3D ultrasound coronarography: first proof of concept study on isolated beating rat hearts

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): Bettencourt Foundation Background/Introduction We demonstrated recently that Ultrafast ultrasound Doppler imaging can image the intramyocardial coronary circulation in beating hearts of large animals and patients [1]. Yet, ultrasound spatial resolution remains limited by wave physics and coronaries smaller than ∼100 µm could not be imaged. Ultrasound Localization Microscopy (ULM) [2] was recently introduced to tackle this issue and exploit the micrometric localization of microbubble contrast agents at ultrafast frame rate in order to image blood flows in micrometer vessels. Purpose The objective of this work was to demonstrate that 3D ultrafast ultrasound with contrast agents can provide the full 3D mapping of the coronary microcirculation with quantitative flow velocity on a beating rat heart. Methods Acquisitions were performed on ex vivo rat hearts (n = 5) with retrograde perfusion (Langendorff model). A flow of a Krebs–Henseleit solution mixed with a diluted microbubbles solution (0.22%) was perfused at controlled pressure into the coronary arteries (between 5 and 15 mL/min). We used a 32 × 32 elements, 8-MHz matrix-array ultrasound transducer connected to a 1024-channel programmable ultrasound scanner. An ultrafast Doppler imaging sequence consisting of 9 plane waves was transmitted at a PRF of 20 kHz during 270 ms and repeated 40 times. After beamforming and SVD clutter filtering, the microbubbles were localized and tracked in 3D. Flow velocity were mapped at baseline and after infusion of Adenosine (10e-5 µMol) at constant coronary perfusion pressure (120 mm Hg). Eventually, the hearts were fixed using formaldehyde perfusion and imaged by µCT after injection of radio opaque agent. Results We successfully imaged the coronary blood flows of entire rat hearts. It revealed the entire vasculature from large main coronaries arteries (cross section up to 1 mm) to small arterioles (smaller than 40 µm). Coronary flow velocities ranged from [1 – 50] cm/s depending on the arteries diameter. Velocity estimates were validated in vitro in tubes of Ø0.58mm and were in good agreement with theoretical values of a Poiseuille’s flow (relative ratio of 10% for maximum velocities). After Adenosine infusion, perfusion flow rates increased 102% ± 50% (p < 0.05) on average. Eventually, anatomy revealed by 3D ultrasound coronarography was in accordance with the anatomy revealed by the µCT. Conclusion(s) We demonstrated the feasibility of 3D ultrasound coronarography on isolated beating rat hearts. This technique has the potential to become a novel imaging tool to investigate the coronary micro-circulation and quantify non-invasively the Coronary Flow Reserve (CFR). Abstract Figure. Ultrasound coronarography