A retrospective 4D‐MRI based on 2D diaphragm profiles for lung cancer patients

Abstract

Introduction 4D‐MRI, compared to 4D‐CT, provides better soft‐tissue contrast for target delineation. However, motion artefacts are often observed due to residual breathing variations. This study is to present a retrospective 4D‐MRI reconstruction method based on 2D diaphragm profiles to improve the quality of 4D‐MR images in the presence of significant breathing variations. Methods The proposed 4D‐MRI reconstruction method utilized diaphragm profiles (2D cine images on a single sagittal plan at the peak diaphragm) in conjunction with 4D‐MR scans (2D‐cine images on multiple pre‐determined coronal planes along the anterior‐posterior direction over a volume of interest). The diaphragm profile images were exploited to sort the 4D‐MR scans by matching respiratory amplitude of diaphragm on the 4D‐MR scans to the diaphragm profiles. To evaluate reconstructed 4D‐MR images (ten 3D‐MR images), sagittal images on ten 3D‐MR images under free breathing (FB) and respiratory guidance (GB) were compared with diaphragm profile images (reference) from 13 healthy volunteers. Results Forty‐four 4D‐MR scan datasets were successfully reconstructed without distinct respiratory‐related motion artefacts even with the presence of breathing variation. The differences in diaphragm profiles between the reference and corresponding reconstructed images in the mean of root mean square were similar between FB (3.5 mm) and GB (3.0 mm), confirming that the 4D‐MRI reconstruction method was effective even with significant breathing variation. Conclusions The diaphragm profiles were utilized to reconstruct 4D‐MR images with spatial reliability and a fixed scan time under FB and GB. Our method can provide reliable 4D information of thoracic and abdominal regions for MRI‐guided radiotherapy.

4D MRI: Robust sorting of free breathing MRI slices for use in interventional settings

“…Cai et al [15] use the body area. Lee et al [16] use sagittal diaphragm profiles and reconstruct one breathing cycle. Tong et al [17] propose a graph-based sorting where the weights are based on image information and semi-automatic assigned respiratory phase although, they are only able to reconstruct one best breathing cycle and not a variety of breathing cycles.…”

An MRI framework for respiratory motion modelling validation

“…4 Indeed, MRI allows safely repeated scans with no additional ionizing radiation, enhanced soft-tissue contrast and fast imaging sequences. 5,6 To investigate intra-and interfraction motion variability and cycle-to-cycle variations, respiratory-correlated (RC) 4-dimensional MRI (4DMRI) can be retrospectively reconstructed, [7][8][9][10][11] and time-resolved (TR) 2D cine-MRI acquired, but limitations are still present in gaining 3D TR images with sufficient spatiotemporal resolution. 6,12 To estimate 3D TR information of tumour and surrounding organs, respiratory motion modelling techniques have been explored from the 2D or 1D information available.…”

Evaluation of MRI-derived surrogate signals to model respiratory motion

“…Due to the inherent trade-off between temporal resolution, spatial resolution and field-of-view in MR imaging, most of the proposed methods are respiratory-correlated 4D-MRI. These techniques use data acquired from different respiratory cycles to produce retrospectively-sorted 3D volumes at different respiratory states (Cai et al 2011 , Deng et al 2016 , Li et al 2017 , Mickevicius and Paulson 2017 , Han et al 2017 , van de Lindt et al 2018 , Lee et al 2019 , van Kesteren et al 2019 ). Real-time guidance information cannot be provided since the images are not available until all data have been acquired.…”