Time-resolved magnetic resonance fingerprinting for radiotherapy motion management

This study aims to develop a novel time-resolved magnetic resonance fingerprinting (TR-MRF) technique for respiratory motion imaging applications.

The TR-MRF technique consists of repeated MRF acquisitions using an unbalanced steady-state free precession sequence with spiral-in-spiral-out trajectory. Time-resolved magnetic resonance fingerprinting was first tested via computer simulation using a four-dimensional (4D) extended cardiac-torso (XCAT) phantom for both regular and irregular breathing profiles, and was tested in three healthy volunteers. Parametric TR-MRF maps at different respiratory phases were subsequently estimated using our TR-MRF sorting and reconstruction techniques. The resulting TR-MRF maps were evaluated using a set of metrices related to radiotherapy applications, including absolute difference in motion amplitude, error in the amplitude of diaphragm motion (ADM), tumor volume error (TVE), signal-to-noise ratio (SNR), and tumor contrast.

TR-MRF maps with regular and irregular breathing were successfully generated in XCAT phantom. Numerical simulations showed that the TVE were 1.6 ± 2.7% and 1.3 ± 2.2%, the average absolute differences in tumor motion amplitude were 0.3 ± 0.7 mm and 0.3 ± 0.6 mm, and the ADM were 4.1 ± 0.9% and 3.5 ± 0.9% for irregular and regular breathing, respectively. The SNR of the T₁ and T₂ maps of the liver and the tumor were generally higher for regular breathing compared to irregular breathing, whereas tumor-to-liver contrast is similar between the two breathing patterns. The proposed technique was successfully implemented on the healthy volunteers.

We have successfully demonstrated in both digital phantom and healthy subjects a novel TR-MRF technique capable of imaging respiratory motions with simultaneous quantification of MR multiparametric maps.