iMinds The Conference: Johan de Mey

• 1.Redefining the Rules of clinical CTImaging with Spectral Imaging and Iterative Reconstruction techniquesJohan de Mey PhD. MD Head of Radiology DepartmentKoenraad Nieboer MDHead of Emergency RadiologyNico Buls PhD. MSc Medical Imaging and Physical SciencesGert Van Gompel PhD. MSc Medical Imaging and Physical SciencesToon Van Cauteren MScMedical Imaging and Physical Sciences

2. Multislice CT and clinical outcome What makes a good CT?accuracy SpatialResolution Temporal CoverageperfusionWorkflow Dose Characterization 3. Low Dose CT = quality ? 1980 2007: Dose optimalisationMore imagesHigher resolution 2008 - :Dose optimalisationLess dose??? Functional???Iterative recon Dual energy 4. Noise suppressionThe straightforward wayincrease # photons, increase patient dose (mAs) 10 mAs 60 mAs120 mAsCTDIv = 0,8 mGy CTDIv = 5 mGy CTDIv = 10 mGy 5. Effective doses for CT procedures(review over 20 published articles) 6. Does the age of the patient at time ofexposure affect the patient risk?Children are 2-10 x more sensitive!Hall Pediatric Radiology Apr 2002 pg 226 7. Iterative reconstructiondetector s source sourceImage acquisitionsinogram Goal:reconstruction 8. Adaptive Statistical Iterative Reconstruction(0%.....100%) and noise45403530noise2520 0.8 mGy15105 16.8 mGy0 0 20406080 100asir (%) GE CT750 HD, 100 kVp, p = 0.9, r = 1, t = 2.5 mm, FOV = 230 mm 9. ASIR (60%) and Spatial resolution FBP1817161514 FBP13 ASIR1211 ASIR 10 9 -6 -4 -20 2 4 6 10. Female 45 Acute dyspnea, suspicion pulmonaryembolism Scan Parameters Range 26 cmNi: 30 120KVASiR: 40% CTDIv 7.44 (mGy) E 2.7 (mSv) 0.625 mm 1.25 mm1.25 mm 11. Body packers: low dose CT / GSIECR EPPOS: C-2068 ECR 2011 K. H. Nieboer, N. Buls, J. de Mey, G. Van Gompel; Brussels/BE, The use of iterativereconstruction in ultra low dose computed tomography for bodypacker screening 12. VeoTM future in dose reduction= Model Based Iterative Reconstruction REAL 3D SYSTEM OPTICS SYSTEM NOISESTATISTICS 13. FBP ASIR VEO 14. Iterative impact on image noise100 90N80FBPo70Relative Noise (SD)i60s50e40 ASIR50 30 20 10VEO00 50 100 150200 250300350 400 450 500 tube current - relative radiation dose Tube current 15. Resolution at standard doseCatphan 504 phantomCTDIvol = 12.5 mGy FBPASIR 50% VEO Performance at 6 lp/cm and 8 lp/cm 16. Ultra low dose Chest follow-up with VEO4mAs, 0.06mSv*FBPASiR VeoTypical CXR effective dose is about 0.06 mSv.Source: Health Physics Society.http://www.hps.org/publicinformation/ate/q2372.html * Determined by internal organ dosimetry on a humanoid phantom 17. Chest CT with Veo Cystic Fibrosis 26y CDTIvol = 0.10 mGy Effective dose = 0.05 mSv* * Obtained by EUR-16262 EN, using a chest factor of 0.017*DLPConversions of CTDI or DLP to effective dose are only rough estimations for children 18. Chest CT with Veo Cystic Fibrosis 15yFBP- Images 2010 FBP-images 2011Veo-images 2011Low dose Ultra low dose Ultra low dose DLP = 54,3 mGy.cmDLP = 2.51 mGy.cm CTDIvol: 1,65 mGyCTDIvol:0,09 mGy Scan protocol: 4 mAs, 80 kV Slice thickness: 0.625mm 19. Chest CT with Veo Cystic Fibrosis 15yFBP reconstruction 2010 VEO reconstruction 201195 % dose reduction 20. Chest CT with Veo Cystic Fibrosis 15yPrevious Chest X-rayVeo-images: Ray Sum CT 21. PATIENT 1 (2010) PATIENT 2 (2011)3Y Old, empyema3 y old, empyemaFBP-images Veo-imagesCTDi 2.1 CTDi 1.25 40% dose reduction 22. Pediatric maxilo-facial CT with Veo 95% dose reduction CT at plain film dose +/- 0.06 mSv1. 70% dose reductionNo quality loss1. 50% dose and low KVpNo quality loss + 50 % contrast reduction1. 40% dose reductionBetter image, less artifacts1. Equal dose and low KVp Up to 60% contrast reduction 30. Dual energyspectral imaging:Need for at least two datasets on different energy levelDiscovery CT750 HDDual tube Fast switching tubeMaking two scansDual layer detectorPhoton counting 31. Dual energyspectral imaging:Need for at least two datasets on different energy levelDiscovery CT750 HDDual tubeFOV max 33 cmSpectral filter on 140Making two scansTime Difference FOV 50 cmDual layer detectorResearch Real TimePhoton countingFuture ? 32. Photon energy with Tungsten? What is dual energy?120 KV scanMean 61.1 KeV80 KV scan Mean 49.9 KeV 140 KV scan Mean 66.1 KeV 33. Spectral material differentiation 34. From 2 datasets to 101 datasets? Is this real?Phantom dataQualitative comparison DECT literature (NIST)ECR: B-851 Monday 14.00G. Van Gompel, N.Buls, K. Nieboer, J. de MeyAccuracy estimation of spectral attenuation curves obtained by dual energy 35. Clinical use?Cadaver blood vessels Filled with CM concentration, GSI scan ROI2: water + CM ROI3: soft plaque ROI4: calcified plaque 36. Relative dose reduction? (No need for non-CE scans) Contrast reduction? (Higher contrast in image) Iodine Water 250CNRD @ iodineconc 1mgI/ml 2001 mg I / ml2 1.8 150 1.6120 1.4 CT value (HU)kV 1.2CNRD 100 1GSI v2 0.8 0.6 500.4 0.2 00 40 60 80 100 120 140 30 50 7090 110130Photon energy (keV) energy (keV) 37. Pleural effusion: Exudate >< transudate? ECR EPOS: C-1996 Y. De Brucker, N. Buls, G. Van Gompel, F. Vandenbroucke, H. Nieboer, T. De Keukeleire, J. de Mey; Brussels/BE Characterization of pleural effusion using dual energy CT: feasibility study. 38. Phantom size on quantitative iodine measurement:dual energy CT >< standard CT.Obtained CT-values (HU) 120 keV GSI 70 keV 23 cm291 9 310 6 17 cm325 5315 8 17 cm + t318 9 314 5range 34 5 VAR10.9%1.6% 39. StandardPhantom size on quantitative iodine measurement:spectral CT >< standard CT. Standard Spectral 120 kV DECT at 70 keV 40. Male 77 Acute dyspnea, suspicionpulmonary embolismAxial recon imageMonochromatic 40 140 KeV Iodine imageWater image 41. Water imageIodine image Scatterplot 42. S GSI: spectral imagingSubsegmental pulmonary embolismClinical value Pulmonary embolism was not Embolism depicted during 1st reading onstandard images MD Iodine shows clearly a hypoperfusion on the left. 2nd reading confirms pulmonaryembolism resulting from a smallthrombus in a small pulmonaryvessel GSI allows physicians todetect pulmonary embolism byshowing subtle lung perfusiondefect 43. Pulmonary EmbolismClinical value Pulmonary embolism wasnot depicted during 1streading on standard images MD Iodine shows clearly ahypo perfusion in left lungresulting from a pulmonaryembolism. Hypo perfusioncould not be seen onstandard CT imageIodine Based Images 70 keV images with lung window 2nd reading confirmspulmonary embolism resultingfrom a small thrombus in asmall pulmonary vessel GSI allows physicians todetect pulmonary embolismby showing subtle lungperfusion defect70 keV Monochromatic Images70 keV Monochromatic Images Images Courtesy of Dr. Jean Louis Sablayrolles; Centre Cardiologique du Nord 44. Standard 120 kV acquisitionGSI: Artefact reduction Monochromatic with MARs 45. Iliac Stent Bleeding Clinical value Endoleak is better visualized on low energy (55 keV) images (green Arrows) 80 keVMD Water (Iodine) MD Iodine suggests Iodine leak GSI allows physicians to better visualize bleeding vs. conventional CTMD Iodine (Water)55 keV Images Courtesy of Pr Dacher, CHU de Rouen 46. Metal Artifact Reduction: Spine GSI acquisition allows tocreate a spectrum ofmonochromatic imageswhere beam hardening ishighly reduced In this case the radiologistcould recover informationpreviously hidden by thebeam hardening 70 keV image 110 keV images Images Courtesy of Dr. Sablayrolles, CCN, France 47. Conclusion: New dimensions in CT Statistical iterative reconstruction: important dose reduction- Dose reduction up to 50% and same quality- Scanning at low KVp and contrast reduction- In some cases alternative for plain film Spectral imaging and real iterative rec.: a new dimension in CT- Ultra low dose at diagnostic quality up to 95% reduction- Artifact reduction and high quality images- Characterization and differentiation- Contrast amount and concentration reductionThank you for your attention