ECR 2018 From Diagnosis to Prognosis ECR 2018 From Diagnosis to Prognosis Thursday, March 1, 2018/08:30-10:00/Room N Les Outils Cliniques de Demain en Scanner Cardiaque Cardiaque Status en 2018 Rodrigo Salgado, MD, PhD Rodrigo Salgado, MD, PhD Dpt. Of Radiology Antwerp University Hospital - Heilig Hart Lier Belgium Dpt. Of Radiology Antwerp University Hospital - Heilig Hart Lier Belgium rodrigo.salgado@uza.be
Cardiac CT was a popstar!
Evolving applications Myocardial perfusion CT 4D evaluation of prosthetic heart valves Clinical applications are evolving as equipment improves Movie courtesy of Ricardo Budde, MD - UMC Utrecht
Acquiring the Data: CT performance Steady increase in performance of CT
CT: The Holy Grail Spatial resolution Temporal resolution Have we now everything that we desire? Speed of execution
New CT technology
New hardware I HAVE A NEW TOY!
Single Energy CT Single x-ray energy spectrum attenuation based on material density & composition CT detector Ca+ Iodine Photons Attenuation characteristics are not necessarily unique to different materials
Dual Energy CT Dual x-ray energy spectrum Specific materials attenuate differently from two X-ray spectra
Dual Energy CT DECT can be performed with single- or double-source systems
Dual Energy CT Continuous dual energy acquisition in every scan Latest technology uses one detector with two layers
Dual Energy CT DECT technology offers new applications
Clinical applications Virtual unenhanced CT from CT-A acquisition native CT-A Virtual unenhanced CT Dual energy CT can provide Ca-score from CT-A images
Clinical applications Thrombus or not? Single contrast-phase detection of thrombus in LAA
Clinical applications Better SNR & CNR 30 ml contrast medium Dual energy CT can lead to >50% CM savings due to better contrast
OK, this is boring! WHAT ELSE DO YOU HAVE?
Dual Energy CT problems Dual X-ray source Fast kvp switching Dual-layer detector 90 phase shift between lowand high-energy data Scattered radiation Relatively high overlap of the energy spectra Noise level may differ between low and high energy images Relatively high overlap of the energy spectra Fixed thresholds Current dual energy CT systems have their disadvantages
New Kid in Town Photon counting CT is coming!
Slide courtesy of R. Symons, MD Detector comparison Energy-Integrating Detectors (EIDs) X-ray photon -> light -> electric signal (slow) Photon-counting detectors (PCDs) X-ray photon -> electric signal (fast) SEPTUM DETECTOR SEPTUM DETECTOR SEPTUM - -- - - - - - INDIRECT CONVERSION DIRECT CONVERSION
Slide courtesy of R. Symons, MD Detector comparison Energy-Integrating Detectors (EIDs) Combine effects of energy and number of photons in intensity value Loss of spectral information Photon-counting detectors (PCDs) Measure number of photons (counts) and energies separately (faster integration time) Photons divided into energy bins Spectral information retained
Slide courtesy of R. Symons, MD Advantages over current technology Spectral information always available: no special protocol required Less overlap between low and high energy data: better material separation Energy thresholds can be adapted to clinical indication: more flexible >2 energies possible: potential for multi-contrast imaging
3 wild mongrel dogs with chronic myocardial infarction Late enhancement: Gadolinium first pass Iodine Gd injection I injection Imaging 10 min 20 sec? By using different contrast media, can we differentiate them with one CT-acquisition?
Multi-energy cardiac imaging 3 wild mongrel dogs with chronic myocardial infarction Late enhancement Gadolinium, first pass Iodine First-pass enhancement of blood late enhancement of subendocardial Virtual unenhanced Multiphase CT acquired in a single acquisition!
Excellent CT-MR-histology correlation of scar
Pulse)Height Intensity Spectral information Time Pulse+Height E4 E3 E2 E1 E bin 4 E bin 3 E bin 2 E bin 1 Time Spatial resolution Noise properties PC-CT also provides better spatial resolution & noise properties
Photon-counting CT Multi-Energy visualization of plaques PC-CT can help in differentiating plaque materials
Software! THE POWER OF THE ROBOTS
Coronary CTA One-pass CT subtraction Subtraction of Calcium & stent material without additional acquisition
Coronary CTA Arrhythmia & low-dose CT 144 BPM Atrial fibrillation 0,19 msv 30 cc contrast medium Cardiac CT is applicable to a larger population
CT-FFR Additional value over conventional angiography CT-FFR requires excellent image quality
CT-FFR Additional value over conventional angiography CT-FFR still has some problems in the critical range
Combining CT Perfusion & FFR Diagnostic performance may improve by CT perfusion & -FFR
From little to big data Morphology Function Comparison with large datasets
Machine learning (ML) Automatic Ca-score determination ML can help in automated Ca-score from CT-A & non-gated scans
Machine learning (ML) CT-FFR has potential to improve FFR algorithms
Machine learning (ML) ML in patients with suspected coronary artery disease Adding Machine Learning for better mortality prediction than other scores
Machine learning (ML) Integration of automated AI results in clinical reporting
Final Thoughts A PRACTICAL POINT OF VIEW
CT detector technology is evolving
New technologies induce new applications
Excellent image quality is nevertgheless required
A radiologist with AI is probably better than one without
Thank you!