattern. These spheres are used to evaluate the scanner’s ability to image volume averaged spheres. The sphere diameters are 2, 4, 6, 8, and 10mm.

17 Sensitometry (CT number linearity) Four or seven high contrast sensitometric targets surround the wire slice thickness ramps. Three are made from the commercial plastics: Teflon, acrylic and low density polyethylene (LDPE). The fourth is air. These targets range from approximately +1000 H to -1000 H. The monitoring of sensitometry target values over time and can provide valuable information, indicating changes in scanner performance. Linear attenuation coefficient µ [units cm-1] KEV Teflon Delrin Acrylic Polystryrene Water 40 0.556 0.327 0.277 0.229 0.240 50 0.447 0.283 0.244 0.209 0.208 60 0.395 0.260 0.227 0.196 0.192 62 0.386 0.256 0.224 0.194 0.190 64 0.380 0.253 0.221 0.192 0.188 66 0.374 0.251 0.219 0.191 0.186 68 0.370 0.248 0.217 0.189 0.184 70 0.363 0.245 0.215 0.188 0.182 72 0.359 0.243 0.214 0.186 0.181 74 0.355 0.240 0.211 0.185 0.179 76 0.351 0.238 0.210 0.184 0.178 78 0.346 0.236 0.208 0.183 0.177 80 0.342 0.234 0.207 0.180 0.175 90 0.328 0.225 0.199 0.175 0.170 100 0.315 0.218 0.194 0.170 0.165

LDPE PMP 0.209 0.189 0.191 0.173 0.181 0.164 0.179 0.162 0.178 0.160 0.177 0.160 0.175 0.158 0.174 0.157 0.172 0.155 0.171 0.155 0.170 0.154 0.168 0.152 0.167 0.151 0.163 0.147 0.158 0.143

Air 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Nominal material formulation and specific gravity Material Formula Specific Gravity* Electron Density (1023e/g) CT # est. Air 75%N, 23.2%O, 1.3%A 0.00 3.007 -1000 PMP [C6H12(CH2)] 0.83 3.435 -200 LDPE [C2H4] 0.92 3.429 -100 Water [H2O] 1.00 3.343 0 Polystryrene [C8H8] 1.05 3.238 -35 Acrylic [C5H8O2] 1.18 3.248 120 Delrin™ Proprietary 1.41 3.209 340 Teflon [CF2] 2.16 2.889 990 Contrast Scale (CS) is formally defined as CS =

µm (E) - µw (E) CTm (E) – CTw (E)

where m is reference medium, and w is water, and E is the effective energy of the CT beam. Alternatively, CS =

µ1 (E) - µ2 (E) CT1 (E) – CT2 (E)

where 1,2 are two materials with low z effective, similar to water (eg. acrylic & air). *For sensitometry inserts The Phantom Laboratory purchases a multiple year supply of material from a single batch. Samples of the purchased material are then measured to determine the actual specific gravity.

18 CTP591 Bead Geometry Module Ø 150.49mm 4 ramps 38mm (high) with 1mm increments and .028mm beads.

2 ramps 6mm (high) with 0.25mm increments and 0.18mm beads

Ø 0.28mm

Ø 0.28mm Ø 0.18mm

Ø 0.18mm

Ø 0.18mm

Ø 0.28mm

60mm

0.25mm

1mm

50µTungsten Wire

40mm

The Bead Geometry Module contains 3 pairs of opposed ramps and 2 individual beads. Two of the ramp pairs have 0.28mm diameter beads, spaced 1mm on center in the z direction. The other ramp pair has 0.18mm diameter beads, spaced 0.25mm on center in the z direction. The 2 individual beads are 0.28mm and 0.18mm in diameter. A 50µ diameter tungsten wire is located 6cm from the center of the module. The wire and beads create point spreads that can be used to calculate the MTF (see the CTP528 section of this manual). The bead ramps can be used to measure the slice width of single or multiple slices following several different methods. •count the beads and multiply by the z axis increment •plot the ssp of the beads (see CTP528 section for additional information) •compare the bead maximum net CT # in thick and thin slices. Thin slice thickness = (( thick slice max net CT#) ÷ (thin slice max net CT#)) * (thick slice width) •z axis length at the full width at half maximum of a bead in a sagital or coronal image •full width at half max for a best-fit curve of the max points of the bead net values in a slice image. Note: Net value = (CT# of the bead) - (CT# of the background)

19 CTP528 High resolution module with 21 line pair per cm gauge and point source

This section has a 1 through 21 line pair per centimeter high resolution test gauge and two impulse sources (beads) which are cast into a uniform material. The beads are positioned along the y axis 20mm above or below the phantom’s center and 2.5 and 10mm past the center of the gauge in the z direction. On older CTP528 modules the bead is aligned in the z axis with the gauge. Bead Point Source for point spread function and MTF Use the impulse source to estimate the point source response function of the CT system. Print out a digitized image of the area surrounding the impulse source. Use the numerical data to determine the two-dimensional array of the CT values arising from the impulse source. The FWHM of the point spread function is determined from the best-fit curve of the point spread function numerical data. The average of several different arrays of impulse response functions is calculated to obtain the average point spread function of the system. These numerical values are used in conjunction with the Fourier Transform Program to provide an estimate of the two-dimensional spatial frequency response characteristics of the CT system (MTF). Illustration is on the next page. The tungsten carbide bead has a diameter of 0.011” or 0.28mm. Because the bead is subpixel sized it is not usually necessary to compensate for its size. However, some MTF programs are designed to compensate for it.

20

Line spread function

PSF -2

-3

-2

0.5

-4

3

17

3

-4

-2

44

100

44

-2

-2

44

100

44

-2

-4

3

17

3

-4

0.5

-2

-3

-2

0.5

90

-11

228

CT numbers

0.5

114

LSF -11

90

228

0 Relative position, x axis

The above illustration shows how by summing the columns (y axis) of numbers in the point spread function (PSF) the line spread function (LSF) for the x axis is obtained. 1.0

0.9 0.8

Average MTF 50% 10% 2%

Cycles/cm 3.84 6.65 9.29

9.0

12.0

0.7

MTF

0.6 0.5

0.4

0.3 0.2

0.1 0.0

0.0

3.0

6.0

Spatial Frequency (1/cm) The MTF curve results from the Fourier transform of the LSF data. Generally it is easiest to use automated software for this operation. Some CT scanners are supplied with software which can calculate the MTF from the Catphan® bead images. Independent software is listed in the Current automated programs available section of the manual. Use of automated scanner MTF programs Many manufacturers include automated MTF software in the standard scanner software packages. Because the bead is cast into an epoxy background which has a different density than water, the software must accept an input for the background. The point size of .28mm must also be selected. While a sphere does produce a different density profile than a cross section of a wire or cylinder, the actual difference is not usually significant in current CT scanners.

21 Bead point source for slice sensitivity profile The bead in this module can be used to calculate the slice sensitivity profile (SSP). Z

Z

Y

SSP(z)

SSP(z)

Z

Z

X

Y

X

3mm Spiral

10mm Spiral

The above image illustrates how the bead will produce an ovoid object in a 3 dimensional reconstruction. The length of the object at the Full Width at Half Maximum signal indicates the SSP. This measurement can be easily obtained on some systems, by making a sagittal or coronal reconstruction through the bead. The bead image in these reconstructions will appear as a small line. By setting the FWHM (use the same technique described in the Scan slice geometry section) measuring the z axis length of the bead image to obtain the SSP. If the scanner does not have the ability to measure z axis lengths in the sagittal or coronal planes, a SSP can be made by incrementing or spiraling the slice through the bead and reconstructing images in positive and negative table directions from the bead (using the smallest available increments) and plotting the peak CT number of the bead image in each slice. The FWHM measurement can then be made from the plotted CT values of the bead as a function of z axis table position. CT# 300 250

200

150

FWHM 100

50

0 -10

-8

-6

-4

6 4 2 0 z axis position in millimeters -2

8

10

22 21 Line pair per centimeter high resolution gauge The 21 line pair/cm gauge has resolution tests for visual evaluation of high resolution ranging from 1 through 21 line pair/cm. The gauge accuracy is ± 0.5 line pair at the 21 line pair test and even better at lower line pair tests. The gauge is cut from 2mm thick aluminum sheets and cast into epoxy. Depending on the choice of slice thickness, the contrast levels will vary due to volume averaging.

Line Pair/cm

Gap Size

Line Pair/cm

Gap Size

1

0.500 cm

11

0.045 cm

2

0.250 cm

12

0.042 cm

3

0.167 cm

13

0.038 cm

4

0.125 cm

14

0.036 cm

5

0.100 cm

15

0.033 cm

6

0.083 cm

16

0.031 cm

7

0.071 cm

17

0.029 cm

8

0.063 cm

18

0.028 cm

9

0.056 cm

19

0.026 cm

10

0.050 cm

20

0.025 cm

21

0.024 cm

Gap