A voxel-based investigation for MRI-only radiotherapy of the brain using ultra short echo times

MRI-Guided Radiation Therapy for Prostate Cancer: The Next Frontier in Ultrahypofractionation

Technological advances in MRI-guided radiation therapy (MRIgRT) have improved real-time visualization of the prostate and its surrounding structures over CT-guided radiation therapy. Seminal studies have demonstrated safe dose escalation achieved through ultrahypofractionation with MRIgRT due to planning target volume (PTV) margin reduction and treatment gating. On-table adaptation with MRI-based technologies can also incorporate real-time changes in target shape and volume and can reduce high doses of radiation to sensitive surrounding structures that may move into the treatment field. Ongoing clinical trials seek to refine ultrahypofractionated radiotherapy treatments for prostate cancer using MRIgRT. Though these studies have the potential to demonstrate improved biochemical control and reduced side effects, limitations concerning patient treatment times and operational workflows may preclude wide adoption of this technology outside of centers of excellence. In this review, we discuss the advantages and limitations of MRIgRT for prostate cancer, as well as clinical trials testing the efficacy and toxicity of ultrafractionation in patients with localized or post-prostatectomy recurrent prostate cancer.

Adoption of Hybrid MRI-Linac Systems for the Treatment of Brain Tumors: A Systematic Review of the Current Literature Regarding Clinical and Technical Features

BACKGROUND

Possible advantages of magnetic resonance (MR)-guided radiation therapy (MRgRT) for the treatment of brain tumors include improved definition of treatment volumes and organs at risk (OARs) that could allow margin reductions, resulting in limited dose to the OARs and/or dose escalation to target volumes. Recently, hybrid systems integrating a linear accelerator and an magnetic resonance imaging (MRI) scan (MRI-linacs, MRL) have been introduced, that could potentially lead to a fully MRI-based treatment workflow.

METHODS

We performed a systematic review of the published literature regarding the adoption of MRL for the treatment of primary or secondary brain tumors (last update November 3, 2022), retrieving a total of 2487 records; after a selection based on title and abstracts, the full text of 74 articles was analyzed, finally resulting in the 52 papers included in this review.

RESULTS AND DISCUSSION

Several solutions have been implemented to achieve a paradigm shift from CT-based radiotherapy to MRgRT, such as the management of geometric integrity and the definition of synthetic CT models that estimate electron density. Multiple sequences have been optimized to acquire images with adequate quality with on-board MR scanner in limited times. Various sophisticated algorithms have been developed to compensate the impact of magnetic field on dose distribution and calculate daily adaptive plans in a few minutes with satisfactory dosimetric parameters for the treatment of primary brain tumors and cerebral metastases. Dosimetric studies and preliminary clinical experiences demonstrated the feasibility of treating brain lesions with MRL.

CONCLUSIONS

The adoption of an MRI-only workflow is feasible and could offer several advantages for the treatment of brain tumors, including superior image quality for lesions and OARs and the possibility to adapt the treatment plan on the basis of daily MRI. The growing body of clinical data will clarify the potential benefit in terms of toxicity and response to treatment.

AAPM Task Group Report 303 endorsed by the ABS: MRI Implementation in HDR Brachytherapy-Considerations from Simulation to Treatment

The Task Group (TG) on Magnetic Resonance Imaging (MRI) Implementation in High Dose Rate (HDR) Brachytherapy - Considerations from Simulation to Treatment, TG 303, was constituted by the American Association of Physicists in Medicine's (AAPM's) Science Council under the direction of the Therapy Physics Committee, the Brachytherapy Subcommittee, and the Working Group on Brachytherapy Clinical Applications. The TG was charged with developing recommendations for commissioning, clinical implementation, and on-going quality assurance (QA). Additionally, the TG was charged with describing HDR brachytherapy (BT) workflows and evaluating practical consideration that arise when implementing MR imaging. For brevity, the report is focused on the treatment of gynecologic and prostate cancer. The TG report provides an introduction and rationale for MRI implementation in BT, a review of previous publications on topics including available applicators, clinical trials, previously published BT related TG reports, and new image guided recommendations beyond CT based practices. The report describes MRI protocols and methodologies, including recommendations for the clinical implementation and logical considerations for MR imaging for HDR BT. Given the evolution from prescriptive to risk-based QA,¹ an example of a risk-based analysis using MRI-based, prostate HDR BT is presented. In summary, the TG report is intended to provide clear and comprehensive guidelines and recommendations for commissioning, clinical implementation, and QA for MRI-based HDR BT that may be utilized by the medical physics community to streamline this process. This report is endorsed by the American Brachytherapy Society (ABS). This article is protected by copyright. All rights reserved.

GAN for synthesizing CT from T2-weighted MRI data towards MR-guided radiation treatment

OBJECTIVE

In medical domain, cross-modality image synthesis suffers from multiple issues , such as context-misalignment, image distortion, image blurriness, and loss of details. The fundamental objective behind this study is to address these issues in estimating synthetic Computed tomography (sCT) scans from T2-weighted Magnetic Resonance Imaging (MRI) scans to achieve MRI-guided Radiation Treatment (RT).

MATERIALS AND METHODS

We proposed a conditional generative adversarial network (cGAN) with multiple residual blocks to estimate sCT from T2-weighted MRI scans using 367 paired brain MR-CT images dataset. Few state-of-the-art deep learning models were implemented to generate sCT including Pix2Pix model, U-Net model, autoencoder model and their results were compared, respectively.

RESULTS

Results with paired MR-CT image dataset demonstrate that the proposed model with nine residual blocks in generator architecture results in the smallest mean absolute error (MAE) value of [Formula: see text], and mean squared error (MSE) value of [Formula: see text], and produces the largest Pearson correlation coefficient (PCC) value of [Formula: see text], SSIM value of [Formula: see text] and peak signal-to-noise ratio (PSNR) value of [Formula: see text], respectively. We qualitatively evaluated our result by visual comparisons of generated sCT to original CT of respective MRI input.

DISCUSSION

The quantitative and qualitative comparison of this work demonstrates that deep learning-based cGAN model can be used to estimate sCT scan from a reference T2 weighted MRI scan. The overall accuracy of our proposed model outperforms different state-of-the-art deep learning-based models.

Pseudo computed tomography image generation from brain magnetic resonance image using integration of PCA & DCNN-UNET: A comparative analysis

MRI-Only Radiation (RT) now avoids some of the issues associated with employing Computed Tomography(CT) in RT chains, such as MRI registration to a separate CT, excess dosage administration, and the cost of recurrent imaging. The fact that MRI signal intensities are unrelated to the biological tissue’s attenuation coefficient poses a problem. This raises workloads, creates uncertainty as a result of the required inter-modality image registrations, and exposes patients to needless radiation. While using only MRI would be preferable, a method for estimating a pseudo-CT (pCT)or synthetic-CT(sCT) for producing electron density maps and patient positioning reference images is required. As Deep Learning(DL) is revolutionized in so many fields these days, an effective and accurate model is required for generating pCT from MRI. So, this paper depicts an efficient DL model in which the following are the stages: a) Data Acquisition where CT and MRI images are collected b) preprocessing these to avoid the anomalies and noises using techniques like outlier elimination, data smoothening and data normalizing c) feature extraction and selection using Principal Component Analysis (PCA) & regression method d) generating pCT from MRI using Deep Convolutional Neural Network and UNET (DCNN-UNET). We here compare both feature extraction (PCA) and classification model (DCNN-UNET) with other methods such as Discrete Wavelet Tranform(DWT), Independent Component Analysis(ICA), Fourier Transform and VGG16, ResNet, AlexNet, DenseNet, CNN (Convolutional Neural Network)respectively. The performance measures used to evaluate these models are Dice Coefficient(DC), Structured Similarity Index Measure(SSIM), Mean Absolute Error(MAE), Mean Squared Error(MSE), Accuracy, Computation Time in which our proposed system outperforms better with 0.94±0.02 over other state-of-art models.

Multi parametric magnetic resonance imaging for radiation treatment planning

In recent years, multi-parametric magnetic resonance imaging (MpMRI) has played a major role in radiation therapy treatment planning. The superior soft tissue contrast, functional or physiological imaging capabilities and the flexibility of site-specific image sequence development has placed MpMRI at the forefront. In this article, the present status of MpMRI for external beam radiation therapy planning is reviewed. Common MpMRI sequences, preprocessing and QA strategies are briefly discussed, and various image registration techniques and strategies are addressed. Image segmentation methods including automatic segmentation and deep learning techniques for organs at risk and target delineation are reviewed. Due to the advancement in MRI guided online adaptive radiotherapy, treatment planning considerations addressing MRI only planning are also discussed. This article is protected by copyright. All rights reserved.

Artificial intelligence-based bone-enhanced magnetic resonance image-a computed tomography/magnetic resonance image composite image modality in nasopharyngeal carcinoma radiotherapy

Background

In the radiotherapy of nasopharyngeal carcinoma (NPC), magnetic resonance imaging (MRI) is widely used to delineate tumor area more accurately. While MRI offers the higher soft tissue contrast, patient positioning and couch correction based on bony image fusion of computed tomography (CT) is also necessary. There is thus an urgent need to obtain a high image contrast between bone and soft tissue to facilitate target delineation and patient positioning for NPC radiotherapy. In this paper, our aim is to develop a novel image conversion between the CT and MRI modalities to obtain clear bone and soft tissue images simultaneously, here called bone-enhanced MRI (BeMRI).

Methods

Thirty-five patients were retrospectively selected for this study. All patients underwent clinical CT simulation and 1.5T MRI within the same week in Shenzhen Second People's Hospital. To synthesize BeMRI, two deep learning networks, U-Net and CycleGAN, were constructed to transform MRI to synthetic CT (sCT) images. Each network used 28 patients' images as the training set, while the remaining 7 patients were used as the test set (~1/5 of all datasets). The bone structure from the sCT was then extracted by the threshold-based method and embedded in the corresponding part of the MRI image to generate the BeMRI image. To evaluate the performance of these networks, the following metrics were applied: mean absolute error (MAE), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR).

Results

In our experiments, both deep learning models achieved good performance and were able to effectively extract bone structure from MRI. Specifically, the supervised U-Net model achieved the best results with the lowest overall average MAE of 125.55 (P<0.05) and produced the highest SSIM of 0.89 and PSNR of 23.84. These results indicate that BeMRI can display bone structure in higher contrast than conventional MRI.

Conclusions

A new image modality BeMRI, which is a composite image of CT and MRI, was proposed. With high image contrast of both bone structure and soft tissues, BeMRI will facilitate tumor localization and patient positioning and eliminate the need to frequently check between separate MRI and CT images during NPC radiotherapy.

Synthesis of pseudo-CT images from pelvic MRI images based on MD-CycleGAN model for radiotherapy

OBJECTIVE

A multi-discriminator-based cycle generative adversarial network (MD-CycleGAN) model was proposed to synthesize higher-quality pseudo-CT from MRI.

APPROACH

The MRI and CT images obtained at the simulation stage with cervical cancer were selected to train the model. The generator adopted the DenseNet as the main architecture. The local and global discriminators based on convolutional neural network jointly discriminated the authenticity of the input image data. In the testing phase, the model was verified by four-fold cross-validation method. In the prediction stage, the data were selected to evaluate the accuracy of the pseudo-CT in anatomy and dosimetry, and they were compared with the pseudo-CT synthesized by GAN with generator based on the architectures of ResNet, sU-Net, and FCN.

MAIN RESULTS

There are significant differences(P<0.05) in the four-fold-cross validation results on peak signal-to-noise ratio and structural similarity index metrics between the pseudo-CT obtained based on MD-CycleGAN and the ground truth CT (CTgt). The pseudo-CT synthesized by MD-CycleGAN had closer anatomical information to the CTgt with root mean square error of 47.83±2.92 HU and normalized mutual information value of 0.9014±0.0212 and mean absolute error value of 46.79±2.76 HU. The differences in dose distribution between the pseudo-CT obtained by MD-CycleGAN and the CTgt were minimal. The mean absolute dose errors of Dosemax, Dosemin and Dosemean based on the planning target volume were used to evaluate the dose uncertainty of the four pseudo-CT. The u-values of the Wilcoxon test were 55.407, 41.82 and 56.208, and the differences were statistically significant. The 2%/2 mm-based gamma pass rate (%) of the proposed method was 95.45±1.91, and the comparison methods (ResNet_GAN, sUnet_GAN and FCN_GAN) were 93.33±1.20, 89.64±1.63 and 87.31±1.94, respectively.

SIGNIFICANCE

The pseudo-CT obtained based on MD-CycleGAN have higher imaging quality and are closer to the CTgt in terms of anatomy and dosimetry than other GAN models.

Virtual magnetic resonance lumbar spine images generated from computed tomography images using conditional generative adversarial networks

INTRODUCTION

The aim of this study was to generate virtual Magnetic resonance (MR) from computed tomography (CT) using conditional generative adversarial networks (cGAN).

METHODS

We selected examinations from 22 adults who obtained their CT and MR lumbar spine examinations. Overall, 4 examinations were used as test data, and 18 examinations were used as training data. A cGAN was trained to generate virtual MR images from the CT images using the corresponding MR images as targets. After training, the generated virtual MR images from test data in epochs 1, 10, 50, 100, 500, and 1000 were compared with the original ones using the mean square error (MSE) and structural similarity index (SSIM). Additionally, two radiologists also performed qualitative assessments.

RESULTS

The MSE of the virtual MR images decreased as the epoch of the cGANs increased from the original CT images: 8876.7 ± 1192.9 (original CT), 1567.5 ± 433.9 (Epoch 1), 1242.4 ± 442.0 (Epoch 10), 1065.8 ± 478.1 (Epoch 50), 1276.1 ± 718.9 (Epoch 100), 1046.7 ± 488.2 (Epoch 500), and 1031.7 ± 400.0 (Epoch 1000). No considerable differences were observed in the qualitative evaluation between the virtual MR images and the original ones, except in the structure of the spinal canal.

CONCLUSION

Virtual MR lumbar spine images using cGANs could be a feasible technique to generate near-MR images from CT without MR examinations for evaluation of the vertebral body and intervertebral disc.

IMPLICATIONS FOR PRACTICE

Virtual MR lumbar spine images using cGANs can offer virtual CT images with sufficient quality for attenuation correction for PET or dose planning in radiotherapy.

Acoustic simulation for transcranial focused ultrasound using GAN-based synthetic CT

Transcranial focused ultrasound (tFUS) is a promising non-invasive technique for treating neurological and psychiatric disorders. One of the challenges for tFUS is the disruption of wave propagation through the skull. Consequently, despite the risks associated with exposure to ionizing radiation, computed tomography (CT) is required to estimate the acoustic transmission through the skull. This study aims to generate synthetic CT (sCT) from T1-weighted magnetic resonance imaging (MRI) and investigate its applicability to tFUS acoustic simulation. We trained a 3D conditional generative adversarial network (3D-cGAN) with 15 subjects. We then assessed image quality with 15 test subjects: mean absolute error (MAE) = 85.72± 9.50 HU (head) and 280.25±24.02 HU (skull), dice coefficient similarity (DSC) = 0.88±0.02 (skull). In terms of skull density ratio (SDR) and skull thickness (ST), no significant difference was found between sCT and real CT (rCT). When the acoustic simulation results of rCT and sCT were compared, the intracranial peak acoustic pressure ratio was found to be less than 4%, and the distance between focal points less than 1 mm.

Synthesizing PET/MR (T1-weighted) images from non-attenuation-corrected PET images

Positron emission tomography (PET) imaging can be used for early detection, diagnosis and postoperative patient monitoring of many diseases. Traditional PET imaging requires not only additional computed tomography (CT) imaging or magnetic resonance imaging (MR) to provide anatomical information but also attenuation correction (AC) map calculation based on CT images or MR images for accurate quantitative estimation. During a patient's treatment, PET/CT or PET/MR scans are inevitably repeated many times, leading to additional doses of ionizing radiation (CT scans) and additional economic and time costs (MR scans). To reduce adverse effects while obtaining high-quality PET/MR images in the course of a patient's treatment, especially in the stage of evaluating the effect of postoperative treatment, in this work, we propose a new method based on deep learning, which can directly obtain synthetic attenuation-corrected PET (sAC PET) and synthetic T1-weighted MR (sMR) images based only on non-attenuation-corrected PET (NAC PET) images. Our model, based on the Wasserstein generative adversarial network, first removes noise and artifacts from the NAC PET images to generate sAC PET images and then generates sMR images from the obtained sAC PET images. To evaluate the performance of this generative model, we evaluated it on paired PET/MR images from a total of eighty clinical patients. Based on qualitative and quantitative analysis, the generated sAC PET and sMR images showed a high degree of similarity to the real AC PET and real MR images. These results indicated that our proposed method can reduce the frequency of additional anatomical imaging scans during PET imaging and has great potential in improving doctors' clinical diagnosis efficiency, saving patients' economic expenditure and reducing the radiation risk brought by CT scanning.

Signal Enhancement of Low Magnetic Field Magnetic Resonance Image Using a Conventional- and Cyclic-Generative Adversarial Network Models With Unpaired Image Sets

In this study, the signal enhancement ratio of low-field magnetic resonance (MR) images was investigated using a deep learning-based algorithm. Unpaired image sets (0.06 Tesla and 1.5 Tesla MR images for different patients) were used in this study following three steps workflow. In the first step, the deformable registration of a 1.5 Tesla MR image into a 0.06 Tesla MR image was performed to ensure that the shapes of the unpaired set matched. In the second step, a cyclic-generative adversarial network (GAN) was used to generate a synthetic MR image of the original 0.06 Tesla MR image based on the deformed or original 1.5 Tesla MR image. Finally, an enhanced 0.06 Tesla MR image could be generated using the conventional-GAN with the deformed or synthetic MR image. The results from the optimized flow and enhanced MR images showed significant signal enhancement of the anatomical view, especially in the nasal septum, inferior nasal choncha, nasopharyngeal fossa, and eye lens. The signal enhancement ratio, signal-to-noise ratio (SNR) and correlation factor between the original and enhanced MR images were analyzed for the evaluation of the image quality. A combined method using conventional- and cyclic-GANs is a promising approach for generating enhanced MR images from low-magnetic-field MR.

Clinical feasibility of MR-generated synthetic CT images of the cervical spine: Diagnostic performance for detection of OPLL and comparison of CT number

We aimed to determine the incremental value of magnetic resonance generated synthetic computed tomography (MRCT), evaluate cervical ossification of the posterior longitudinal ligament (OPLL), and compare the computed tomography (CT) numbers between MRCT and conventional CT.Twenty-two patients who underwent magnetic resonance imaging (MRI) with MRCT protocols and CT were enrolled. MRCT images were generated from 3D-T2-weighted imaging, 3D-pointwise-encoding time reduction with radial acquisition, 3D-T1-Dixon, and 3D-time-of-flight sequences. Two radiologists independently evaluated the presence of OPLL at each cervical spine level during sessions 1 (MRI alone) and 2 (MRI + MRCT). CT was the reference standard for the presence of OPLL. One reader measured the mean CT number of the vertebral body and spinous process at each cervical spine level in the MRCT and CT images.Sensitivity for the detection of OPLL was markedly higher in session 2 (MRI + MRCT) than in session 1 (MRI alone), as measured by both readers (47% vs. 90%, reader 1; 63% vs. 93%, reader 2). The mean CT number of MRCT and CT showed a moderate to strong positive correlation (ρ = .42-.72, P < .001).The combined use of MRCT and MRI showed improved sensitivity for the evaluation of cervical OPLL. The mean CT number of MRCT and CT showed a positive correlation.

Dosimetric evaluation of synthetic CT image generated using a neural network for MR-only brain radiotherapy

PURPOSE AND BACKGROUND

The magnetic resonance (MR)-only radiotherapy workflow is urged by the increasing use of MR image for the identification and delineation of tumors, while a fast generation of synthetic computer tomography (sCT) image from MR image for dose calculation remains one of the key challenges to the workflow. This study aimed to develop a neural network to generate the sCT in brain site and evaluate the dosimetry accuracy.

MATERIALS AND METHODS

A generative adversarial network (GAN) was developed to translate T1-weighted MRI to sCT. First, the "U-net" shaped encoder-decoder network with some image translation-specific modifications was trained to generate sCT, then the discriminator network was adversarially trained to distinguish between synthetic and real CT images. We enrolled 37 brain cancer patients acquiring both CT and MRI for treatment position simulation. Twenty-seven pairs of 2D T1-weighted MR images and rigidly registered CT image were used to train the GAN model, and the remaining 10 pairs were used to evaluate the model performance through the metric of mean absolute error. Furthermore, the clinical Volume Modulated Arc Therapy plan was calculated on both sCT and real CT, followed by gamma analysis and comparison of dose-volume histogram.

RESULTS

On average, only 15 s were needed to generate one sCT from one T1-weighted MRI. The mean absolute error between synthetic and real CT was 60.52 ± 13.32 Housefield Unit over 5-fold cross validation. For dose distribution on sCT and CT, the average pass rates of gamma analysis using the 3%/3 mm and 2%/2 mm criteria were 99.76% and 97.25% over testing patients, respectively. For parameters of dose-volume histogram for both target and organs at risk, no significant differences were found between both plans.

CONCLUSION

The GAN model can generate synthetic CT from one single MRI sequence within seconds, and a state-of-art accuracy of CT number and dosimetry was achieved.

Synthetic computed tomography data allows for accurate absorbed dose calculations in a magnetic resonance imaging only workflow for head and neck radiotherapy

•

The geometry of the synthetic CT is comparable to the CT in the H&N region.

•

Synthetic CT in the H&N region provides similar absorbed dose calculation as the CT.

•

Absorbed dose calculations in the dental region could benefit from using synthetic CT.

Background and purpose

Few studies on magnetic resonance imaging (MRI) only head and neck radiation treatment planning exist, and none using a generally available software. The aim of this study was to evaluate the accuracy of absorbed dose for head and neck synthetic computed tomography data (sCT) generated by a commercial convolutional neural network-based algorithm.

Materials and methods

For 44 head and neck cancer patients, sCT were generated and the geometry was validated against computed tomography data (CT). The clinical CT based treatment plan was transferred to the sCT and recalculated without re-optimization, and differences in relative absorbed dose were determined for dose-volume-histogram (DVH) parameters and the 3D volume.

Results

For overall body, the results of the geometric validation were (Mean ± 1sd): Mean error −5 ± 10 HU, mean absolute error 67 ± 14 HU, Dice similarity coefficient 0.98 ± 0.05, and Hausdorff distance difference 4.2 ± 1.7 mm. Water equivalent depth difference for region Th1-C7, mid mandible and mid nose were −0.3 ± 3.4, 1.1 ± 2.0 and 0.7 ± 3.8 mm respectively. The maximum mean deviation in absorbed dose for all DVH parameters was 0.30% (0.12 Gy). The absorbed doses were considered equivalent (p-value < 0.001) and the mean 3D gamma passing rate was 99.4 (range: 95.7–99.9%).

Conclusions

The convolutional neural network-based algorithm generates sCT which allows for accurate absorbed dose calculations for MRI-only head and neck radiation treatment planning. The sCT allows for statistically equivalent absorbed dose calculations compared to CT based radiotherapy.

Synthetic CT Generation of the Pelvis in Patients With Cervical Cancer: A Single Input Approach Using Generative Adversarial Network

Multi-modality imaging constitutes a foundation of precision medicine, especially in oncology where reliable and rapid imaging techniques are needed in order to insure adequate diagnosis and treatment. In cervical cancer, precision oncology requires the acquisition of 18F-labeled 2-fluoro-2-deoxy-D-glucose (FDG) positron emission tomography (PET), magnetic resonance (MR), and computed tomography (CT) images. Thereafter, images are co-registered to derive electron density attributes required for FDG-PET attenuation correction and radiation therapy planning. Nevertheless, this traditional approach is subject to MR-CT registration defects, expands treatment expenses, and increases the patient's radiation exposure. To overcome these disadvantages, we propose a new framework for cross-modality image synthesis which we apply on MR-CT image translation for cervical cancer diagnosis and treatment. The framework is based on a conditional generative adversarial network (cGAN) and illustrates a novel tactic that addresses, simplistically but efficiently, the paradigm of vanishing gradient vs. feature extraction in deep learning. Its contributions are summarized as follows: 1) The approach -termed sU-cGAN-uses, for the first time, a shallow U-Net (sU-Net) with an encoder/decoder depth of 2 as generator; 2) sU-cGAN's input is the same MR sequence that is used for radiological diagnosis, i.e. T2-weighted, Turbo Spin Echo Single Shot (TSE-SSH) MR images; 3) Despite limited training data and a single input channel approach, sU-cGAN outperforms other state of the art deep learning methods and enables accurate synthetic CT (sCT) generation. In conclusion, the suggested framework should be studied further in the clinical settings. Moreover, the sU-Net model is worth exploring in other computer vision tasks.

Magnetic resonance imaging-based synthetic computed tomography of the lumbar spine for surgical planning: a clinical proof-of-concept

OBJECTIVE

Computed tomography scanning of the lumbar spine incurs a radiation dose ranging from 3.5 mSv to 19.5 mSv as well as relevant costs and is commonly necessary for spinal neuronavigation. Mitigation of the need for treatment-planning CT scans in the presence of MRI facilitated by MRI-based synthetic CT (sCT) would revolutionize navigated lumbar spine surgery. The authors aim to demonstrate, as a proof of concept, the capability of deep learning-based generation of sCT scans from MRI of the lumbar spine in 3 cases and to evaluate the potential of sCT for surgical planning.

METHODS

Synthetic CT reconstructions were made using a prototype version of the "BoneMRI" software. This deep learning-based image synthesis method relies on a convolutional neural network trained on paired MRI-CT data. A specific but generally available 4-minute 3D radiofrequency-spoiled T1-weighted multiple gradient echo MRI sequence was supplemented to a 1.5T lumbar spine MRI acquisition protocol.

RESULTS

In the 3 presented cases, the prototype sCT method allowed voxel-wise radiodensity estimation from MRI, resulting in qualitatively adequate CT images of the lumbar spine based on visual inspection. Normal as well as pathological structures were reliably visualized. In the first case, in which a spiral CT scan was available as a control, a volume CT dose index (CTDIvol) of 12.9 mGy could thus have been avoided. Pedicle screw trajectories and screw thickness were estimable based on sCT findings.

CONCLUSIONS

The evaluated prototype BoneMRI method enables generation of sCT scans from MRI images with only minor changes in the acquisition protocol, with a potential to reduce workflow complexity, radiation exposure, and costs. The quality of the generated CT scans was adequate based on visual inspection and could potentially be used for surgical planning, intraoperative neuronavigation, or for diagnostic purposes in an adjunctive manner.

MR-guided proton therapy: a review and a preview

Background

The targeting accuracy of proton therapy (PT) for moving soft-tissue tumours is expected to greatly improve by real-time magnetic resonance imaging (MRI) guidance. The integration of MRI and PT at the treatment isocenter would offer the opportunity of combining the unparalleled soft-tissue contrast and real-time imaging capabilities of MRI with the most conformal dose distribution and best dose steering capability provided by modern PT. However, hybrid systems for MR-integrated PT (MRiPT) have not been realized so far due to a number of hitherto open technological challenges. In recent years, various research groups have started addressing these challenges and exploring the technical feasibility and clinical potential of MRiPT. The aim of this contribution is to review the different aspects of MRiPT, to report on the status quo and to identify important future research topics.

Methods

Four aspects currently under study and their future directions are discussed: modelling and experimental investigations of electromagnetic interactions between the MRI and PT systems, integration of MRiPT workflows in clinical facilities, proton dose calculation algorithms in magnetic fields, and MRI-only based proton treatment planning approaches.

Conclusions

Although MRiPT is still in its infancy, significant progress on all four aspects has been made, showing promising results that justify further efforts for research and development to be undertaken. First non-clinical research solutions have recently been realized and are being thoroughly characterized. The prospect that first prototype MRiPT systems for clinical use will likely exist within the next 5 to 10 years seems realistic, but requires significant work to be performed by collaborative efforts of research groups and industrial partners.

Generation of Pseudo-CT using High-Degree Polynomial Regression on Dual-Contrast Pelvic MRI Data

Increasing interests in using magnetic resonance imaging only in radiation therapy require methods for predicting the computed tomography numbers from MRI data. Here we propose a simple voxel method to generate the pseudo-CT (pCT) image using dual-contrast pelvic MRI data. The method is first trained with the CT data and dual-contrast MRI data (two sets of MRI with different sequences) of multiple patients, where the anatomical structures in the images after deformable image registration are segmented into several regions, and after MRI intensity normalizations a regression analysis is used to determine a two-variable polynomial function for each region to relate a voxel's two MRI intensity values to its CT number. We first evaluate the accuracy via the Hounsfield unit (HU) difference between the pseudo-CT and reference-CT (rCT) images and obtain the average mean absolute error as 40.3 ± 2.9 HU from leave-one-out-cross-validation (LOOCV) across all six patients, which is better than most previous results and comparable to another study using the more complicated atlas-based method. We also perform a dosimetric evaluation of the treatment plans based on pCT and rCT images and find the average passing rate within 2% dose difference to be 95.4% in point-to-point dose comparisons. Therefore, our method shows encouraging results in predicting the CT numbers. This polynomial method needs less computer storage than the interpolation method and can be readily extended to the case of more than two MRI sequences.

Deep learning-based MR-to-CT synthesis: The influence of varying gradient echo-based MR images as input channels

PURPOSE

To study the influence of gradient echo-based contrasts as input channels to a 3D patch-based neural network trained for synthetic CT (sCT) generation in canine and human populations.

METHODS

Magnetic resonance images and CT scans of human and canine pelvic regions were acquired and paired using nonrigid registration. Magnitude MR images and Dixon reconstructed water, fat, in-phase and opposed-phase images were obtained from a single T1 -weighted multi-echo gradient-echo acquisition. From this set, 6 input configurations were defined, each containing 1 to 4 MR images regarded as input channels. For each configuration, a UNet-derived deep learning model was trained for synthetic CT generation. Reconstructed Hounsfield unit maps were evaluated with peak SNR, mean absolute error, and mean error. Dice similarity coefficient and surface distance maps assessed the geometric fidelity of bones. Repeatability was estimated by replicating the training up to 10 times.

RESULTS

Seventeen canines and 23 human subjects were included in the study. Performance and repeatability of single-channel models were dependent on the TE-related water-fat interference with variations of up to 17% in mean absolute error, and variations of up to 28% specifically in bones. Repeatability, Dice similarity coefficient, and mean absolute error were statistically significantly better in multichannel models with mean absolute error ranging from 33 to 40 Hounsfield units in humans and from 35 to 47 Hounsfield units in canines.

CONCLUSION

Significant differences in performance and robustness of deep learning models for synthetic CT generation were observed depending on the input. In-phase images outperformed opposed-phase images, and Dixon reconstructed multichannel inputs outperformed single-channel inputs.

Dosimetric evaluation of synthetic CT generated with GANs for MRI-only proton therapy treatment planning of brain tumors

PURPOSE

The purpose of this study was to address the dosimetric accuracy of synthetic computed tomography (sCT) images of patients with brain tumor generated using a modified generative adversarial network (GAN) method, for their use in magnetic resonance imaging (MRI)-only treatment planning for proton therapy.

METHODS

Dose volume histogram (DVH) analysis was performed on CT and sCT images of patients with brain tumor for plans generated for intensity-modulated proton therapy (IMPT). All plans were robustly optimized using a commercially available treatment planning system (RayStation, from RaySearch Laboratories) and standard robust parameters reported in the literature. The IMPT plan was then used to compute the dose on CT and sCT images for dosimetric comparison, using RayStation analytical (pencil beam) dose algorithm. We used a second, independent Monte Carlo dose calculation engine to recompute the dose on both CT and sCT images to ensure a proper analysis of the dosimetric accuracy of the sCT images.

RESULTS

The results extracted from RayStation showed excellent agreement for most DVH metrics computed on the CT and sCT for the nominal case, with a mean absolute difference below 0.5% (0.3 Gy) of the prescription dose for the clinical target volume (CTV) and below 2% (1.2 Gy) for the organs at risk (OARs) considered. This demonstrates a high dosimetric accuracy for the generated sCT images, especially in the target volume. The metrics obtained from the Monte Carlo doses mostly agreed with the values extracted from RayStation for the nominal and worst-case scenarios (mean difference below 3%).

CONCLUSIONS

This work demonstrated the feasibility of using sCT generated with a GAN-based deep learning method for MRI-only treatment planning of patients with brain tumor in intensity-modulated proton therapy.

Multi-sequence MR image-based synthetic CT generation using a generative adversarial network for head and neck MRI-only radiotherapy

PURPOSE

The purpose of this study is to investigate the effect of different magnetic resonance (MR) sequences on the accuracy of deep learning-based synthetic computed tomography (sCT) generation in the complex head and neck region.

METHODS

Four sequences of MR images (T1, T2, T1C, and T1DixonC-water) were collected from 45 patients with nasopharyngeal carcinoma. Seven conditional generative adversarial network (cGAN) models were trained with different sequences (single channel) and different combinations (multi-channel) as inputs. To further verify the cGAN performance, we also used a U-net network as a comparison. Mean absolute error, structural similarity index, peak signal-to-noise ratio, dice similarity coefficient, and dose distribution were evaluated between the actual CTs and sCTs generated from different models.

RESULTS

The results show that the cGAN model with multi-channel (i.e., T1 + T2 + T1C + T1DixonC-water) as input to predict sCT achieves higher accuracy than any single MR sequence model. The T1-weighted MR model achieves better results than T2, T1C, and T1DixonC-water models. The comparison between cGAN and U-net shows that the sCTs predicted by cGAN retains additional image details are less blurred and more similar to the actual CT.

CONCLUSIONS

Conditional generative adversarial network with multiple MR sequences as model input shows the best accuracy. The T1-weighted MR images provide sufficient image information and are suitable for sCT prediction in clinical scenarios with limited acquisition sequences or limited acquisition time.

Pseudo-CT generation from multi-parametric MRI using a novel multi-channel multi-path conditional generative adversarial network for nasopharyngeal carcinoma patients

PURPOSE

To develop and evaluate a novel method for pseudo-CT generation from multi-parametric MR images using multi-channel multi-path generative adversarial network (MCMP-GAN).

METHODS

Pre- and post-contrast T1-weighted (T1-w), T2-weighted (T2-w) MRI, and treatment planning CT images of 32 nasopharyngeal carcinoma (NPC) patients were employed to train a pixel-to-pixel MCMP-GAN. The network was developed based on a 5-level Residual U-Net (ResU-Net) with the channel-based independent feature extraction network to generate pseudo-CT images from multi-parametric MR images. The discriminator with five convolutional layers was added to distinguish between the real CT and pseudo-CT images, improving the nonlinearity and prediction accuracy of the model. Eightfold cross validation was implemented to validate the proposed MCMP-GAN. The pseudo-CT images were evaluated against the corresponding planning CT images based on mean absolute error (MAE), peak signal-to-noise ratio (PSNR), Dice similarity coefficient (DSC), and Structural similarity index (SSIM). Similar comparisons were also performed against the multi-channel single-path GAN (MCSP-GAN), the single-channel single-path GAN (SCSP-GAN).

RESULTS

It took approximately 20 h to train the MCMP-GAN model on a Quadro P6000, and less than 10 s to generate all pseudo-CT images for the subjects in the test set. The average head MAE between pseudo-CT and planning CT was 75.7 ± 14.6 Hounsfield Units (HU) for MCMP-GAN, significantly (P-values < 0.05) lower than that for MCSP-GAN (79.2 ± 13.0 HU) and SCSP-GAN (85.8 ± 14.3 HU). For bone only, the MCMP-GAN yielded a smaller mean MAE (194.6 ± 38.9 HU) than MCSP-GAN (203.7 ± 33.1 HU), SCSP-GAN (227.0 ± 36.7 HU). The average PSNR of MCMP-GAN (29.1 ± 1.6) was found to be higher than that of MCSP-GAN (28.8 ± 1.2) and SCSP-GAN (28.2 ± 1.3). In terms of metrics for image similarity, MCMP-GAN achieved the highest SSIM (0.92 ± 0.02) but did not show significantly improved bone DSC results in comparison with MCSP-GAN.

CONCLUSIONS

We developed a novel multi-channel GAN approach for generating pseudo-CT from multi-parametric MR images. Our preliminary results in NPC patients showed that the MCMP-GAN method performed apparently superior to the U-Net-GAN and SCSP-GAN, and slightly better than MCSP-GAN.

Towards a generalised development of synthetic CT images and assessment of their dosimetric accuracy

Background: The interest in generating "synthetic computed tomography (CT) images" from magnetic resonance (MR) images has been increasing over the past years due to advances in MR guidance for radiotherapy. A variety of methods for synthetic CT creation have been developed, from simple bulk density assignment to complex machine learning algorithms.Material and methods: In this study, we present a general method to determine simplistic synthetic CTs and evaluate them according to their dosimetric accuracy. It separates the requirements on the MR image and the associated calculation effort to generate a synthetic CT. To evaluate the significance of the dosimetric accuracy under realistic conditions, clinically common uncertainties including position shifts and Hounsfield lookup table (HLUT) errors were simulated. To illustrate our approach, we first translated CT images from a test set of six pelvic cancer patients to relative electron density (ED) via a clinical HLUT. For each patient, seven simplified ED images (simED) were generated at different levels of complexity, ranging from one to four tissue classes. Then, dose distributions optimised on the reference ED image and the simEDs were compared to each other in terms of gamma pass rates (2 mm/2% criteria) and dose volume metrics.Results: For our test set, best results were obtained for simEDs with four tissue classes representing fat, soft tissue, air, and bone. For this simED, gamma pass rates of 99.95% (range: 99.72-100%) were achieved. The decrease in accuracy from ED simplification was smaller in this case than the influence of the uncertainty scenarios on the reference image, both for gamma pass rates and dose volume metrics.Conclusions: The presented workflow helps to determine the required complexity of synthetic CTs with respect to their dosimetric accuracy. The investigated cases showed potential simplifications, based on which the synthetic CT generation could be faster and more reproducible.

Feasibility of synthetic computed tomography generated with an adversarial network for multi-sequence magnetic resonance-based brain radiotherapy

The aim of this work is to generate synthetic computed tomography (sCT) images from multi-sequence magnetic resonance (MR) images using an adversarial network and to assess the feasibility of sCT-based treatment planning for brain radiotherapy. Datasets for 15 patients with glioblastoma were selected and 580 pairs of CT and MR images were used. T1-weighted, T2-weighted and fluid-attenuated inversion recovery MR sequences were combined to create a three-channel image as input data. A conditional generative adversarial network (cGAN) was trained using image patches. The image quality was evaluated using voxel-wise mean absolute errors (MAEs) of the CT number. For the dosimetric evaluation, 3D conformal radiotherapy (3D-CRT) and volumetric modulated arc therapy (VMAT) plans were generated using the original CT set and recalculated using the sCT images. The isocenter dose and dose-volume parameters were compared for 3D-CRT and VMAT plans, respectively. The equivalent path length was also compared. The mean MAEs for the whole body, soft tissue and bone region were 108.1 ± 24.0, 38.9 ± 10.7 and 366.2 ± 62.0 hounsfield unit, respectively. The dosimetric evaluation revealed no significant difference in the isocenter dose for 3D-CRT plans. The differences in the dose received by 2% of the volume (D2%), D50% and D98% relative to the prescribed dose were <1.0%. The overall equivalent path length was shorter than that for real CT by 0.6 ± 1.9 mm. A treatment planning study using generated sCT detected only small, clinically negligible differences. These findings demonstrated the feasibility of generating sCT images for MR-only radiotherapy from multi-sequence MR images using cGAN.

Automated air region delineation on MRI for synthetic CT creation

Automatically and accurately separating air from other low signal regions (especially bone, liver, etc) in an MRI is difficult because these tissues produce similar MR intensities, resulting in errors in synthetic CT generation for MRI-based radiation therapy planning. This work aims to develop a technique to accurately and automatically determine air-regions for MR-guided adaptive radiation therapy. CT and MRI scans (T2-weighted) of phantoms with fabricated air-cavities and abdominal cancer patients were used to establish an MR intensity threshold for air delineation. From the phantom data, air/tissue boundaries in MRI were identified by CT-MRI registration. A formula relating the MRI intensities of air and surrounding materials was established to auto-threshold air-regions. The air-regions were further refined by using quantitative image texture features. A naive Bayesian classifier was trained using the extracted features with a leave-one-out cross validation technique to differentiate air from non-air voxels. The multi-step air auto-segmentation method was tested against the manually segmented air-regions. The dosimetry impacts of the air-segmentation methods were studied. Air-regions in the abdomen can be segmented on MRI within 1 mm accuracy using a multi-step auto-segmentation method as compared to manually delineated contours. The air delineation based on the MR threshold formula was improved using the MRI texture differences between air and tissues, as judged by the area under the receiver operating characteristic curve of 81% when two texture features (autocorrelation and contrast) were used. The performance increased to 82% with using three features (autocorrelation, sum-variance, and contrast). Dosimetric analysis showed no significant difference between the auto-segmentation and manual MR air delineation on commonly used dose volume parameters. The proposed techniques consisting of intensity-based auto-thresholding and image texture-based voxel classification can automatically and accurately segment air-regions on MRI, allowing synthetic CT to be generated quickly and precisely for MR-guided adaptive radiation therapy.

MRI-guided proton therapy planning: accounting for an inline MRI fringe field

MRI-guided proton therapy is being pursued for its promise to provide a more conformal, accurate proton therapy. However, the presence of the magnetic field imposes a challenge for the beam delivery as protons are deflected due to the Lorenz force. In this study, the impact of realistic inline MRI fringe field on IMPT plan delivery is investigated for a water phantom, liver tumor and prostate cancer differing in target volume, shape, and field configuration using Monte Carlo simulations. A method to correct for the shift of the beam spot positions in the presence of the inline magnetic field is presented. Results show that when not accounting for the effect of the magnetic field on the pencil beam delivery, the spot positions are substantially shifted and the quality of delivered plans is significantly deteriorated leading to dose inhomogeneities and creation of hot and cold spots. However, by correcting the pencil beam delivery, the dose quality of the IMPT plans is restored to a high degree. Nevertheless, adaptation of beam delivery alone is not robust regarding different treatment sites. By fully accounting during plan optimization for the dose distortions caused by the fringe and imaging fields, highly conformal IMPT plans are achieved. These results demonstrate proton pencil beam scanning and treatment planning can be adapted for precise delivery of state-of-the-art IMPT plans in MR-guided proton therapy in the presence of an inline MRI fringe field.

MRI-Based Proton Treatment Planning for Base of Skull Tumors

Purpose

To introduce a novel, deep-learning method to generate synthetic computed tomography (SCT) scans for proton treatment planning and evaluate its efficacy.

Materials and Methods

50 Patients with base of skull tumors were divided into 2 nonoverlapping training and study cohorts. Computed tomography and magnetic resonance imaging pairs for patients in the training cohort were used for training our novel 3-dimensional generative adversarial network (cycleGAN) algorithm. Upon completion of the training phase, SCT scans for patients in the study cohort were predicted based on their magnetic resonance images only. The SCT scans obtained were compared against the corresponding original planning computed tomography scans as the ground truth, and mean absolute errors (in Hounsfield units [HU]) and normalized cross-correlations were calculated. Proton plans of 45 Gy in 25 fractions with 2 beams per plan were generated for the patients based on their planning computed tomographies and recalculated on SCT scans. Dose-volume histogram endpoints were compared. A γ-index analysis along 3 cardinal planes intercepting at the isocenter was performed. Proton distal range along each beam was calculated.

Results

Image quality metrics show agreement between the generated SCT scans and the ground truth with mean absolute error values ranging from 38.65 to 65.12 HU and an average of 54.55 ± 6.81 HU and a normalized cross-correlation average of 0.96 ± 0.01. The dosimetric evaluation showed no statistically significant differences (p > 0.05) within planning target volumes for dose-volume histogram endpoints and other metrics studied, with the exception of the dose covering 95% of the target volume, with a relative difference of 0.47%. The γ-index analysis showed an average passing rate of 98% with a 10% threshold and 2% and 2-mm criteria. Proton ranges of 48 of 50 beams (96%) in this study were within clinical tolerance adopted by 4 institutions.

Conclusions

This study shows our method is capable of generating SCT scans with acceptable image quality, dose distribution agreement, and proton distal range compared with the ground truth. Our results set a promising approach for magnetic resonance imaging-based proton treatment planning.

Tissue segmentation-based electron density mapping for MR-only radiotherapy treatment planning of brain using conventional T1-weighted MR images

PURPOSE

Magnetic resonance imaging (MRI) is the primary modality for targeting brain tumors in radiotherapy treatment planning (RTP). MRI is not directly used for dose calculation since image voxel intensities of MRI are not associated with EDs of tissues as those of computed tomography (CT). The purpose of the present study is to develop and evaluate a tissue segmentation-based method to generate a synthetic-CT (sCT) by mapping EDs to corresponding tissues using only T1-weighted MR images for MR-only RTP.

METHODS

Air regions were contoured in several slices. Then, air, bone, brain, cerebrospinal fluid (CSF), and other soft tissues were automatically segmented with an in-house algorithm based on edge detection and anatomical information and relative intensity distribution. The intensities of voxels in each segmented tissue were mapped into their CT number range to generate a sCT. Twenty-five stereotactic radiosurgery and stereotactic ablative radiotherapy patients' T1-weighted MRI and coregistered CT images from two centers were retrospectively evaluated. The CT was used as ground truth. Distances between bone contours of the external skull of sCT and CT were measured. The mean error (ME) and mean absolute error (MAE) of electron density represented by standardized CT number was calculated in HU.

RESULTS

The average distance between the contour of the external skull in sCT and the contour in coregistered CT is 1.0 ± 0.2 mm (mean ± 1SD). The ME and MAE differences for air, soft tissue and whole body voxels within external body contours are -4 HU/24 HU, 2 HU/26 HU, and -2 HU/125 HU, respectively.

CONCLUSIONS

A MR-sCT generation technique was developed based on tissue segmentation and voxel-based tissue ED mapping. The generated sCT is comparable to real CT in terms of anatomical position of tissues and similarity to the ED assignment. This method provides a feasible method to generate sCT for MR-only radiotherapy treatment planning.

MR-based treatment planning in radiation therapy using a deep learning approach

Purpose

To develop and evaluate the feasibility of deep learning approaches for MR‐based treatment planning (deepMTP) in brain tumor radiation therapy.

Methods and materials

A treatment planning pipeline was constructed using a deep learning approach to generate continuously valued pseudo CT images from MR images. A deep convolutional neural network was designed to identify tissue features in volumetric head MR images training with co‐registered kVCT images. A set of 40 retrospective 3D T1‐weighted head images was utilized to train the model, and evaluated in 10 clinical cases with brain metastases by comparing treatment plans using deep learning generated pseudo CT and using an acquired planning kVCT. Paired‐sample Wilcoxon signed rank sum tests were used for statistical analysis to compare dosimetric parameters of plans made with pseudo CT images generated from deepMTP to those made with kVCT‐based clinical treatment plan (CTTP).

Results

deepMTP provides an accurate pseudo CT with Dice coefficients for air: 0.95 ± 0.01, soft tissue: 0.94 ± 0.02, and bone: 0.85 ± 0.02 and a mean absolute error of 75 ± 23 HU compared with acquired kVCTs. The absolute percentage differences of dosimetric parameters between deepMTP and CTTP was 0.24% ± 0.46% for planning target volume (PTV) volume, 1.39% ± 1.31% for maximum dose and 0.27% ± 0.79% for the PTV receiving 95% of the prescribed dose (V95). Furthermore, no significant difference was found for PTV volume (P = 0.50), the maximum dose (P = 0.83) and V95 (P = 0.19) between deepMTP and CTTP.

Conclusions

We have developed an automated approach (deepMTP) that allows generation of a continuously valued pseudo CT from a single high‐resolution 3D MR image and evaluated it in partial brain tumor treatment planning. The deepMTP provided dose distribution with no significant difference relative to a kVCT‐based standard volumetric modulated arc therapy plans.

MRI-only based synthetic CT generation using dense cycle consistent generative adversarial networks

PURPOSE

Automated synthetic computed tomography (sCT) generation based on magnetic resonance imaging (MRI) images would allow for MRI-only based treatment planning in radiation therapy, eliminating the need for CT simulation and simplifying the patient treatment workflow. In this work, the authors propose a novel method for generation of sCT based on dense cycle-consistent generative adversarial networks (cycle GAN), a deep-learning based model that trains two transformation mappings (MRI to CT and CT to MRI) simultaneously.

METHODS AND MATERIALS

The cycle GAN-based model was developed to generate sCT images in a patch-based framework. Cycle GAN was applied to this problem because it includes an inverse transformation from CT to MRI, which helps constrain the model to learn a one-to-one mapping. Dense block-based networks were used to construct generator of cycle GAN. The network weights and variables were optimized via a gradient difference (GD) loss and a novel distance loss metric between sCT and original CT.

RESULTS

Leave-one-out cross-validation was performed to validate the proposed model. The mean absolute error (MAE), peak signal-to-noise ratio (PSNR), and normalized cross correlation (NCC) indexes were used to quantify the differences between the sCT and original planning CT images. For the proposed method, the mean MAE between sCT and CT were 55.7 Hounsfield units (HU) for 24 brain cancer patients and 50.8 HU for 20 prostate cancer patients. The mean PSNR and NCC were 26.6 dB and 0.963 in the brain cases, and 24.5 dB and 0.929 in the pelvis.

CONCLUSION

We developed and validated a novel learning-based approach to generate CT images from routine MRIs based on dense cycle GAN model to effectively capture the relationship between the CT and MRIs. The proposed method can generate robust, high-quality sCT in minutes. The proposed method offers strong potential for supporting near real-time MRI-only treatment planning in the brain and pelvis.

MRI-based synthetic CT generation using semantic random forest with iterative refinement

Target delineation for radiation therapy treatment planning often benefits from magnetic resonance imaging (MRI) in addition to x-ray computed tomography (CT) due to MRI's superior soft tissue contrast. MRI-based treatment planning could reduce systematic MR-CT co-registration errors, medical cost, radiation exposure, and simplify clinical workflow. However, MRI-only based treatment planning is not widely used to date because treatment-planning systems rely on the electron density information provided by CTs to calculate dose. Additionally, air and bone regions are difficult to separategiven their similar intensities in MR imaging. The purpose of this work is to develop a learning-based method to generate patient-specific synthetic CT (sCT) from a routine anatomical MRI for use in MRI-only radiotherapy treatment planning. An auto-context model with patch-based anatomical features was integrated into a classification random forest to generate and improve semantic information. The semantic information along with anatomical features was then used to train a series of regression random forests based on the auto-context model. After training, the sCT of a new MRI can be generated by feeding anatomical features extracted from the MRI into the well-trained classification and regression random forests. The proposed algorithm was evaluated using 14 patient datasets withT1-weighted MR and corresponding CT images of the brain. The mean absolute error (MAE), peak signal-to-noise ratio (PSNR), and normalized cross correlation (NCC) were 57.45  ±  8.45 HU, 28.33  ±  1.68 dB, and 0.97  ±  0.01. We also compared the difference between dose maps calculated on the sCT and those on the original CT, using the same plan parameters. The average DVH differences among all patients are less than 0.2 Gy for PTVs, and less than 0.02 Gy for OARs. The sCT generation by the proposed method allows for dose calculation based MR imaging alone, and may be a useful tool for MRI-based radiation treatment planning.

Feasibility of MRI-only photon and proton dose calculations for pediatric patients with abdominal tumors

The purpose of this study was to develop a method enabling synthetic computed tomography (sCT) generation of the whole abdomen using magnetic resonance imaging (MRI) scans of pediatric patients with abdominal tumors. The proposed method relies on an automatic atlas-based segmentation of bone and lungs followed by an MRI intensity to synthetic Hounsfield unit conversion. Separate conversion algorithms were used for bone, lungs and soft-tissue. Rigidly registered CT and T2-weighted MR images of 30 patients in treatment position and with the same field of view were used for the evaluation of the atlas and the conversion algorithms. The dose calculation accuracy of the generated sCTs was verified for volumetric modulated arc therapy (VMAT) and pencil beam scanning (PBS). VMAT and PBS plans were robust optimized on an internal target volume (ITV) against a patient set-up uncertainty of 5 mm. Average differences between CT and sCT dose calculations for the ITV V _95% were 0.5% (min 0.0%; max 5.0%) and 0.0% (min  -0.1%; max 0.1%) for VMAT and PBS dose distributions, respectively. Average differences for the mean dose to the organs at risk were  <0.2% (min  -0.6%; max 1.2%) and  <0.2% (min  -2.0%; max 2.6%) for VMAT and PBS dose distributions, respectively. Results show that MRI-only photon and proton dose calculations are feasible for children with abdominal tumors.

Integrating a low-field open MR scanner with a static proton research beam line: proof of concept

On-line image guidance using magnetic resonance (MR) imaging is expected to improve the targeting accuracy of proton therapy. However, to date no combined system exists. In this study, for the first time a low-field open MR scanner was integrated with a static proton research beam line to test the feasibility of simultaneous irradiation and imaging. The field-of-view of the MR scanner was aligned with the beam by taking into account the Lorentz force induced beam deflection. Various imaging sequences for extremities were performed on a healthy volunteer and on a patient with a soft-tissue sarcoma of the upper arm, both with the proton beam line switched off. T ₁-weighted spin echo images of a tissue-mimicking phantom were acquired without beam, with energised beam line magnets and during proton irradiation. Beam profiles were acquired for the MR scanner's static magnetic field alone and in combination with the dynamic gradient fields during the acquisition of different imaging sequences. It was shown that MR imaging is feasible in the electromagnetically contaminated environment of a proton therapy facility. The observed quality of the anatomical MR images was rated to be sufficient for target volume definition and positioning. The tissue-mimicking phantom showed no visible beam-induced image degradation. The beam profiles depicted no influence due to the dynamic gradient fields of the imaging sequences. This study proves that simultaneous irradiation and in-beam MR imaging is technically feasible with a low-field MR scanner integrated with a static proton research beam line.

Technical Note: U-net-generated synthetic CT images for magnetic resonance imaging-only prostate intensity-modulated radiation therapy treatment planning

PURPOSE

Clinical implementation of magnetic resonance imaging (MRI)-only radiotherapy requires a method to derive synthetic CT image (S-CT) for dose calculation. This study investigated the feasibility of building a deep convolutional neural network for MRI-based S-CT generation and evaluated the dosimetric accuracy on prostate IMRT planning.

METHODS

A paired CT and T2-weighted MR images were acquired from each of 51 prostate cancer patients. Fifteen pairs were randomly chosen as tested set and the remaining 36 pairs as training set. The training subjects were augmented by applying artificial deformations and feed to a two-dimensional U-net which contains 23 convolutional layers and 25.29 million trainable parameters. The U-net represents a nonlinear function with input an MR slice and output the corresponding S-CT slice. The mean absolute error (MAE) of Hounsfield unit (HU) between the true CT and S-CT images was used to evaluate the HU estimation accuracy. IMRT plans with dose 79.2 Gy prescribed to the PTV were applied using the true CT images. The true CT images then were replaced by the S-CT images and the dose matrices were recalculated on the same plan and compared to the one obtained from the true CT using gamma index analysis and absolute point dose discrepancy.

RESULTS

The U-net was trained from scratch in 58.67 h using a GP100-GPU. The computation time for generating a new S-CT volume image was 3.84-7.65 s. Within body, the (mean ± SD) of MAE was (29.96 ± 4.87) HU. The 1%/1 mm and 2%/2 mm gamma pass rates were over 98.03% and 99.36% respectively. The DVH parameters discrepancy was less than 0.87% and the maximum point dose discrepancy within PTV was less than 1.01% respect to the prescription.

CONCLUSION

The U-net can generate S-CT images from conventional MR image within seconds with high dosimetric accuracy for prostate IMRT plan.

Proton range shift analysis on brain pseudo-CT generated from T1 and T2 MR

BACKGROUND

In radiotherapy, MR imaging is only used because it has significantly better soft tissue contrast than CT, but it lacks electron density information needed for dose calculation. This work assesses the feasibility of using pseudo-CT (pCT) generated from T1w/T2w MR for proton treatment planning, where proton range comparisons are performed between standard CT and pCT.

MATERIAL AND METHODS

MR and CT data from 14 glioblastoma patients were used in this study. The pCT was generated by using conversion libraries obtained from tissue segmentation and anatomical regioning of the T1w/T2w MR. For each patient, a plan consisting of three 18 Gy beams was designed on the pCT, for a total of 42 analyzed beams. The plan was then transferred onto the CT that represented the ground truth. Range shift (RS) between pCT and CT was computed at R₈₀ over 10 slices. The acceptance threshold for RS was according to clinical guidelines of two institutions. A γ-index test was also performed on the total dose for each patient.

RESULTS

Mean absolute error and bias for the pCT were 124 ± 10 and -16 ± 26 Hounsfield Units (HU), respectively. The median and interquartile range of RS was 0.5 and 1.4 mm, with highest absolute value being 4.4 mm. Of the 42 beams, 40 showed RS less than the clinical range margin. The two beams with larger RS were both in the cranio-caudal direction and had segmentation errors due to the partial volume effect, leading to misassignment of the HU.

CONCLUSIONS

This study showed the feasibility of using T1w and T2w MRI to generate a pCT for proton therapy treatment, thus avoiding the use of a planning CT and allowing better target definition and possibilities for online adaptive therapies. Further improvements of the methodology are still required to improve the conversion from MRI intensities to HUs.

Comparative study of algorithms for synthetic CT generation from MRI: Consequences for MRI-guided radiation planning in the pelvic region

PURPOSE

Magnetic resonance imaging (MRI)-guided radiation therapy (RT) treatment planning is limited by the fact that the electron density distribution required for dose calculation is not readily provided by MR imaging. We compare a selection of novel synthetic CT generation algorithms recently reported in the literature, including segmentation-based, atlas-based and machine learning techniques, using the same cohort of patients and quantitative evaluation metrics.

METHODS

Six MRI-guided synthetic CT generation algorithms were evaluated: one segmentation technique into a single tissue class (water-only), four atlas-based techniques, namely, median value of atlas images (ALMedian), atlas-based local weighted voting (ALWV), bone enhanced atlas-based local weighted voting (ALWV-Bone), iterative atlas-based local weighted voting (ALWV-Iter), and a machine learning technique using deep convolution neural network (DCNN).

RESULTS

Organ auto-contouring from MR images was evaluated for bladder, rectum, bones, and body boundary. Overall, DCNN exhibited higher segmentation accuracy resulting in Dice indices (DSC) of 0.93 ± 0.17, 0.90 ± 0.04, and 0.93 ± 0.02 for bladder, rectum, and bones, respectively. On the other hand, ALMedian showed the lowest accuracy with DSC of 0.82 ± 0.20, 0.81 ± 0.08, and 0.88 ± 0.04, respectively. DCNN reached the best performance in terms of accurate derivation of synthetic CT values within each organ, with a mean absolute error within the body contour of 32.7 ± 7.9 HU, followed by the advanced atlas-based methods (ALWV: 40.5 ± 8.2 HU, ALWV-Iter: 42.4 ± 8.1 HU, ALWV-Bone: 44.0 ± 8.9 HU). ALMedian led to the highest error (52.1 ± 11.1 HU). Considering the dosimetric evaluation results, ALWV-Iter, ALWV, DCNN and ALWV-Bone led to similar mean dose estimation within each organ at risk and target volume with less than 1% dose discrepancy. However, the two-dimensional gamma analysis demonstrated higher pass rates for ALWV-Bone, DCNN, ALMedian and ALWV-Iter at 1%/1 mm criterion with 94.99 ± 5.15%, 94.59 ± 5.65%, 93.68 ± 5.53% and 93.10 ± 5.99% success, respectively, while ALWV and water-only resulted in 86.91 ± 13.50% and 80.77 ± 12.10%, respectively.

CONCLUSIONS

Overall, machine learning and advanced atlas-based methods exhibited promising performance by achieving reliable organ segmentation and synthetic CT generation. DCNN appears to have slightly better performance by achieving accurate automated organ segmentation and relatively small dosimetric errors (followed closely by advanced atlas-based methods, which in some cases achieved similar performance). However, the DCNN approach showed higher vulnerability to anatomical variation, where a greater number of outliers was observed with this method. Considering the dosimetric results obtained from the evaluated methods, the challenge of electron density estimation from MR images can be resolved with a clinically tolerable error.

MRI-based pseudo CT synthesis using anatomical signature and alternating random forest with iterative refinement model

We develop a learning-based method to generate patient-specific pseudo computed tomography (CT) from routinely acquired magnetic resonance imaging (MRI) for potential MRI-based radiotherapy treatment planning. The proposed pseudo CT (PCT) synthesis method consists of a training stage and a synthesizing stage. During the training stage, patch-based features are extracted from MRIs. Using a feature selection, the most informative features are identified as an anatomical signature to train a sequence of alternating random forests based on an iterative refinement model. During the synthesizing stage, we feed the anatomical signatures extracted from an MRI into the sequence of well-trained forests for a PCT synthesis. Our PCT was compared with original CT (ground truth) to quantitatively assess the synthesis accuracy. The mean absolute error, peak signal-to-noise ratio, and normalized cross-correlation indices were 60.87 ± 15.10    HU , 24.63 ± 1.73    dB , and 0.954 ± 0.013 for 14 patients' brain data and 29.86 ± 10.4    HU , 34.18 ± 3.31    dB , and 0.980 ± 0.025 for 12 patients' pelvic data, respectively. We have investigated a learning-based approach to synthesize CTs from routine MRIs and demonstrated its feasibility and reliability. The proposed PCT synthesis technique can be a useful tool for MRI-based radiation treatment planning.

Emerging role of MRI in radiation therapy

Advances in multimodality imaging, providing accurate information of the irradiated target volume and the adjacent critical structures or organs at risk (OAR), has made significant improvements in delivery of the external beam radiation dose. Radiation therapy conventionally has used computed tomography (CT) imaging for treatment planning and dose delivery. However, magnetic resonance imaging (MRI) provides unique advantages: added contrast information that can improve segmentation of the areas of interest, motion information that can help to better target and deliver radiation therapy, and posttreatment outcome analysis to better understand the biologic effect of radiation. To take advantage of these and other potential advantages of MRI in radiation therapy, radiologists and MRI physicists will need to understand the current radiation therapy workflow and speak the same language as our radiation therapy colleagues. This review article highlights the emerging role of MRI in radiation dose planning and delivery, but more so for MR-only treatment planning and delivery. Some of the areas of interest and challenges in implementing MRI in radiation therapy workflow are also briefly discussed. Level of Evidence: 5 Technical Efficacy: Stage 5 J. Magn. Reson. Imaging 2018;48:1468-1478.

Cone beam CT for QA of synthetic CT in MRI only for prostate patients

Purpose

Magnetic resonance imaging (MRI)‐only radiotherapy is performed without computed tomography (CT). A synthetic CT (sCT) is used for treatment planning. The aim of this study was to develop a clinically feasible quality assurance (QA) procedure for sCT using the kV‐cone beam CT (CBCT), in an MRI‐only workflow for prostate cancer patients.

Material and method

Three criteria were addressed; stability in Hounsfield Units (HUs), deviations in HUs between the CT and CBCT, and validation of the QA procedure. For the two first criteria, weekly phantom measurements were performed. For the third criteria, sCT, CT, and CBCT for ten patients were used. Treatment plans were created based on the sCT (MriPlanner^TM). CT and CBCT images were registered to the sCT. The treatment plan was copied to the CT and CBCT and recalculated. Dose–volume histogram (DVH) metrics were used to evaluate dosimetric differences between the sCT plan and the recalculated CT and CBCT plans. HU distributions in sCT, CT, and CBCT were compared. Well‐defined errors were introduced in the sCT for one patient to evaluate efficacy of the QA procedure.

Results

The kV‐CBCT system was stable in HU over time (standard deviation <40 HU). Variation in HUs between CT and CBCT was <60 HU. The differences between sCT–CT and sCT–CBCT dose distributions were below or equal to 1.0%. The highest mean dose difference for the CT and CBCT dose distribution was 0.6%. No statistically significant difference was found between total mean dose deviations from recalculated CT and CBCT plans, except for femoral head. Comparing HU distributions, the CBCT appeared to be similar to the CT. All introduced errors were identified by the proposed QA procedure, except all tissue compartments assigned as water.

Conclusion

The results in this study shows that CBCT can be used as a clinically feasible QA procedure for MRI‐only radiotherapy of prostate cancer patients.

Magnetic resonance imaging-based pseudo computed tomography using anatomic signature and joint dictionary learning

Magnetic resonance imaging (MRI) provides a number of advantages over computed tomography (CT) for radiation therapy treatment planning; however, MRI lacks the key electron density information necessary for accurate dose calculation. We propose a dictionary-learning-based method to derive electron density information from MRIs. Specifically, we first partition a given MR image into a set of patches, for which we used a joint dictionary learning method to directly predict a CT patch as a structured output. Then a feature selection method is used to ensure prediction robustness. Finally, we combine all the predicted CT patches to obtain the final prediction for the given MR image. This prediction technique was validated for a clinical application using 14 patients with brain MR and CT images. The peak signal-to-noise ratio (PSNR), mean absolute error (MAE), normalized cross-correlation (NCC) indices and similarity index (SI) for air, soft-tissue and bone region were used to quantify the prediction accuracy. The mean ± std of PSNR, MAE, and NCC were: 22.4±1.9  dB , 82.6±26.1 HU, and 0.91±0.03 for the 14 patients. The SIs for air, soft-tissue, and bone regions are 0.98±0.01 , 0.88±0.03 , and 0.69±0.08 . These indices demonstrate the CT prediction accuracy of the proposed learning-based method. This CT image prediction technique could be used as a tool for MRI-based radiation treatment planning, or for PET attenuation correction in a PET/MRI scanner.

Predicting CT Image From MRI Data Through Feature Matching With Learned Nonlinear Local Descriptors

Attenuation correction for positron-emission tomography (PET)/magnetic resonance (MR) hybrid imaging systems and dose planning for MR-based radiation therapy remain challenging due to insufficient high-energy photon attenuation information. We present a novel approach that uses the learned nonlinear local descriptors and feature matching to predict pseudo computed tomography (pCT) images from T1-weighted and T2-weighted magnetic resonance imaging (MRI) data. The nonlinear local descriptors are obtained by projecting the linear descriptors into the nonlinear high-dimensional space using an explicit feature map and low-rank approximation with supervised manifold regularization. The nearest neighbors of each local descriptor in the input MR images are searched in a constrained spatial range of the MR images among the training dataset. Then the pCT patches are estimated through k-nearest neighbor regression. The proposed method for pCT prediction is quantitatively analyzed on a dataset consisting of paired brain MRI and CT images from 13 subjects. Our method generates pCT images with a mean absolute error (MAE) of 75.25 ± 18.05 Hounsfield units, a peak signal-to-noise ratio of 30.87 ± 1.15 dB, a relative MAE of 1.56 ± 0.5% in PET attenuation correction, and a dose relative structure volume difference of 0.055 ± 0.107% in , as compared with true CT. The experimental results also show that our method outperforms four state-of-the-art methods.

MRI-only treatment planning: benefits and challenges

Over the past decade, the application of magnetic resonance imaging (MRI) has increased, and there is growing evidence to suggest that improvements in the accuracy of target delineation in MRI-guided radiation therapy may improve clinical outcomes in a variety of cancer types. However, some considerations should be recognized including patient motion during image acquisition and geometric accuracy of images. Moreover, MR-compatible immobilization devices need to be used when acquiring images in the treatment position while minimizing patient motion during the scan time. Finally, synthetic CT images (i.e. electron density maps) and digitally reconstructed radiograph images should be generated from MRI images for dose calculation and image guidance prior to treatment. A short review of the concepts and techniques that have been developed for implementation of MRI-only workflows in radiation therapy is provided in this document.