Accelerating Dynamic Magnetic Resonance Imaging (MRI) for Lung Tumor Tracking Based on Low-Rank Decomposition in the Spatial–Temporal Domain: A Feasibility Study Based on Simulation and Preliminary Prospective Undersampled MRI

Abstract

Purpose To evaluate a low-rank decomposition method to reconstruct down-sampled k-space data for the purpose of tumor tracking. Methods and Materials Seven retrospective lung cancer patients were included in the simulation study. The fully-sampled k-space data were first generated from existing 2-dimensional dynamic MR images and then down-sampled by 5 × −20 × before reconstruction using a Cartesian undersampling mask. Two methods, a low-rank decomposition method using combined dynamic MR images (k-t SLR based on sparsity and low-rank penalties) and a total variation (TV) method using individual dynamic MR frames, were used to reconstruct images. The tumor trajectories were derived on the basis of autosegmentation of the resultant images. To further test its feasibility, k-t SLR was used to reconstruct prospective data of a healthy subject. An undersampled balanced steady-state free precession sequence with the same undersampling mask was used to acquire the imaging data. Results In the simulation study, higher imaging fidelity and low noise levels were achieved with the k-t SLR compared with TV. At 10 × undersampling, the k-t SLR method resulted in an average normalized mean square error <0.05, as opposed to 0.23 by using the TV reconstruction on individual frames. Less than 6% showed tracking errors >1 mm with 10 × down-sampling using k-t SLR, as opposed to 17% using TV. In the prospective study, k-t SLR substantially reduced reconstruction artifacts and retained anatomic details. Conclusions Magnetic resonance reconstruction using k-t SLR on highly undersampled dynamic MR imaging data results in high image quality useful for tumor tracking. The k-t SLR was superior to TV by better exploiting the intrinsic anatomic coherence of the same patient. The feasibility of k-t SLR was demonstrated by prospective imaging acquisition and reconstruction.

Accelerating volumetric cine MRI (VC-MRI) using undersampling for real-time 3D target localization/tracking in radiation therapy: a feasibility study

“…It is also worthy to mention that only retrospective patient data are included in the current study to evaluate the feasibility of accelerating the VC-MRI technique. Compared to prospective studies, retrospective studies have the advantages of providing the ground-truth images for evaluation and having the flexibility to investigate the effects of various under sampling strategies as well as various amounts of under sampling(Sarma et al , 2014; Subashi et al , 2013; Wang et al , 2016). However, retrospective studies do not account for certain challenges in prospective in-vivo applications of the technique, such as undersampling artifacts, timing errors, gradient imperfections and Eddy currents in the MR acquisition.…”

“…Compared to CBCT, MRI has better soft tissue contrast with no ionizing radiation dose to the patient. Different MRI techniques have been developed for localization of moving targets, including 2D cine MRI and 4D MRI (Cai et al , 2011; Sarma et al , 2014; Cervino et al , 2011). 2D cine MRI tracks the 2D in-plane motion of the target by continuously generating 2D images of the patient.…”

“…The previous technique (Harris et al , 2016) used a single, fully sampled 2D cine image to generate VC-MRI at a temporal resolution of 3–5 frames/s, which may not be sufficient to track a moving target for patients with short or irregular breathing cycles (Cervino et al , 2011; Sarma et al , 2014). Higher temporal resolution is also needed for VC-MRI to achieve better accuracy for post treatment accumulated dose evaluation and for deriving the tumor motion probability density function (PDF) for probability based treatments (Zhang et al , 2012).…”

Lung dynamic MRI deblurring using low‐rank decomposition and dictionary learning

“…The same principle was sucessfully used to reconstruct undersampled dynamic MRI images. 21 In comparison, deblurring single images separately without considering the coherence of 2D images in the dynamic series resulted in increase in image artifacts that offset the gains in blood vessel signals.…”

Functional Image-guided Radiotherapy Planning for Normal Lung Avoidance

“…Greater use of MR in radiotherapy is on the horizon [46 48], from delineation of tumour and organs at risk [49,50] and assessment of lung motion [51,52] to MR-only planning [53 56]. Additionally, the roll-out of hybrid MR treatment machines such as cobalt systems [57] and MR-linacs [58,59] provides further incentive for the advancement of both gas and novel Several groups have investigated the issues related to hyperpolarised gas MRI-based lung avoidance planning [17, 60 64].…”