Magnetic Particle Imaging for High Temporal Resolution Assessment of Aneurysm Hemodynamics

Abstract

Purpose: The purpose of this work was to demonstrate the capability of magnetic particle imaging (MPI) to assess the hemodynamics in a realistic 3D aneurysm model obtained by additive manufacturing. MPI was compared with magnetic resonance imaging (MRI) and dynamic digital subtraction angiography (DSA).

Materials and methods: The aneurysm model was of saccular morphology (7 mm dome height, 5 mm cross-section, 3-4 mm neck, 3.5 mm parent artery diameter) and connected to a peristaltic pump delivering a physiological flow (250 mL/min) and pulsation rate (70/min). High-resolution (4 h long) 4D phase contrast flow quantification (4D pc-fq) MRI was used to directly assess the hemodynamics of the model. Dynamic MPI, MRI, and DSA were performed with contrast agent injections (3 mL volume in 3 s) through a proximally placed catheter.

Results and discussion: 4D pc-fq measurements showed distinct pulsatile flow velocities (20-80 cm/s) as well as lower flow velocities and a vortex inside the aneurysm. All three dynamic methods (MPI, MRI, and DSA) also showed a clear pulsation pattern as well as delayed contrast agent dynamics within the aneurysm, which is most likely caused by the vortex within the aneurysm. Due to the high temporal resolution of MPI and DSA, it was possible to track the contrast agent bolus through the model and to estimate the average flow velocity (about 60 cm/s), which is in accordance with the 4D pc-fq measurements.

Conclusions: The ionizing radiation free, 4D high resolution MPI method is a very promising tool for imaging and characterization of hemodynamics in human. It carries the possibility of overcoming certain disadvantages of other modalities like considerably lower temporal resolution of dynamic MRI and limited 2D characteristics of DSA. Furthermore, additive manufacturing is the key for translating powerful pre-clinical techniques into the clinic.

Fig 1. Photo of the aneurysm model.

Regions or volumes of interest (inlet, aneurysm, outlet) were chosen at the indicated positions of the model in the corresponding 3D or 2D image data, respectively. The signal within these regions was plotted and analyzed to compare the imaging methods.

Fig 2. False color (left) and vector field (right) visualization of the 4D pc-fq MRI at the time point of maximal flow velocity during one pulsation cycle.

Lower flow velocity and different flow directions are clearly visible inside the aneurysm. Movies showing all phases of the pulsation are available as S1 and S2 Videos.

Fig 3. Cyclic repetition of the 4D pc-fq MRI (top) shows distinct pulsation and a much lower flow velocity inside the aneurysm.

Dynamic MRI (second top, cubic spline interpolated), MPI (second bottom), and DSA (bottom) also show distinct pulsation and delayed contrast agent dynamics inside the aneurysm. Regions were chosen in the corresponding image data as depicted in Fig 1. Pulsations and bolus injections did not perfectly match between the different methods, due to the resetting of the experiments.

Fig 4. Multiple frames of dynamic MRI, MPI, and DSA measurements during the passage of a local maximum of contrast agent concentration (see time points 2.5 s-2.8 s in Fig 2).

The passage of the contrast agent bolus through the main vessel of the model can be estimated in MPI and DSA, but not in MRI. A movie showing all frames of the dynamic scans is available as S3 Video.