Software and Hardware in CCTA. Elly Castellano PhD

Transcription

1 Software and Hardware in CCTA Elly Castellano PhD

2 Outline technical requirements for coronary CTA the modern cardiac CT scanner ECG-gating technology image reconstruction algorithms 2

3 Technical requirements for coronary CTA 3

4 Technical requirements for coronary CTA ECG-gating capability freezes cardiac motion high temporal resolution reduces blurring due to cardiac motion high contrast and spatial resolution enhances conspicuity of narrow vessels large detector coverage per rotation facilitates short scan times without ECG gating images courtesy of Willi Kalender with ECG gating 4

5 ECG gating two physiological windows of opportunity HR < 65 HR > 65 R-R interval 100% V V t t ~ 65-75% ~ 25-35% end diastole end systole choice in timing and width of acquisition window to capture one / several / all phases of cardiac cycle 5

6 Temporal resolution determined by time taken to acquire projection data rotation time 180º+ or 360º scan conventional CT or DSCT at 60 bpm require < 200 ms possible with 0.4 s rotation time and 180º+ scan 180º+ scan 360º scan 6

7 determined by Spatial resolution in-plane: reconstruction kernel out-of-plane: slice thickness require sharp reconstruction kernels increase spatial resolution, but increase image noise < 1 mm slices increase spatial resolution, but increase image noise Corresponding CT and conventional angiogram images of RCA 7

8 Contrast resolution determined by difference in HU between arteries and background kv setting partial volume effects image noise require contrast enhancement low kv to increase sensitivity to iodine, but image noise and likelihood of photon starvation increase high maximum tube ma for use with lower kv settings < 1 mm slices to reduce partial volume effects, but image noise increases 8

9 determined by Coverage per rotation width of detector array with thin detector elements require > 30 mm of sub-mm detector rows 4-5 rotations to cover heart all data acquisition during optimal contrast enhancement image-stitching artefacts reduced 9

10 The modern cardiac CT scanner 10

11 The basic model: cardiac-enabled 64-DCT scanners* supplier example scanner model detector array rows x element mm z-coverage mm slices per rotation x thinnest slice mm minimum rot s GE Optima CT x x Siemens Definition Edge 64 x x Philips Brilliance x x Toshiba Aquilion CXL 64 x x NB manufacturers with z-ffs or detector resampling double number of slices per rotation * NB scanner specifications are continually changing 11

12 Top MDCT scanners 12

13 Revolution series: Rev CT 512-slice scanner 256 detector rows with resampling improved anatomical coverage 160 mm detector covers heart improved spatial resolution new x-ray tube HD mode using xy-ffs improved contrast resolution Gemstone Clarity detector in 3 sections increased detector efficiency lower detector afterglow image courtesy of GE 13

14 Brilliance ict / ict Elite 256-slice scanner 128 detector rows with z-ffs improved anatomical coverage 80 mm detector faster rotation time 0.27 s with AirGlide technology and imrc tube images courtesy of Philips 14

15 Definition Force 2 imaging chains mounted at ~94º 50 and 35 cm FOV improved temporal resolution faster acquisition of projection data set improved anatomical coverage 96 detector rows, 57.6 mm Turbo Flash mode high pitch helical mode image courtesy of Siemens 15

16 Aquilion One ViSION 640-slice scanner 320 detector rows with double sampling improved anatomical coverage 160 mm detector covers heart volume mode no table feed image courtesy of Toshiba 16

17 Top MDCT scanners: summary* supplier scanner model detector design rows x element mm z-coverage mm slices per rotation x thinnest slice mm minimum rot s GE Siemens Revolution CT Definition Force 256 x x x 0.6 x x Philips ict / Elite 128 x x Toshiba Aquilion One Vision 320 x x NB manufacturers with z-ffs or detector resampling double number of slices per rotation *from manufacturers information. NB scanner specifications are continually changing 17

18 ECG gating technology 18

19 ECG gating of acquisition essentially two types: prospective retrospective different image sets generated different radiation burden to patient 19

20 Prospective gating acquisition triggered by ECG set time after R wave acquisition window matched to minimum acquisition time no padding acquisition image sets in selected phase acquisition window extended beyond minimum acquisition time with padding acquisition at full or reduced ma image sets in 1+ selected phases can adjust for heart rate changes ma ma prescribed mas prescribed mas t t padding mas 20

21 Prospective gating choice of table motions step-and-shoot slow helical high-pitch helical (Flash) none axial images reconstructed from partial projection data sets MPRs, VR image sets patient dose 1 5 msv CXRs partial projection data set 21

22 Prospective gating Flash helical mode LAD with stent Image courtesy of Universität Ulm, Germany, and Philips images courtesy of Institute of Medical Physics / Erlangen, Germany, and Siemens 22

23 Retrospective gating long acquisition fast rotation time choice of table motion slow helical none pitch typically projections tagged with ECG signal ECG-gated ma pulsing selectable temporal window position and width restricts reconstruction to selected cardiac phase(s) 23

24 Retrospective gating images reconstructed from partial projection data set to improve temporal resolution single or multi-segment recon determines temporal resolution reconstruction window can be adjusted essential for arrhythmic patients typical patient dose msv CXRs single-segment recon multi-segment recon 24

25 Retrospective gating coronary CTA from images reconstructed at one cardiac phase cardiac function from images reconstructed at several cardiac phases Functional Analysis, EF 80% images courtesy of Dr Rybicki, BWH, and Toshiba 25

26 Dealing with irregular heart rates arrhythmia detection algorithms acquisition delayed until rhythm re-established compensation for heart rate changes acquisition window automatically widened to capture required phase(s) may impact on radiation dose to patient 26

27 Image reconstruction algorithms 27

28 Filtered back-projection choice of convolution kernels images courtesy of Willi Kalender 28

29 Iterative reconstruction from Beister et al

30 IR: vendor solutions supplier IR algorithms GE ASiR Veo ASiR-V Siemens SAFIRE ADMIRE Philips idose4 IMR Toshiba AIDR 3D proprietary black boxes! 30

31 IR solutions in raw data and image domains used by Philips (idose), Siemens and Toshiba based on statistical modelling of x-ray detection from Willemink et al

32 IR solutions in raw data domain used by GE, Philips (IMR) based on statistical modelling of x-ray detection geometric modelling of imaging chain (Veo) physics modelling (ASiR-V) from Willemink et al

33 Iterative reconstruction images presented different IR strengths mixed with FBP lower image noise than FBP opportunity for dose reduction FBP SAFIRE images courtesy of Siemens 33

34 Iterative reconstruction slower recon than FBP IR-specific artefacts may be introduced disappearance of hypodense seam from Beister et al

35 Conclusions coronary CT angiography requires cardiacenabled scanners 64+ DCT choice of ECG gating techniques iterative reconstruction algorithms should be carefully validated 35

36 Acknowledgements Dee Mears, RMH Nina Arcuri, RBH Willi Kalender ImPACT group Philips Siemens Toshiba GE 36