Rachid Fahmi PhD, Presenter: Institutional Grant support, Koninklijke Philips NV

Brendan Eck, Abstract Co-Author: Research Grant, Koninklijke Philips NV

Jacob Levi MS, Abstract Co-Author: Institutional Research Grant, Koninklijke Philips NV

Anas Fares MD, Abstract Co-Author: Institutional Research Grant, Koninklijke Philips NV

Amar Dhanantwari, Abstract Co-Author: Employee, Koninklijke Philips NV

Mani Vembar MS, Abstract Co-Author: Employee, Koninklijke Philips NV

Hiram Bezerra, Abstract Co-Author: Nothing to Disclose

David L. Wilson PhD, Abstract Co-Author: Co-owner, BioInVision Inc Research Grant, Koninklijke Philips NV

We used a prototype spectral detector CT (SDCT) scanner (Philips Healthcare) and advanced processing for cardiac CT perfusion (CTP) in a porcine model. This scanner limits beam hardening (BH), eliminates partial scan artifacts with fast 360deg scans, and limits errors in material decomposition. Our initial studies aim at developing a methodology and determining the effect of mono-energetic (MonoE) reconstructions on quality of CTP measurements.

A porcine model mimics coronary stenosis through partial occlusion of the LAD artery with a balloon catheter as adjusted with FFR measurements. Dynamic CT (120kVp/100mAs) scans were acquired at end-systole (45%RR). Projection-based MonoE images were reconstructed for energies [50,120]keV at 10keV increments. We applied 3D cubic B-spline normalized mutual-information to register reconstructed data to a common reference and created a mean CT volume. Semiautomatic segmentation of the LV myocardium was performed on the average volume and propagated across the 4D sequences. Absolute myocardial blood flow (MBF) was computed using a deconvolution-based approach.

At baseline (FFR=1), MonoE images at peak enhancement, and iodine maps gave relatively uniform values within the entire myocardium, with no obvious BH artifacts, whereas at FFR=0.7, clear deficits in the LAD regions were noticed mainly for lower keV’s. Mean MBF=(100.99±26.1mL/min/100g) at FFR=1 and (31.43±13.3mL/min/100g) at FFR=0.7, which corresponds to about 69% decrease in blood flow. Qualitatively, 70keV images provide a more uniform MBF map with a high contrast between ischemic and normal tissue. This contrast depends on the keV, and we used flow-CNR (or CNRf) to evaluate this dependence. Two equally sized ROI’s (same for all keVs) were manually drawn in the LAD and remote areas. We computed CNRf as: CNRf= (µMBF(remote) - μMBF(LAD)) / σMBF(remote). Mean CNRf=8.32±4.27(50keV), 7.33±3.4(70keV), and 0.94±0.52(120keV).

The SDCT scanner can be used to acquire high quality CTP images which are relatively artifact free. MBF decrease was only noticed in LAD territory between ischemic and baseline conditions. Flow CNR between occluded LAD and remote areas decreases for higher keVs.

Adding robust cardiac CTP to coronary CTA will create a powerful non-invasive diagnostic test for cardiovascular disease and an ideal gatekeeper exam for cardiac catheterization.

Fahmi, R, Eck, B, Levi, J, Fares, A, Dhanantwari, A, Vembar, M, Bezerra, H, Wilson, D, Dynamic Myocardial Perfusion in a Porcine Ischemia Model Using Spectral Detector CT.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14012181.html