The future of MRI in radiation therapy: Challenges and opportunities for the MR community

Radiation therapy is a major component of cancer treatment pathways worldwide. The main aim of this treatment is to achieve tumor control through the delivery of ionizing radiation while preserving healthy tissues for minimal radiation toxicity. Because radiation therapy relies on accurate localization of the target and surrounding tissues, imaging plays a crucial role throughout the treatment chain. In the treatment planning phase, radiological images are essential for defining target volumes and organs-at-risk, as well as providing elemental composition (e.g., electron density) information for radiation dose calculations. At treatment, onboard imaging informs patient setup and could be used to guide radiation dose placement for sites affected by motion. Imaging is also an important tool for treatment response assessment and treatment plan adaptation. MRI, with its excellent soft tissue contrast and capacity to probe functional tissue properties, holds great untapped potential for transforming treatment paradigms in radiation therapy. The MR in Radiation Therapy ISMRM Study Group was established to provide a forum within the MR community to discuss the unmet needs and fuel opportunities for further advancement of MRI for radiation therapy applications. During the summer of 2021, the study group organized its first virtual workshop, attended by a diverse international group of clinicians, scientists, and clinical physicists, to explore our predictions for the future of MRI in radiation therapy for the next 25 years. This article reviews the main findings from the event and considers the opportunities and challenges of reaching our vision for the future in this expanding field.

Rapid 4D-MRI reconstruction using a deep radial convolutional neural network: Dracula

BACKGROUND AND PURPOSE

4D and midposition MRI could inform plan adaptation in lung and abdominal MR-guided radiotherapy. We present deep learning-based solutions to overcome long 4D-MRI reconstruction times while maintaining high image quality and short scan times.

METHODS

Two 3D U-net deep convolutional neural networks were trained to accelerate the 4D joint MoCo-HDTV reconstruction. For the first network, gridded and joint MoCo-HDTV-reconstructed 4D-MRI were used as input and target data, respectively, whereas the second network was trained to directly calculate the midposition image. For both networks, input and target data had dimensions of 256 × 256 voxels (2D) and 16 respiratory phases. Deep learning-based MRI were verified against joint MoCo-HDTV-reconstructed MRI using the structural similarity index (SSIM) and the naturalness image quality evaluator (NIQE). Moreover, two experienced observers contoured the gross tumour volume and scored the images in a blinded study.

RESULTS

For 12 subjects, previously unseen by the networks, high-quality 4D and midposition MRI (1.25 × 1.25 × 3.3 mm3) were each reconstructed from gridded images in only 28 seconds per subject. Excellent agreement was found between deep-learning-based and joint MoCo-HDTV-reconstructed MRI (average SSIM ≥ 0.96, NIQE scores 7.94 and 5.66). Deep-learning-based 4D-MRI were clinically acceptable for target and organ-at-risk delineation. Tumour positions agreed within 0.7 mm on midposition images.

CONCLUSION

Our results suggest that the joint MoCo-HDTV and midposition algorithms can each be approximated by a deep convolutional neural network. This rapid reconstruction of 4D and midposition MRI facilitates online treatment adaptation in thoracic or abdominal MR-guided radiotherapy.

Synthetic 4D-CT of the thorax for treatment plan adaptation on MR-guided radiotherapy systems

MR-guided radiotherapy treatment planning utilises the high soft-tissue contrast of MRI to reduce uncertainty in delineation of the target and organs at risk. Replacing 4D-CT with MRI-derived synthetic 4D-CT would support treatment plan adaptation on hybrid MR-guided radiotherapy systems for inter- and intrafractional differences in anatomy and respiration, whilst mitigating the risk of CT to MRI registration errors. Three methods were devised to calculate synthetic 4D and midposition (time-weighted mean position of the respiratory cycle) CT from 4D-T1w and Dixon MRI. The first approach employed intensity-based segmentation of Dixon MRI for bulk-density assignment (sCTD). The second step added spine density information using an atlas of CT and Dixon MRI (sCTDS). The third iteration used a polynomial function relating Hounsfield units and normalised T1w image intensity to account for variable lung density (sCTDSL). Motion information in 4D-T1w MRI was applied to generate synthetic CT in midposition and in twenty respiratory phases. For six lung cancer patients, synthetic 4D-CT was validated against 4D-CT in midposition by comparison of Hounsfield units and dose-volume metrics. Dosimetric differences found by comparing sCTD,DS,DSL and CT were evaluated using a Wilcoxon signed-rank test (p   =  0.05). Compared to sCTD and sCTDS, planning on sCTDSL significantly reduced absolute dosimetric differences in the planning target volume metrics to less than 98 cGy (1.7% of the prescribed dose) on average. When comparing sCTDSL and CT, average radiodensity differences were within 97 Hounsfield units and dosimetric differences were significant only for the planning target volume D99% metric. All methods produced clinically acceptable results for the organs at risk in accordance with the UK SABR consensus guidelines and the LungTech EORTC phase II trial. The overall good agreement between sCTDSL and CT demonstrates the feasibility of employing synthetic 4D-CT for plan adaptation on hybrid MR-guided radiotherapy systems.

Super-resolution T2-weighted 4D MRI for image guided radiotherapy

BACKGROUND AND PURPOSE

The superior soft-tissue contrast of 4D-T2w MRI motivates its use for delineation in radiotherapy treatment planning. We address current limitations of slice-selective implementations, including thick slices and artefacts originating from data incompleteness and variable breathing.

MATERIALS AND METHODS

A method was developed to calculate midposition and 4D-T2w images of the whole thorax from continuously acquired axial and sagittal 2D-T2w MRI (1.5 × 1.5 × 5.0 mm3). The method employed image-derived respiratory surrogates, deformable image registration and super-resolution reconstruction. Volunteer imaging and a respiratory motion phantom were used for validation. The minimum number of dynamic acquisitions needed to calculate a representative midposition image was investigated by retrospectively subsampling the data (10-30 dynamic acquisitions).

RESULTS

Super-resolution 4D-T2w MRI (1.0 × 1.0 × 1.0 mm3, 8 respiratory phases) did not suffer from data incompleteness and exhibited reduced stitching artefacts compared to sorted multi-slice MRI. Experiments using a respiratory motion phantom and colour-intensity projection images demonstrated a minor underestimation of the motion range. Midposition diaphragm differences in retrospectively subsampled acquisitions were <1.1 mm compared to the full dataset. 10 dynamic acquisitions were found sufficient to generate midposition MRI.

CONCLUSIONS

A motion-modelling and super-resolution method was developed to calculate high quality 4D/midposition T2w MRI from orthogonal 2D-T2w MRI.

T2-Weighted 4D Magnetic Resonance Imaging for Application in Magnetic Resonance-Guided Radiotherapy Treatment Planning

OBJECTIVES

The aim of this study was to develop and verify a method to obtain good temporal resolution T2-weighted 4-dimensional (4D-T2w) magnetic resonance imaging (MRI) by using motion information from T1-weighted 4D (4D-T1w) MRI, to support treatment planning in MR-guided radiotherapy.

MATERIALS AND METHODS

Ten patients with primary non-small cell lung cancer were scanned at 1.5 T axially with a volumetric T2-weighted turbo spin echo sequence gated to exhalation and a volumetric T1-weighted stack-of-stars spoiled gradient echo sequence with golden angle spacing acquired in free breathing. From the latter, 20 respiratory phases were reconstructed using the recently developed 4D joint MoCo-HDTV algorithm based on the self-gating signal obtained from the k-space center. Motion vector fields describing the respiratory cycle were obtained by deformable image registration between the respiratory phases and projected onto the T2-weighted image volume. The resulting 4D-T2w volumes were verified against the 4D-T1w volumes: an edge-detection method was used to measure the diaphragm positions; the locations of anatomical landmarks delineated by a radiation oncologist were compared and normalized mutual information was calculated to evaluate volumetric image similarity.

RESULTS

High-resolution 4D-T2w MRI was obtained. Respiratory motion was preserved on calculated 4D-T2w MRI, with median diaphragm positions being consistent with less than 6.6 mm (2 voxels) for all patients and less than 3.3 mm (1 voxel) for 9 of 10 patients. Geometrical positions were coherent between 4D-T1w and 4D-T2w MRI as Euclidean distances between all corresponding anatomical landmarks agreed to within 7.6 mm (Euclidean distance of 2 voxels) and were below 3.8 mm (Euclidean distance of 1 voxel) for 355 of 470 pairs of anatomical landmarks. Volumetric image similarity was commensurate between 4D-T1w and 4D-T2w MRI, as mean percentage differences in normalized mutual information (calculated over all respiratory phases and patients), between corresponding respiratory phases of 4D-T1w and 4D-T2w MRI and the tie-phase of 4D-T1w and 3-dimensional T2w MRI, were consistent to 0.41% ± 0.37%. Four-dimensional T2w MRI displayed tumor extent, structure, and position more clearly than corresponding 4D-T1w MRI, especially when mobile tumor sites were adjacent to organs at risk.

CONCLUSIONS

A methodology to obtain 4D-T2w MRI that retrospectively applies the motion information from 4D-T1w MRI to 3-dimensional T2w MRI was developed and verified. Four-dimensional T2w MRI can assist clinicians in delineating mobile lesions that are difficult to define on 4D-T1w MRI, because of poor tumor-tissue contrast.

First- and second-order motion perception in Gabor micropattern stimuli: psychophysics and computational modelling

This paper examines the perception of first- and second-order motion in human vision. In an extension of previous work by Boulton and Baker [J.B. Boulton, C.L. Baker, Motion detection is dependent on spatial frequency not size, Vision Res., 31 (1991) 77-87; J.B. Boulton, C.L. Baker, Different parameters control motion perception above and below a critical density, Vision Res., 33 (1993) 1803-1811], the direction of two-frame apparent motion is measured for stimuli composed of Gabor or Gaussian micropatterns. Three conditions are investigated. Condition 1 is that used by Boulton and Baker, in which motion is defined by the displacement of Gabor micropatterns. In condition 2, motion is defined by the displacement of Gaussian micropatterns. In condition 3, the envelopes of Gabor micropatterns are displaced while their carriers remain static. Using sparsely distributed micropatterns, direction judgements in all three conditions are determined by the spacing of the micropatterns. With a dense stimulus, direction judgements vary as a function of displacement in qualitatively different ways for the three conditions. The psychophysical results are predicted by a two-channel computational model. In one channel, motion is calculated directly from stimulus luminance, while in the other it is preceded by a texture-grabbing operation. The relative activities of the two channels dictates which governs direction judgements for any given stimulus.