Ultra-high field cardiac MRI in large animals and humans for translational cardiovascular research

Figure 1. Key components of the translational MRI infrastructure. (A) Operating theater dedicated to large animal surgery, including a used C-arm x-ray fluoroscopy machine for monitoring of instrumentation of the animals, e.g., balloon-occlusion for induction of myocardial ischemia or infarction. Radiative shielding of the room's walls was implemented to fulfill legal requirements for x-ray scanning. Moreover, Doppler ultrasound allows for further characterization of the animals. (B) The animal waiting area is adjacent to the magnet room. Animals are positioned on a magnetic trolley (blue arrow), on which the animals can be transported through the sliding door and magnetic material detector to the magnet room. Access to pressurized air, and anesthesia gas extraction system allow for mechanical ventilation of the anesthetized animals before they are granted access to the magnet room through the magnet detectors (vertical, with currently green lights on top). To reduce anesthesia duration and animal stress, every attempt is made to keep this phase as short as possible. (C) Magnet room during animal MRI, including MR-compatible mechanical ventilator (MV), MR-compatible infusion machines (I),…

Figure 2. RF coil concept of in-house developed coils for experimental imaging. (A): (1) Coil for cardiac MRI in small animals (tunable to the 1H and 19F Larmor frequency) suitable for mice up to 40 g body weight. This coil is for use with the small animal MRI system. All other coils shown in this figure were developed in-house for use with the whole-body 7 T MRI system: (2) 8Tx/16Rx pTX coil for imaging excised organs from large animals in the whole-body 7 T MRI system (c.f., Figure 5). (3) 8Tx/16Rx array for 7 T cardiac MRI in human subjects. Panel (B) Dedicated 8Tx/16Rx arrays for cardiac MRI in pigs of different weights: (a) 60–90 kg, (b) 40–60 kg, (c) 20–40 kg. The shape of each coil is adapted to the typical shape of a German landrace pig. The three differently sized coils allow for optimal coil filling factor and, thus, SNR. The scanner bore size as well as, to a minor extent, the thickness of the RF coils, limit the scanning of pigs to 90 kg or less. (C) Visualization of the housing of the three coils in (B). The different shape of the thorax of different sized animals is visible. No posterior coil elements were used because their signal contribution for the MRI signal of the heart…

Figure 3. Imaging in a pig model 90 min after occlusion of the left circumflex artery and subsequent reperfusion. (A) The high-resolution first-pass perfusion image obtained in the early (day 3) acute phase of myocardial infarction in a short-axis view after injection of 0.0167 mmol/kg BW Gd-DOTA demonstrates a large perfusion defect. Extensive perfusion heterogeneity is observed with hypoperfused myocardium (green arrows). Smaller regions (blue arrows) appear with perfusion almost equal to that of remote myocardium (white arrows). Spatial resolution was 0.95 × 1.4 × 6 mm3. (B) T2*-weighted imaging at the same examination demonstrates similar inhomogeneities with hypointensities probably induced by acute bleeding. Spatial resolution: 2.2 × 2.5 × 6 mm3. (C) In the late acute phase, homogenization of T2*-contrast probably demonstrates microstructural changes, e.g., myocardial remodeling. (D) High-resolution imaging of chronic myocardial infarction (day 60 after infarction) using the late gadolinium enhancement (LGE) technique following injection of a total of 0.1 mmol/kg BW Gd-DOTA (spatial resolution 1.0 × 1.2 × 6.0 mm3). (E) LGE imaging in-situ 5 min after cardiac arrest. Without…

Figure 4. Serial imaging in a pig using the coil concept of Figure 2. Up to 60 days post-infarction. Images were scaled to show the approximately identical size of the heart despite the obvious growth that can be determined from the mass determination given for each measurement. (A) High image quality can be obtained throughout the whole series of measurements. Please also note the improved visualization of the left anterior descending (LAD) coronary artery (red arrows) at 7 T. Representative short axis cine image acquired with moderate (B) and high (C) resolution, and a measurement obtained in another 40 kg pig at 3 T using a commercial four channel flexible coil. (D). At 7 T, and with the concept of size- and shape adapted RF-coils, a marked improvement of spatial resolution and image detail is observed when compared to 3 T. For this measurement in another animal, a prior test identified the best-fitting coil which gave the best image quality at 3 T. This was the commercial flexible four-channel coil used for this measurement.

Figure 5. High-resolution ex-vivo imaging of excised pig hearts using an (A) in-house built 16 element transmit/receive coil. This coil provides an optimal coil filling factor allowing for highest resolution imaging and image quality e.g., in (B) high-resolution (0.15 × 0.15 × 0.5 mm3) susceptibility weighted imaging in a mid-ventricular short-axis view. Distinct signal intensity variations in the healthy pig heart are depicted in the endo- and epicardial left-ventricular myocardium, and in the right ventricle. The source of this image contrast is not clear at this stage but may be a consequence of transmural variation of the myofiber helix angle, and thus, orientationally variable susceptibility effects in otherwise normal myocardial tissue. Dark dot-like structures denote blood vessels. (C,D) Anatomical imaging of late-acute myocardial infarction with 0.1 × 0.1 × 0.8 mm3 spatial resolution. The dotted rectangle denotes the approximate position of the magnified portion of the image in (C). A B1-related signal void is seen in the top left portion of subfigure (C). Acquisition time for 12 slices was 40 min. Scar tissue structure is well visualized on an almost histologic level. Subfigure…

Figure 6. Experimental imaging in excised mouse hearts using a liquid nitrogen-cooled RF coil. (A) Diffusion tensor imaging-based tracking of myocardial myofibers with 100 µm3 isotropic spatial resolution allows for a visualization of the change of the transmural change of fiber orientation and high precision quantification of the myofiber helix angle. In an isoproterenol myocardial damage model, distinct change of myofiber orientation as represented by the helix angle in subendocardial regions was observed at different positions in the mouse heart (grey bars: control group. Red bars; isoproterenterol model). Reprinted under CC-BY license from (61). (B) DTI-based fiber-tracking in mouse aorta with 80 µm3 isotropic spatial resolution allows for visualization of the orientation of intimal vascular smooth muscle cells. (C) T2-weighted imaging of an excised porcine aorta with 0.3 mm slice thickness allows for an almost histologic visualization of vascular anatomy. (D) Dynamic MRI in the excised aorta following exposition to 30 mM TEMPOL as an T1-modifying contrast agent. The TEMPOL-induced change of the T1 relaxation times from outside and inside the vessel allows for the study of its kinetics…

Figure 7. UHF-MRI in humans. (A) Short axis CINE imaging in a healthy volunteer shows signal inhomogeneities because of B0 and B1 inhomogeneities typical for human imaging at 7 T, but imaging of the left-ventricular outflow (B) tract nicely visualized the aortic valve (red arrow) and the left main coronary artery (yellow arrow). (C) Comparison of stroke volume as obtained from functional CINE imaging with phase-contrast flow measurement in the ascending aorta demonstrates excellent agreement and, thus, reliability of UHF-MRI-based quantitative analysis of this type of measurement. For these measurements in subfigures (A–C), the vendor-supplied 1Tx/16RX cardiac coil was used. Using a new 8 Tx/16Rx MRI coil (D) allows for better image quality in a patient (E) which was scanned during a routine check-up where ECG abnormalities were detected. In combination with subsequent clinical imaging [e.g., (B)], prior ischemic myocardial damage was diagnosed. Although the 3 T scan in this patient (F) is not of particularly good image quality, this image comparison demonstrates the potential of the image quality inherent in UHF-MRI. The black dot in the 7 T-image (red arrow) is an artifact from image…

Figure 8. Summary of translational imaging concept. The described research infrastructure allows imaging for the full spectrum of samples relevant for modern translational research, from (A) cardiac organoid (30 × 30 × 300 µm3) with yellow marked regions of interest for subsequent data analysis in the central and peripheral region of the organoid, (B) Arrested zebrafish heart ex-vivo (25 × 25 × 50 µm3, c.f. Supplementary video S1), (C) CINE MRI in a mouse heart in-vivo (0.2 × 0.2 × 1.0 mm3, c.f. Supplementary video S2), (D) pig heart in-vivo (0.4 × 0.4 × 6 mm3), and (E) human heart in-vivo (0.6 × 0.6 × 6.0 mm3, c.f. Supplemental video S3). All imaging sections were positioned in a mid-ventricular short-axis view.