Four-dimensional magnetic resonance imaging (4D-MRI) using image-based respiratory surrogate: A feasibility study

Purpose: Four-dimensional computed tomography (4D-CT) has been widely used in radiation therapy to assess patient-specific breathing motion for determining individual safety margins. However, it has two major drawbacks: low soft-tissue contrast and an excessive imaging dose to the patient. This research aimed to develop a clinically feasible four-dimensional magnetic resonance imaging (4D-MRI) technique to overcome these limitations. Methods: The proposed 4D-MRI technique was achieved by continuously acquiring axial images throughout the breathing cycle using fast 2D cine-MR imaging, and then retrospectively sorting the images by respiratory phase. The key component of the technique was the use of body area (BA) of the axial MR images as an internal respiratory surrogate to extract the breathing signal. The validation of the BA surrogate was performed using 4D-CT images of 12 cancer patients by comparing the respiratory phases determined using the BA method to those determined clinically using the Real-time position management (RPM) system. The feasibility of the 4D-MRI technique was tested on a dynamic motion phantom, the 4D extended Cardiac Torso (XCAT) digital phantom, and two healthy human subjects. Results: Respiratory phases determined from the BA matched closely to those determined from the RPM: mean (SD) difference in phase: -3.9 (6.4); mean (SD) absolute difference in phase: 10.40 (3.3); mean (SD) correlation coefficient: 0.93 (0.04). In the motion phantom study, 4D-MRI clearly showed the sinusoidal motion of the phantom; image artifacts observed were minimal to none. Motion trajectories measured from 4D-MRI and 2D cine-MRI (used as a reference) matched excellently: the mean (SD) absolute difference in motion amplitude: -0.3 (0.5) mm. In the 4D-XCAT phantom study, the simulated 4D-MRI images showed good consistency with the original 4D-XCAT phantom images. The motion trajectory of the hypothesized tumor matched excellently between the two, with a mean (SD) absolute difference in motion amplitude of 0.5 (0.4) mm. 4D-MRI was able to reveal the respiratory motion of internal organs in both human subjects; superior-inferior (SI) maximum motion of the left kidney of Subject #1 and the diaphragm of Subject #2 measured from 4D-MRI was 0.88 and 1.32 cm, respectively. Conclusions: Preliminary results of our study demonstrated the feasibility of a novel retrospective 4D-MRI technique that uses body area as a respiratory surrogate.