Principles of CT scanning • Alan Britten • From slides by Jim Weston, Im. PACT, • St George’s Hospital Im. PACT
Principles of CT scanning • • Basic CT Principles Displaying the image Helical CT Multi-slice CT Im. PACT
In practice Cables Aperture / bore X-Ray Tube Fan beam Detectors Im. PACT
CT image • Cross-sectional image – what is it showing? – how do we get the image? Im. PACT
Data acquisition • Measure attenuation of the xray beam – Mainly compton interactions x-ray tube eattenuation • Relates to the linear attenuation coefficient (µ) ‘summed’ along the path between tube and detectors Im. PACT detector
Data acquisition • Take many attenuation measurements • Sample at each detector position to generate a profile • Gives view of patient – at one orientation attenuation PROJECTION Im. PACT Detector position
Data acquisition • Collect projections from all around patient & ~1000 projections 90° attenuation ~1000 positions attenuation detector position – rotate tube and detectors a small amount and repeat the measurements 0° detector position Im. PACT
Filtered back projection (See SPECT talk slides) 2 projections 8 projections 64 projections (not filtered) Im. PACT 4 projections 64 projections 16 projections
Filtered back projection original object Im. PACT back projected image filtered back projected image
Iterative Reconstruction Iterative reconstruction is now being used in X-ray CT Potential to reduce radiation dose and maintain image Quality, with system modelling. (See slides on SPECT for explanation of Iterative reconstruction) Im. PACT
CT number • Original measure m = linear attenuation coefficient • Values are normalised to give the CT number – fixed points: water & air HOUNSFIELD UNITS (HU) water = 0 air = -1000 bone ≈ +1000 Im. PACT
CT image • Cross-sectional image – what is it showing? • Linear attenuation coefficient average within a voxel – how do we get the image? • reconstruct as filtered back projection, or iterative reconstruction Im. PACT
Scanning a volume • “Axial scans” or “sequential scanning” • “Step and Shoot” Im. PACT
Helical scanning • • Also known as spiral scanning Introduced in late 1980 s Continuous rotation Continuous table feed Im. PACT
Helical CT data collection • With constant rotation and table feed, scan data forms a helical path Power Data Im. PACT
Helical pitch • Pitch table travel per rotation x-ray beam width Travel Beam Pitch Im. PACT = 10 mm/rot = 10 mm =1 Travel Beam Pitch = 20 mm/rot = 10 mm =2
Data sets – axial scans • Single plane scanned • Projection sets all in same plane • Reconstruction plane location fixed by scan plane Im. PACT
Data sets– helical scans • Length scanned • Projection sets not in same plane • Reconstruction plane not dictated Im. PACT
Advantages of helical scanning • Speed – – no need to pause between scans for table movement longer scan lengths possible within single breath-hold pitches greater than 1 possible reduced patient movement artefacts • Flexibility of reconstruction – any position – any interval • overlapping reconstructions possible Im. PACT
MULTI SLICE CT Im. PACT
Multi-slice CT • Dual slice – Elscint (1991) • Four slice – GE, Philips, Siemens, Toshiba (1998) • Sixteen slice – GE, Philips, Siemens, Toshiba (2002) • 32, 40, 64 slice … – and beyond. . . Im. PACT
Multi slice scanning • Multiple parallel rows of detectors Multi slice • For single irradiation width, four simultaneous data sets acquired – 4 axial images – 4 helical sets of data – (or 8, or 16, or …) Banks of detectors Im. PACT
Detector arrays • Not all detectors have to be the same size – eg 16 slice 24 mm Courtesy: Siemens 4 x 1. 5 Im. PACT 16 x 0. 75 4 x 1. 5 z-axis
Helical MSCT • Analogous to the move from axial to helical for single slice • But new factors to consider Im. PACT
Pitch in MSCT • Pitch can be defined relative to total x-ray collimation (x), or individual detector width (d) • In practice, pitchx is useful when looking at dose, pitchd for image quality • Standard definition (IEC) Pitchx = table travel x-ray beam width (x) Also Pitchd = Im. PACT table travel detector width (d) x d
Cone beam artefacts • As number of slices increases, beam is less planar – cone shaped Cone beam reconstruction methods used Im. PACT
Basic CT Quality Control Im. PACT
Image noise Scanner’s water phantom • • • CT 01 Measured daily – weekly 40 % of phantom diameter Standard deviation Central slice Im. PACT Noise image
CT number values • Draw a ROI inside each different material • ROI should be entirely within the material • Use approximately the same size ROI each time • Note the mean CT number of each ROI Im. PACT
Measurement of spatial resolution • Standard deviation technique – ROI method 100 90 80 MTF (%) 70 60 50 40 30 20 10 0 0 2 4 6 Frequency (lp/cm) 8 Droege RT & Morin RL, ‘A practical method to routinely monitor the MTF of CT Scanners’, Medical Physics 9(5): 758 -780 Im. PACT 10
Scan plane localisation from alignment lights • Now a 1 – 3 monthly test • Test alignment of light to xray beam – at the start of your quality control session • helps subsequent set up of phantoms – important clinically to have accurate alignment • for example if setting up patient with angled gantry to avoid eyes • Can easily become misaligned during service Im. PACT
CT QC: dose check • CT Dose Index = CTDI • Ionisation chamber in air, or in a standard Perspex cylinder CTDI dose in a perspex Phantom Im. PACT
CT QA – for SPECT/CT & PET/CT • Standard CT QC – daily noise – quarterly CTDI check • CT number accuracy Noise CTDI Dose check • Geometrical alignment • Attenuation correction accuracy with clinical protocols Im. PACT CT Number accuracy: Need to check for low dose and Large patients CT Number Accuracy
PET/CT WB Phantom Scan: Clinical Image Quality test • 2 bed position acquisition • Clinical image protocol • Inspect image for: – Uniformity per slice – Continuity across slices – Visibility of slice overlap Im. PACT
QC for SPECT/CT or PET/CT systems • Routine checks (noise, CT number accuracy) • Geometrical alignment of SPECT & CT/PET images • Check of accuracy of correction – – head and body sized water phantoms – Is the CT attenuation correct? Homogeneous phantom – AC Emission values constant across the phantom? Im. PACT
