Presentation is loading. Please wait.

Presentation is loading. Please wait.

Seeram Chapter 11: Image Quality

Similar presentations


Presentation on theme: "Seeram Chapter 11: Image Quality"— Presentation transcript:

1 Seeram Chapter 11: Image Quality
CT Seeram Chapter 11: Image Quality

2 CT Image Quality Parameters
Spatial Resolution Contrast Resolution Image Noise Artifacts

3 Factors Influencing CT Image Quality
Beam Characteristics Subject Transmissivity Dose Slice Thickness Scatter Display Resolution Reconstruction Algorithm

4 Spatial Resolution Quantifies image blurring
“Ability to discriminate objects of varying density a small distance apart against a uniform background” Minimum separation required between two high contrast objects for them to be resolved as two objects

5 Spatial Resolution

6 Resolvable Object Size & Limiting Resolution
Smallest resolvable high contrast object Often expressed as line pairs / cm “Pair” is one object + one space One Pair

7 Resolvable Object Size: Limiting Resolution
Smallest resolvable high contrast object is half the reciprocal of spatial frequency Example: Limited resolution = 15 line pairs per cm Pair is 1/15th cm Object is half of pair 1/15th / 2 1/30th cm .033 cm 0.33 mm 1/15th cm 1/30th cm

8 Geometric Factors affecting Spatial Resolution
Focal spot size detector aperture width slice thickness or collimation Less variation likely for thinner slices attenuation variations within a voxel are averaged partial volume effect

9 Geometric Factors affecting Spatial Resolution
Finite focal spot size focal spot - detector distance focal spot - isocenter distance

10 Geometric Unsharpness & CT
Decreased spatial resolution if object blurred over several detectors Detector aperture size must be < object for object to be resolved Focal Spot Small Object to be Imaged Detectors

11 Non-geometric Factors affecting Spatial Resolution
# of projections Display matrix size 512 X 512 pixels standard Reconstruction algorithms smoothing or enhancing of edges

12 Reconstruction Algorithm & Spatial Resolution
Back projecting blurs image Algorithms may be anatomically specific Special algorithms edge enhancement noise reduction smoothing soft tissue or bone emphasis

13 Hi-Resolution CT Technique
Very small slice thicknesses 1-2 mm High spatial frequency algorithms increases resolution increases noise Noise can be offset by using higher doses Optimized window / level settings Small field of view (FOV) Known as “targeting”

14 Contrast Resolution Ability of an imaging system to demonstrate small changes in tissue contrast The difference in contrast necessary to resolve 2 large areas in image as separate structures

15 CT Contrast Resolution
Significantly better than radiography CT can demonstrate very small differences in density and atomic # This’ll be on your test. I guarantee it. Contrast Resolution Radiography 10% CT <1%

16 CT Contrast Resolution Depends Upon
reconstruction algorithm low spatial frequency algorithm smooths image Loss of spatial resolution Reduces noise enhances perceptibility of low contrast lesions image display

17 CT Contrast Resolution Depends on Noise

18 CT Contrast Resolution
Contrast depends on noise Noise depends on # photons detected # photons detected depends on …

19 # of Photons Detected Depends Upon
photon flux (x-ray technique) slice thickness patient size Detector efficiency Note: Good contrast resolution requires that detector sensitivity be capable of discriminating small differences in intensity

20

21 Small Contrast Difference Harder to Identify in Presence of Noise

22 CT Image Noise Fluctuation of CT #’s in an image of uniform material (water) Usually described as standard deviation of pixel values

23 CT Image Noise Standard deviation of pixel values S(xi - xmean)2
Xi = individual pixel value Xmean = average of all pixel values in ROI n =total # pixels in ROI

24 Noise Level Units CT numbers (HU’s) or % contrast

25 Noise Measurement in CT
Scan water phantom Select regions of interest (ROI) Take mean & standard deviation in each region Standard deviation measures noise in ROI

26 CT Noise Levels Depend Upon
# detected photons quantum noise matrix size (pixel size) slice thickness algorithm electronic noise scattered radiation object size Photon flux to detectors…

27 Photon Flux to Detectors
Tube output flux (intensity) depends upon kVp mAs beam filtration Flux is combination of beam quality & quantity Flux to detectors modified by patient Larger patient = less photons to detector

28 Slice Thickness Thinner slices mean
less scatter better contrast less active detector area less photons detected More noise To achieve equivalent noise with thinner slices, dose (technique factors) must be increased

29 Noise Levels in CT: Increasing slice width = less noise BUT
Increasing slice width degrades spatial resolution less uniformity inside a larger pixel partial volume effect

30 CT Image Quality in Equation Form
s2(m) = kT/(td3R) Where s is variance resulting from noise k is a conversion factor (constant) T is transmissivity (inverse of attenuation) t is slice thickness d is pixel size R is dose

31 Noise Levels in CT: When dose increases, noise decreases
dose increases # detected photons Doubling spatial resolution (2X lp/mm) requires an 8X increase in dose for equivalent noise Smaller voxels mean less radiation per voxel

32 CT Image Quality Trade-off
s2(m) = kT/(td3R) To hold noise constant If slice thickness goes down by 2 Dose must go up by 2

33 Measurements of Image Quality
PSF = Point Spread Function LSF = Line Spread Function CTF = Contrast Transfer Function MTF = Modulation Traffic Function

34 Point Spread Function PSF
“Point” object imaged as circle due to blurring Causes finite focal spot size finite detector size finite matrix size Finite separation between object and detector Ideally zero Finite distance to focal spot Ideally infinite

35 Quantifying Blurring Object point becomes image circle
Difficult to quantify total image circle size difficult to identify beginning & end of object Intensity ?

36 Quantifying Blurring Full Width at Half Maximum (FWHM)
width of point spread function at half its maximum value Maximum value easy to identify Half maximum value easy to identify Easy to quantify width at half maximum Maximum Half Maximum FWHM

37 Line Spread Function LSF
Line object image blurred Image width larger than object width Intensity ?

38 Contrast Response Function CTF or CRF
Measures contrast response of imaging system as function of spatial frequency Lower Frequency Higher Frequency Loss of contrast between light and dark areas as bars & spaces get narrower. Bars & spaces blur into one another.

39 Contrast Response Function CTF or CRF
Blurring causes loss of contrast darks get lighter lights get darker Lower Frequency Higher Frequency Higher Contrast Lower Contrast

40 CT Phantoms Measure spatial resolution noise contrast resolution
Available from CT manufacturer private phantom manufacturers American Association of Physicists in Medicine AAPM Measure noise spatial resolution contrast resolution slice thickness dose

41 CT Spatial vs. Contrast Resolution
Spatial & contrast resolution interact High contrast objects are easier to resolve Omprove one at the expense of the other Can only improve both by increasing dose Increasing object size Increasing contrast

42 Increasing object size
Contrast & Detail Larger objects easy to see even at low contrast Increasing object size Increasing contrast

43 Increasing object size
Contrast & Detail Small objects only visible at high contrast Increasing object size Increasing contrast

44 Contrast – Detail Relationship
Contrast vs. object diameter less contrast means object must be larger to resolve Visibility Increasing object size Difference in CT # Object Diameter Increasing contrast

45 Modulation Transfer Function MTF
Fraction of contrast reproduced as a function of frequency Freq. = line pairs / cm 1 MTF 50% Recorded Contrast (reduced by blur) Contrast provided to film frequency

46 MTF Can be derived from MTF = 1 means MTF = 0 means
point spread function line spread function MTF = 1 means all contrast reproduced at this frequency MTF = 0 means no contrast reproduced at this frequency

47 MTF If MTF = 1 all contrast reproduced at this frequency
Contrast provided to film Recorded Contrast

48 MTF If MTF = 0.5 half of contrast reproduced at this frequency
Contrast provided to film Recorded Contrast

49 MTF If MTF = 0 no contrast reproduced at this frequency
Contrast provided to film Recorded Contrast

50 CT Number Calculated from reconstructed pixel attenuation coefficient
(mt - mW) CT # = 1000 X mW Where: ut = linear attenuation coefficient for tissue in pixel uW = linear attenuation coefficient for water

51 Linearity Linear relationship of CT #’s to object linear attenuation coefficients Checked with phantom of several known materials average CT # of each material obtained from ROI analysis Compare CT #’s with known coefficients 77 -100 325 50 -44

52 CatPhan

53

54 Cross-Field Uniformity
Use uniform phantom (water) CT pixel values should be uniform anyplace in image Take 5 ROI 1 center ROI 4 corners ROI’s Compare standard deviation between ROI’s

55 CT Artifacts Distortion
Areas where image not faithful to subject Sources patient image process equipment

56 CT Artifacts Distortion
Phantoms with evenly distributed objects

57 Preview! CT Artifacts: Causes
motion metal & high-contrast sharp edges beam hardening partial volume averaging sampling detectors

58 Motion Artifacts Causes streaks in image
Algorithms have trouble coping because of inconsistent data

59 Artifacts: Abrupt High Contrast Changes
Examples: prostheses dental fillings surgical clips Electrodes bone Metal absorbs all radiation in ray causes star-shaped artifact Can be reduced by software

60 CT Artifacts: Beam Hardening
Increase in mean energy of polychromatic beam as it passes through patient Can cause broad dark bands or streaks cupping artifact Reduced by beam hardening correction algorithms

61 CT Artifacts: Partial Volume Effect
CT #’s based on linear attenuation coefficient for tissue voxels If voxel non-uniform (contains several materials), detection process will average

62 Partial Volume Effect Can appear as Minimizing incorrect densities
streaks bands Minimizing Use thinner slices

63 Image Artifacts: Ring Artifact in 3rd Generation
Causes 1 or more bad detectors small offset or gain difference of 1 detector compared to neighbors detector calibration required Reason: rays measured by a given detector are all tangent to same circle

64 Quality Control in CT spatial resolution contrast resolution noise
Performance tested at prescribed intervals Image Quality Tests spatial resolution contrast resolution noise slice width kVp waveform average & standard deviation of water phantom CT # scatter & leakage


Download ppt "Seeram Chapter 11: Image Quality"

Similar presentations


Ads by Google