The goal of the work was to build a low cost cardiac phantom that mimics cardiac motion and measure thermometric profile based on applied B₁⁺ fields during MR imaging. The phantom could be used to validate novel acquisition and reconstruction strategies specifically focusing on cardiac motion and Specific Absorption Rate (SAR).

The cardiac phantom consists of two square boxes made out of acrylic material, sizes of which were 200´200´200mm3 and 350´350´350mm3. The smaller box housed the heart that was prepared using Poly Vinyl Alcohol (PVA). The larger box housed the mechanical gear system that produces the piston motion. 10% of PVA solution was prepared as in ref. [1] and cardiac mold was prepared as per the specification based on American Society of Echocardiography using polymer clay. The cardiac motion inside the phantom was simulated using the mechanical gear setup. The gear setup consisted of the one oblique shaped oval and small disc, the rotational motion between the two was converted to horizontal piston motion. The piston was connected to pumping system to pump in Copper Sulphate (CuSO4) solution that mimic the blood contrast. The cardiac phantom setup is as shown in the figure 1.

In-vivo data was acquired on a 1.5T scanner using cardiac cine protocol in short axis and four chamber views with the following acquisition parameters: TR/TE=66.8/1.42ms, FA=80, number of measurements 10. Small acrylic box housed PVA cardiac phantom was placed in the head coil and the mechanical setup housed in large acrylic box was placed on the patient table for pumping the CuSO4 solution. A thread was connected to the gear setup inside the large box, the thread was pulled to get the piston motion. The piston would push the pumping system to pump in the CuSO4 solution. PVA phantom had two valves connected to pipes, one was for pumping-in CuSO4 solution and other one was kept open to avoid air bubbles in the cardiac chamber during pump-in.

Thermometry:

The cardiac phantom was positioned in the iso-center, supported with foam pads on the sides and paper tapes. Tap water was filled in the heart phantom. To obtain the thermometric profile of the cardiac phantom 12 probes were inserted into the phantom as shown in figure 3(d,e), with required thermal insulation. One probe was placed inside the bore and one probe inside the head coil to serve as controls. The above thermometry experiments were carried out at 3T GE Signa scanner. The scan protocol included a gradient echo sequence was part of a B₁⁺ stress test for a single channel transmit head coil which produced a B1max of 3.2µT. This experiment was repeated twice with RF being turned on and off for half hour each.

Four chamber view and oblique short axis image of the phantom can be seen as shown in Figure 2 and show cardiac motion. This depicts change in the cardiac wall (myocardium) in the four chamber view and change in the left and right ventricle size in the oblique short axis view at 10 different time points. Vertical dotted lines are plotted to separate the left and right chamber and horizontal dotted line to show the changes in the size of left and right ventricle during CuSO4 pump in. Figure 3(a) shows the probe poked into the cardiac phantom prepared for thermometric measurement. Figure 3(b) shows the coil sensitivity of the image. Cardiac phantom was placed vertically in the head coil and bottom part of the phantom was placed close to the coil elements which resulted in hyperintensity in the bottom part of the cardiac phantom image. Figure 3(c) shows the short axis view of the cardiac phantom. Temperature increased with the application of RF and reduced when it was turned off. This can be observed in Figure 4. We have achieved 40 to 50 beats per minute and currently working on matching the normal human heart beat of approximately 70. The phantom provides for a detailed study of thermometry with and without the effect of perfusion. The phantom also provides an opportunity to correlate local SAR findings with temperature measurements in the heart phantom. This is critical as accelerated cardiac MRI is mostly restricted by SAR constraints.