Workshop Summary
Anthony Butler Radiologist and Physicist Disclaimer
I have many biases - I am a clinician by job and physicist by hobby - My university has been a member of Medipix since 2006 - I am a founder of MARS Bioimaging Ltd Overview
Participants and submissions
Context
Review of talks
Conclusions Disclaimer
NZ has a strange view of the world Disclaimer
NZ has a strange view of the world Participants
Context
Review of talks
Conclusions Scientific committee
Michael Campbell, CERN
Anthony Butler, Univ. of Canterbury & MARS Bioimaging
Steffen Kappler, Siemens Healthineers
Ewald Roessl, Philips Research Laboratories
Katsuyuki (Ken) Taguchi, Johns Hopkins University
Richard Thompson, Canon Medical Research
Brian Yanoff, GE Global Research
… with lots of assistance from Patricia Mage-Granados
Initiation only for several reasons Participants 119 registrants
2017 => 130 2013 => 102 2015 => 112 2011 => 76
50 Industry; 46 Scientists; 23 Medical >85% are returning guests Participants and submissions
Context
Review of talks
Conclusions Context The electromagnetic spectrum
X-ray colour: Wavelength, Frequency, or Energy Context
Spectral CT is true colour x-ray imaging Context
Hounsfield, Nobel Prize 1979
Computed Tomography is 3D X-rays Context
300,000,000 people per year get a CT
Hounsfield, Nobel Prize 1979
Computed Tomography is 3D X-rays Context
Rontgen 1895, to CERN technology 2018 Context Before SpecXray
Earnest Rutherford:
We have no money, therefore we must think
Particle physics pushes frontiers of knowledge Context Before SpecXray
Early ’80s, direct Si detectors – Erik Heijne, Robert Klanner, Gehard Lutz
Their role is recognized by the 2017 High Energy and Particle Physics Prize of EPS
Mid ’90s, Medipix – Michael Campbell “Various applications like Medical Imaging should profit” Context SpecXray 2011
Mouse, Mouse, Human, Anthony Butler Ewald Roessel Jerry Arenson
MARS, Philips, and GE Healthcare Context 2011 to 2019 PCD 2017
2011
Rajiv Gupta, Mass.Gen.Hosp Context 2011 to 2019 PCD 2017
2011 2013-2017
Rajiv Gupta, Mass.Gen.Hosp Context 2011 to 2019 PCD 2017
2011 2013-2017 2019
Rajiv Gupta, Mass.Gen.Hosp Context The world has changed
More focus on value, not excellence … … scanners for emerging markets … United Imaging Ltd … health-pathways and efficient care
Radiology market has changed Context The world has changed
More focus on value, not excellence … … scanners for emerging markets … United Imaging Ltd … health-pathways and efficient care
Earlier decision making… … growth of Urgent and Point-of-Care … Artificial Intelligence
Radiology market has changed I’m lucky if I can learn one thing per talk
Participants
Context
Review of talks
Conclusions Invited talk
A patient in Tuba City, Arizona (Navajo Nation), 5hr drive from Albuquerque
Scanners with no moving parts in order to meet the community’s needs
Re-imagining the imaging system Rajiv Gupta, Mass.Gen.Hosp Invited talk
Image processing is a critical part of the system
Reimaging the imaging system Rajiv Gupta, Mass.Gen.Hosp Taking it to humans
The world is watching Anthony Butler, Uni. Of Canterbury Taking it to humans
Imaging where the clinician wants to see
Universal protocols practical for users
Strong focus on clinical benefits Cynthia McCollough, Mayo Clinic Clinical applications
PCD
EID Multiple image metrics showing PCD gives better image quality
Photon counting gives better image quality Jayasai Rajagopol, NIH Clinical Centre Clinical applications
Image quality
=
Can I make a diagnosis?
Need to assist clinicians Thorsten Fleiter, Uni. of Maryland Clinical applications
Iodine and Bismuth used simultaneously ( animal models of tumors )
Multiple contrast agents for radiological surgery William Pritchard, NIH Clinical Centre Clinical applications
Structure and mineralization Gout Micro-fractures
Lots of easily accessible bone applications Aamir Raja, Uni. of Otago Prototype systems
Building systems is hard!
Showed complexity of systems Pete Edic, GE Global Research Prototype systems
Protocols vary, but potentially big dose advantages
PCT allows high resolution and low dose Shuai Leng, Mayo Clinic Prototype systems
2 sources + 2 bin detectors = 4 energies
Number of bins and energy separation is critical Shuai Leng, Mayo Clinic Detector systems
Gd/Ho K-edges: 6 keV apart
Very close materials can be separated Magdalena Bazalova-Carter, Uni. of Victoria Detector systems
1 mm CdTe
Internal Escape XRF peak
Improved spectral resolution Improved spatial resolution
Timepix3 is a REALLY advanced detector Jan Jakubek, ADVACAM Detector systems
Singe chip Compton camera for thyroid Jan Jakubek, ADVACAM Detector systems
4 side buttable, 5 energy, high flux
Kris also talked about who might benefit from spectral imaging
CZT and commercial detectors have improved Kris Iniewski, Redlen Technologies Lots of time for discussion
Coppet and its chateau Thanks to Michael and Patricia Detector modelling
Charge transport over time Yi Qiang, Cannon Medical Research Simulations and Recon
…. may be enough to get good images
Just measure co-incident, but don’t correct Scott Hsieh, UCLA Simulations and Recon
Spectral phase contrast – an important future Thorsten Sellerer, Technical Uni. Munich Simulations and Recon
What is the optimal basis for describing tissues?
Spectra -> Eigen-basis -> Materials
Eigen-tissue decomposition Mikael Simard, Uni. of Montreal Detectors of HEP
Very clear description of LHC, ATLAS, and CMS Walter Snoeys, CERN Molecular imaging
Two contrast agents - tumor vascularization - nano-particle distribution
Nano-particles are used in cancer research Ananth Annapragada, Baylor College Molecular imaging Ex-vivo imaging means he knows what he wants to see in humans
Cardiovascular disease causes 37% of EU deaths Steven Gieseg, Uni. Canterbury Material Decomposition
Improvements in projection based MD Artur Sossin, Philips Innovative Technology Material Decomposition
More flexible: - apply image constrains - variable energy bins - non-ideal detectors - flux dependent effects - variable input spectra Projections Materials
One step reconstruction and MD Taly Schmidt, Marquette Uni. High-Z semi-conductors
Instability more noticeable at high energy thresholds
Th1 Th2 Th3 Th4 Th5
Stability counts
>200 Mcps/mm2 Integrated achievable!
High photon flux is a challenge Georgios Prekas, Redlen Technology ASICs
Excellent for pushing the limits of spectral x-ray technology
65nm technology Pixel matrix of 512 x 448 pixels (55 µm x 55 µm) Particle identification and tracking (Data driven) Imaging (frame based with CRW sequential readout) Sub-ns time binning ~195ps Max data output ≤163.84 Gbps Energy resolution <1 KeV 4 side buttable due to TSV
Timepix4 is almost here Xavi Llopart, CERN ASICs
Increase detector area for real clinical CTs
TSV
New 4 side buttable ASIC for CT Edgar Goederer, Siemens Healthineers ASICs
Increase detector area for real clinical CTs
New 4 side buttable ASIC for CT Edgar Goederer, Siemens Healthineers ASICs
Charging sharing and pulse pile-up are the main challenges
Charge sharing correction scheme Pawl Grybos, AGH Uni.of Science and Technology ASICs
Want to have: charge sharing correction, pile-up correction, more energies Currently: Future:
2 energies Pile-up rejection Charge sharing correction 256 energy bins-real time 2 side “buttable” Energy resolution 7 to 9% OCR= 8 Mcps/pixel https://ctcicblog.mayo.edu/hubca https://www.nih.gov/news-events/news-releases/nih-uses-photon-counting-ct-scanner-patients-first-time p/our-scanners/
Moving beyond 2 energies for clinical CT Loick Verger, CEA LETI ASICs My favorite talk !
A great review of competing requirements Rafael Ballabriga, CERN Participants
Context
Review of talks
Conclusions Conclusion
“This is what happens when you bring people with good ideas together to form a community”
Stephan Kappler, Scientific committee Conclusion
Looking back what have we learned since 2011? The medical applications of spectral x-ray detectors are: - new information - higher resolution - lower dose Conclusion
Looking back what have we learned since 2011? The medical applications of spectral x-ray detectors are: - new information - higher resolution - lower dose
2011 – would any of this work? Conclusion
What have we learned in this workshop? Spectral CT is going to happen - all of the challenges are successfully being address - it will reach the clinics (timelines remain uncertain)
At least 300,000,000 people per year will benefit Thanks you
Michael Campbell, CERN
Anthony Butler, Univ. of Canterbury & MARS Bioimaging
Steffen Kappler, Siemens Healthineers
Ewald Roessl, Philips Research Laboratories
Katsuyuki (Ken) Taguchi, Johns Hopkins University
Richard Thompson, Canon Medical Research
Brian Yanoff, GE Global Research