Publications

@article{Whelan2024,

title = {Open‐source hardware and software for the measurement, characterization, reporting, and correction of geometric distortion in MRI},

author = {Brendan M. Whelan and Paul Z. Y. Liu and Shanshan Shan and David E. J. Waddington and Bin Dong and Michael G. Jameson and Paul J. Keall},

doi = {10.1002/mp.17342},

issn = {2473-4209},

year  = {2024},

date = {2024-11-00},

journal = {Medical Physics},

volume = {51},

number = {11},

pages = {8399--8410},

publisher = {Wiley},

abstract = {AbstractBackgroundGeometric distortion is a serious problem in MRI, particularly in MRI guided therapy. A lack of affordable and adaptable tools in this area limits research progress and harmonized quality assurance.PurposeTo develop and test a suite of open‐source hardware and software tools for the measurement, characterization, reporting, and correction of geometric distortion in MRI.MethodsAn open‐source python library was developed, comprising modules for parametric phantom design, data processing, spherical harmonics, distortion correction, and interactive reporting. The code was used to design and manufacture a distortion phantom consisting of 618 oil filled markers covering a sphere of radius 150 mm. This phantom was imaged on a CT scanner and a novel split‐bore 1.0 T MRI magnet. The CT images provide distortion‐free dataset. These data were used to test all modules of the open‐source software.ResultsAll markers were successfully extracted from all images. The distorted MRI markers were mapped to undistorted CT data using an iterative search approach. Spherical harmonics reconstructed the fitted gradient data to 1.0 ± 0.6% of the input data. High resolution data were reconstructed via spherical harmonics and used to generate an interactive report. Finally, distortion correction on an independent data set reduced distortion inside the DSV from 5.5 ± 3.1 to 1.6 ± 0.8 mm.ConclusionOpen‐source hardware and software for the measurement, characterization, reporting, and correction of geometric distortion in MRI have been developed. The utility of these tools has been demonstrated via their application on a novel 1.0 T split bore magnet.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Choi2024,

title = {Dynamic contrast enhanced high field magnetic resonance imaging for canine primary intracranial neoplasia},

author = {Simon Choi and Caterina Brighi and Sam Long},

doi = {10.3389/fvets.2024.1468831},

issn = {2297-1769},

year  = {2024},

date = {2024-10-04},

journal = {Front. Vet. Sci.},

volume = {11},

publisher = {Frontiers Media SA},

abstract = {IntroductionDistinguishing meningiomas from other intracranial neoplasms is clinically relevant as the prognostic and therapeutic implications differ greatly and influence clinical decision making. Dynamic contrast-enhanced MRI (DCE-MRI) is an imaging technique that assists with characterisation of physiologic alterations such as blood flow and tissue vascular permeability. Quantitative pharmacokinetic analysis utilising DCE-MRI has not been studied in canine neuro-oncology.MethodsA retrospective study was performed in canine patients that underwent DCE-MRI with an imaging diagnosis of an intracranial meningioma and surgery for histopathological diagnosis. Kinetic parameters Ktrans and cerebral blood flow were measured and compared to assess whether differences could be identified between meningiomas and other intracranial neoplasms.ResultsSix dogs with meningiomas and 3 dogs with other intracranial neoplasms were included for statistical analysis. Cerebral blood flow values were found to be statistically higher within meningiomas compared to other intracranial neoplasms. Ktrans values were higher within meningiomas than in other types of intracranial tumours, however this difference did not reach statistical significance.DiscussionBased on the results of this study cerebral blood flow measurement can be utilised to differentiate canine intracranial meningiomas from other similar appearing intracranial tumours.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Chin2024,

title = {Cardiac Substructure Dose and Survival in Stereotactic Radiotherapy for Lung Cancer: Results of the Multi-Centre SSBROC Trial},

author = {V. Chin and P. Chlap and R. Finnegan and E. Hau and A. Ong and X. Ma and J. Descallar and J. Otton and L. Holloway and G.P. Delaney and S.K. Vinod},

doi = {10.1016/j.clon.2024.07.005},

issn = {0936-6555},

year  = {2024},

date = {2024-10-00},

journal = {Clinical Oncology},

volume = {36},

number = {10},

pages = {642--650},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Gardner2024,

title = {Data-driven rapid 4D cone-beam CT reconstruction for new generation linacs},

author = {Mark Gardner and Owen Dillon and Hilary Byrne and Paul Keall and Ricky O’Brien},

doi = {10.1088/1361-6560/ad780a},

issn = {1361-6560},

year  = {2024},

date = {2024-09-21},

journal = {Phys. Med. Biol.},

volume = {69},

number = {18},

publisher = {IOP Publishing},

abstract = {Abstract

               

                  Objective. Newer generation linear accelerators (Linacs) allow 20 s cone-beam CT (CBCT) acquisition which reduces radiation therapy treatment time. However, the current clinical application of these rapid scans is only 3DCBCT. In this paper we propose a novel data-driven rapid 4DCBCT reconstruction method for new generation linacs. Approach. This method relies on estimating the magnitude of the diaphragm motion from an initial 3D reconstruction. This estimated motion is used to linearly approximate a deformation vector field (DVF) for each respiration phase. These DVFs are then used for motion compensated Feldkamp–Davis–Kress (MCFDK) reconstructions. This method, named MCFDK Data Driven (MCFDK-DD), was compared to a MCFDK reconstruction using a prior motion model (MCFDK-Prior), a 3D-FDK reconstruction, and a conventional acquisition (4 mins) conventional reconstruction 4DCBCT (4D-FDK). The data used in this paper were derived from 4DCT volumes from 12 patients from The Cancer Imaging Archives. Image quality was quantified using RMSE of line plots centred on the tumour, tissue interface width (TIW), the mean square error (MSE) and structural similarity index measurement (SSIM). Main Results. The tumour line plots in the Superior-Inferior direction showed reduced RMSE for the MCFDK-DD compared to the 3D-FDK method, indicating the MCFDK-DD method provided a more accurate tumour location. Similarly, the TIW values from the MCFDK-DD reconstructions (median 8.6 mm) were significantly reduced for the MCFDK-DD method compared to the 3D-FDK reconstructions (median 14.8 mm, (p < 0.001). The MCFDK-DD, MCFDK-Prior and 3D-FDK had median MSE values of 

                     

                     

                        

                           1.08

                           ×

                           

                              

                                 10

                                 

                                    −

                                    6

                                 

                              

                           

                           

                           

                              m

                           

                           

                              

                                 

                                    m

                                 

                                 

                                    −

                                    1

                                 

                              

                           

                        

                     

                  , 

                     

                     

                        

                           1.11

                           ×

                           

                              

                                 10

                                 

                                    −

                                    6

                                 

                              

                           

                           

                           

                              m

                           

                           

                              

                                 

                                    m

                                 

                                 

                                    −

                                    1

                                 

                              

                           

                        

                     

                   and 

                     

                     

                        

                           1.17

                           ×

                           

                              

                                 10

                                 

                                    −

                                    6

                                 

                              

                           

                           

                           

                              m

                           

                           

                              

                                 

                                    m

                                 

                                 

                                    −

                                    1

                                 

                              

                           

                        

                     

                   respectively. The corresponding median SSIM values were 0.93, 0.92 and 0.92 respectively indicating the MCFDK-DD had good agreement with the conventional 4D-FDK reconstructions. Significance. These results demonstrate the feasibility of creating accurate data-driven 4DCBCT images for rapid scans on new generation linacs. These findings could lead to increased clinical usage of 4D information on newer generation linacs.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Byrne2024,

title = {Prospective Randomized Trial Comparing 2 Devices for Deep Inspiration Breath Hold Management in Breast Radiation Therapy: Results of the BRAVEHeart Trial},

author = {Hilary L. Byrne and Elisabeth Steiner and Jeremy Booth and Gillian Lamoury and Marita Morgia and Susan Carroll and Kylie Richardson and Leigh Ambrose and Kuldeep Makhija and Cameron Stanton and Benjamin Zwan and Michael Carr and Maegan Stewart and Regina Bromley and John Atyeo and Shona Silvester and Natalie Plant and Paul Keall},

doi = {10.1016/j.adro.2024.101572},

issn = {2452-1094},

year  = {2024},

date = {2024-09-00},

journal = {Advances in Radiation Oncology},

volume = {9},

number = {9},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Huang2024,

title = {Radiotherapy protocol compliance in routine clinical practice for patients with stages I–III non‐small‐cell lung cancer},

author = {Xiaoshui Huang and Matthew Field and Shalini Vinod and Helen Ball and Vikneswary Batumalai and Paul Keall and Lois Holloway},

doi = {10.1111/1754-9485.13727},

issn = {1754-9485},

year  = {2024},

date = {2024-09-00},

journal = {J Med Imag Rad Onc},

volume = {68},

number = {6},

pages = {729--739},

publisher = {Wiley},

abstract = {AbstractIntroductionDespite the availability of radiotherapy treatment protocols for lung cancer, considerable treatment variation occurs in clinical practice. This study assessed compliance with a radiotherapy protocol for the treatment of patients with stages I–III non‐small‐cell lung cancer (NSCLC) in routine clinical practice and to identify factors that were associated with compliance.MethodsThe Cancer Institute New South Wales eviQ treatment protocol for external beam radiotherapy of stages I–III NSCLC was taken as the reference to measure compliance. All inoperable patients with stages I–III NSCLC and documented ECOG performance status treated with radiotherapy between 2007 and 2019 at two radiotherapy facilities were available for analysis. Protocol compliance rates were calculated. Univariate and multivariate logistic regression models with 23 input factors were used to determine factors significantly associated with compliance. Survival analysis was conducted for both compliant and non‐compliant treatments.ResultsOverall, 656 patients met the inclusion criteria. Protocol compliance was 16%. Alternative dose/fractionation was responsible for 49% of non‐compliant treatments with 30% receiving an alternative curative fractionation. Five of 23 factors (age at the start of radiotherapy, stage group, ECOG performance status, tumour location and alcoholism history) showed significant associations with protocol compliance on multivariate analysis. There was no significant difference in median survival between patients receiving protocol compliant treatment (15.1 months) and non‐compliant treatment (15.6 months).ConclusionAdherence to the eviQ curative radiotherapy protocol for stages I–III NSCLC was low. Alternative dose/fractionation schemes were the main reason for non‐compliance. Protocol compliance was not associated with outcome.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Chin2024b,

title = {Dosimetric Impact of Delineation and Motion Uncertainties on the Heart and Substructures in Lung Cancer Radiotherapy},

author = {V. Chin and R.N. Finnegan and P. Chlap and L. Holloway and D.I. Thwaites and J. Otton and G.P. Delaney and S.K. Vinod},

doi = {10.1016/j.clon.2024.04.002},

issn = {0936-6555},

year  = {2024},

date = {2024-07-00},

journal = {Clinical Oncology},

volume = {36},

number = {7},

pages = {420--429},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Abdel-Wahab2024,

title = {Addressing challenges in low-income and middle-income countries through novel radiotherapy research opportunities},

author = {May Abdel-Wahab and C Norman Coleman and Jesper Grau Eriksen and Peter Lee and Ryan Kraus and Ekaterina Harsdorf and Becky Lee and Adam Dicker and Ezra Hahn and Jai Prakash Agarwal and Pataje G S Prasanna and Michael MacManus and Paul Keall and Nina A Mayr and Barbara Alicja Jereczek-Fossa and Francesco Giammarile and In Ah Kim and Ajay Aggarwal and Grant Lewison and Jiade J Lu and Douglas Guedes de Castro and Feng-Ming (Spring) Kong and Haidy Afifi and Hamish Sharp and Verna Vanderpuye and Tajudeen Olasinde and Fadi Atrash and Luc Goethals and Benjamin W Corn},

doi = {10.1016/s1470-2045(24)00038-x},

issn = {1470-2045},

year  = {2024},

date = {2024-06-00},

journal = {The Lancet Oncology},

volume = {25},

number = {6},

pages = {e270--e280},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Hindley2024,

title = {A statistical learning framework for mapping indirect measurements of ergodic systems to emergent properties},

author = {Nicholas Hindley and Stephen J. DeVience and Ella Zhang and Leo L. Cheng and Matthew S. Rosen},

doi = {10.1016/j.jmro.2024.100151},

issn = {2666-4410},

year  = {2024},

date = {2024-06-00},

journal = {Journal of Magnetic Resonance Open},

volume = {19},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Madden2024,

title = {CBCT-DRRs superior to CT-DRRs for target-tracking applications for pancreatic SBRT},

author = {Levi Madden and Abdella Ahmed and Maegan Stewart and Danielle Chrystall and Adam Mylonas and Ryan Brown and Doan Trang Nguyen and Paul Keall and Jeremy Booth},

doi = {10.1088/2057-1976/ad3bb9},

issn = {2057-1976},

year  = {2024},

date = {2024-05-01},

journal = {Biomed. Phys. Eng. Express},

volume = {10},

number = {3},

publisher = {IOP Publishing},

abstract = {Abstract

               

                  Objective. In current radiograph-based intra-fraction markerless target-tracking, digitally reconstructed radiographs (DRRs) from planning CTs (CT-DRRs) are often used to train deep learning models that extract information from the intra-fraction radiographs acquired during treatment. Traditional DRR algorithms were designed for patient alignment (i.e. bone matching) and may not replicate the radiographic image quality of intra-fraction radiographs at treatment. Hypothetically, generating DRRs from pre-treatment Cone-Beam CTs (CBCT-DRRs) with DRR algorithms incorporating physical modelling of on-board-imagers (OBIs) could improve the similarity between intra-fraction radiographs and DRRs by eliminating inter-fraction variation and reducing image-quality mismatches between radiographs and DRRs. In this study, we test the two hypotheses that intra-fraction radiographs are more similar to CBCT-DRRs than CT-DRRs, and that intra-fraction radiographs are more similar to DRRs from algorithms incorporating physical models of OBI components than DRRs from algorithms omitting these models.

               

                  Approach. DRRs were generated from CBCT and CT image sets collected from 20 patients undergoing pancreas stereotactic body radiotherapy. CBCT-DRRs and CT-DRRs were generated replicating the treatment position of patients and the OBI geometry during intra-fraction radiograph acquisition. To investigate whether the modelling of physical OBI components influenced radiograph-DRR similarity, four DRR algorithms were applied for the generation of CBCT-DRRs and CT-DRRs, incorporating and omitting different combinations of OBI component models. The four DRR algorithms were: a traditional DRR algorithm, a DRR algorithm with source-spectrum modelling, a DRR algorithm with source-spectrum and detector modelling, and a DRR algorithm with source-spectrum, detector and patient material modelling. Similarity between radiographs and matched DRRs was quantified using Pearson’s correlation and Czekanowski’s index, calculated on a per-image basis. Distributions of correlations and indexes were compared to test each of the hypotheses. Distribution differences were determined to be statistically significant when Wilcoxon’s signed rank test and the Kolmogorov-Smirnov two sample test returned p ≤ 0.05 for both tests.

               

                  Main results. Intra-fraction radiographs were more similar to CBCT-DRRs than CT-DRRs for both metrics across all algorithms, with all p ≤ 0.007. Source-spectrum modelling improved radiograph-DRR similarity for both metrics, with all p < 10−6. OBI detector modelling and patient material modelling did not influence radiograph-DRR similarity for either metric.

               

                  Significance. Generating DRRs from pre-treatment CBCT-DRRs is feasible, and incorporating CBCT-DRRs into markerless target-tracking methods may promote improved target-tracking accuracies. Incorporating source-spectrum modelling into a treatment planning system’s DRR algorithms may reinforce the safe treatment of cancer patients by aiding in patient alignment.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Ma2024,

title = {Fully automatic online geometric calibration for non‐circular cone‐beam CT orbits using fiducials with unknown placement},

author = {Yiqun Q. Ma and Tess Reynolds and Tina Ehtiati and Clifford Weiss and Kelvin Hong and Nicholas Theodore and Grace J. Gang and J. Webster Stayman},

doi = {10.1002/mp.17041},

issn = {2473-4209},

year  = {2024},

date = {2024-05-00},

journal = {Medical Physics},

volume = {51},

number = {5},

pages = {3245--3264},

publisher = {Wiley},

abstract = {AbstractBackgroundCone‐beam CT (CBCT) with non‐circular scanning orbits can improve image quality for 3D intraoperative image guidance. However, geometric calibration of such scans can be challenging. Existing methods typically require a prior image, specialized phantoms, presumed repeatable orbits, or long computation time.PurposeWe propose a novel fully automatic online geometric calibration algorithm that does not require prior knowledge of fiducial configuration. The algorithm is fast, accurate, and can accommodate arbitrary scanning orbits and fiducial configurations.MethodsThe algorithm uses an automatic initialization process to eliminate human intervention in fiducial localization and an iterative refinement process to ensure robustness and accuracy. We provide a detailed explanation and implementation of the proposed algorithm. Physical experiments on a lab test bench and a clinical robotic C‐arm scanner were conducted to evaluate spatial resolution performance and robustness under realistic constraints.ResultsQualitative and quantitative results from the physical experiments demonstrate high accuracy, efficiency, and robustness of the proposed method. The spatial resolution performance matched that of our existing benchmark method, which used a 3D‐2D registration‐based geometric calibration algorithm.ConclusionsWe have demonstrated an automatic online geometric calibration method that delivers high spatial resolution and robustness performance. This methodology enables arbitrary scan trajectories and should facilitate translation of such acquisition methods in a clinical setting.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Lau2023,

title = {Faster and lower dose imaging: evaluating adaptive, constant gantry velocity and angular separation in fast low‐dose 4D cone beam CT imaging},

author = {Benjamin K. F. Lau and Owen Dillon and Shalini K. Vinod and Ricky T. O'Brien and Tess Reynolds},

doi = {10.1002/mp.16585},

issn = {2473-4209},

year  = {2024},

date = {2024-02-00},

journal = {Medical Physics},

volume = {51},

number = {2},

pages = {1364--1382},

publisher = {Wiley},

abstract = {AbstractBackgroundThe adoption of four‐dimensional cone beam computed tomography (4DCBCT) for image‐guided lung cancer radiotherapy is increasing, especially for hypofractionated treatments. However, the drawbacks of 4DCBCT include long scan times (∼240 s), inconsistent image quality, higher imaging dose than necessary, and streaking artifacts. With the emergence of linear accelerators that can acquire 4DCBCT scans in a short period of time (9.2 s) there is a need to examine the impact that these very fast gantry rotations have on 4DCBCT image quality.PurposeThis study investigates the impact of gantry velocity and angular separation between x‐ray projections on image quality and its implication for fast low‐dose 4DCBCT with emerging systems, such as the Varian Halcyon that provide fast gantry rotation and imaging. Large and uneven angular separation between x‐ray projections is known to reduce 4DCBCT image quality through increased streaking artifacts. However, it is not known when angular separation starts degrading image quality. The study assesses the impact of constant and adaptive gantry velocity and determines the level when angular gaps impair image quality using state‐of‐the‐art reconstruction methods.MethodsThis study considers fast low‐dose 4DCBCT acquisitions (60–80 s, 200‐projection scans). To assess the impact of adaptive gantry rotations, the angular position of x‐ray projections from adaptive 4DCBCT acquisitions from a 30‐patient clinical trial were analyzed (referred to as patient angular gaps). To assess the impact of angular gaps, variable and static angular gaps (20°, 30°, 40°) were introduced into evenly separated 200 projections (ideal angular separation). To simulate fast gantry rotations, which are on emerging linacs, constant gantry velocity acquisitions (9.2 s, 60 s, 120 s, 240 s) were simulated by sampling x‐ray projections at constant intervals using the patient breathing traces from the ADAPT clinical trial (ACTRN12618001440213). The 4D Extended Cardiac‐Torso (XCAT) digital phantom was used to simulate projections to remove patient‐specific image quality variables.Image reconstruction was performed using Feldkamp‐Davis‐Kress (FDK), McKinnon‐Bates (MKB), and Motion‐Compensated‐MKB (MCMKB) algorithms. Image quality was assessed using Structural Similarity‐Index‐Measure (SSIM), Contrast‐to‐Noise‐Ratio (CNR), Signal‐to‐Noise‐Ratio (SNR), Tissue‐Interface‐Width‐Diaphragm (TIW‐D), and Tissue‐Interface‐Width‐Tumor (TIW‐T).ResultsPatient angular gaps and variable angular gap reconstructions produced similar results to ideal angular separation reconstructions, while static angular gap reconstructions produced lower image quality metrics. For MCMKB‐reconstructions, average patient angular gaps produced SSIM‐0.98, CNR‐13.6, SNR‐34.8, TIW‐D‐1.5 mm, and TIW‐T‐2.0 mm, static angular gap 40° produced SSIM‐0.92, CNR‐6.8, SNR‐6.7, TIW‐D‐5.7 mm, and TIW‐T‐5.9 mm and ideal produced SSIM‐1.00, CNR‐13.6, SNR‐34.8, TIW‐D‐1.5 mm, and TIW‐T‐2.0 mm. All constant gantry velocity reconstructions produced lower image quality metrics than ideal angular separation reconstructions regardless of the acquisition time. Motion compensated reconstruction (MCMKB) produced the highest contrast images with low streaking artifacts.ConclusionVery fast 4DCBCT scans can be acquired provided that the entire scan range is adaptively sampled, and motion‐compensated reconstruction is performed. Importantly, the angular separation between x‐ray projections within each individual respiratory bin had minimal effect on the image quality of fast low‐dose 4DCBCT imaging. The results will assist the development of future 4DCBCT acquisition protocols that can now be achieved in very short time frames with emerging linear accelerators.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Lau2023b,

title = {Faster and lower dose imaging: evaluating adaptive, constant gantry velocity and angular separation in fast low‐dose 4D cone beam CT imaging},

author = {Benjamin K. F. Lau and Owen Dillon and Shalini K. Vinod and Ricky T. O'Brien and Tess Reynolds},

doi = {10.1002/mp.16585},

issn = {2473-4209},

year  = {2024},

date = {2024-02-00},

journal = {Medical Physics},

volume = {51},

number = {2},

pages = {1364--1382},

publisher = {Wiley},

abstract = {AbstractBackgroundThe adoption of four‐dimensional cone beam computed tomography (4DCBCT) for image‐guided lung cancer radiotherapy is increasing, especially for hypofractionated treatments. However, the drawbacks of 4DCBCT include long scan times (∼240 s), inconsistent image quality, higher imaging dose than necessary, and streaking artifacts. With the emergence of linear accelerators that can acquire 4DCBCT scans in a short period of time (9.2 s) there is a need to examine the impact that these very fast gantry rotations have on 4DCBCT image quality.PurposeThis study investigates the impact of gantry velocity and angular separation between x‐ray projections on image quality and its implication for fast low‐dose 4DCBCT with emerging systems, such as the Varian Halcyon that provide fast gantry rotation and imaging. Large and uneven angular separation between x‐ray projections is known to reduce 4DCBCT image quality through increased streaking artifacts. However, it is not known when angular separation starts degrading image quality. The study assesses the impact of constant and adaptive gantry velocity and determines the level when angular gaps impair image quality using state‐of‐the‐art reconstruction methods.MethodsThis study considers fast low‐dose 4DCBCT acquisitions (60–80 s, 200‐projection scans). To assess the impact of adaptive gantry rotations, the angular position of x‐ray projections from adaptive 4DCBCT acquisitions from a 30‐patient clinical trial were analyzed (referred to as patient angular gaps). To assess the impact of angular gaps, variable and static angular gaps (20°, 30°, 40°) were introduced into evenly separated 200 projections (ideal angular separation). To simulate fast gantry rotations, which are on emerging linacs, constant gantry velocity acquisitions (9.2 s, 60 s, 120 s, 240 s) were simulated by sampling x‐ray projections at constant intervals using the patient breathing traces from the ADAPT clinical trial (ACTRN12618001440213). The 4D Extended Cardiac‐Torso (XCAT) digital phantom was used to simulate projections to remove patient‐specific image quality variables.Image reconstruction was performed using Feldkamp‐Davis‐Kress (FDK), McKinnon‐Bates (MKB), and Motion‐Compensated‐MKB (MCMKB) algorithms. Image quality was assessed using Structural Similarity‐Index‐Measure (SSIM), Contrast‐to‐Noise‐Ratio (CNR), Signal‐to‐Noise‐Ratio (SNR), Tissue‐Interface‐Width‐Diaphragm (TIW‐D), and Tissue‐Interface‐Width‐Tumor (TIW‐T).ResultsPatient angular gaps and variable angular gap reconstructions produced similar results to ideal angular separation reconstructions, while static angular gap reconstructions produced lower image quality metrics. For MCMKB‐reconstructions, average patient angular gaps produced SSIM‐0.98, CNR‐13.6, SNR‐34.8, TIW‐D‐1.5 mm, and TIW‐T‐2.0 mm, static angular gap 40° produced SSIM‐0.92, CNR‐6.8, SNR‐6.7, TIW‐D‐5.7 mm, and TIW‐T‐5.9 mm and ideal produced SSIM‐1.00, CNR‐13.6, SNR‐34.8, TIW‐D‐1.5 mm, and TIW‐T‐2.0 mm. All constant gantry velocity reconstructions produced lower image quality metrics than ideal angular separation reconstructions regardless of the acquisition time. Motion compensated reconstruction (MCMKB) produced the highest contrast images with low streaking artifacts.ConclusionVery fast 4DCBCT scans can be acquired provided that the entire scan range is adaptively sampled, and motion‐compensated reconstruction is performed. Importantly, the angular separation between x‐ray projections within each individual respiratory bin had minimal effect on the image quality of fast low‐dose 4DCBCT imaging. The results will assist the development of future 4DCBCT acquisition protocols that can now be achieved in very short time frames with emerging linear accelerators.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Brighi2024,

title = {Reply to Letter from Price et al, re: Translation of focused ultrasound for blood-brain barrier opening in glioma},

author = {Caterina Brighi and Ekaterina Salimova and Michael de Veer and Simon Puttick and Gary Egan},

doi = {10.1016/j.jconrel.2022.06.042},

issn = {0168-3659},

year  = {2024},

date = {2024-02-00},

journal = {Journal of Controlled Release},

volume = {366},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{BenBouchta2024,

title = {The Remove-the-Mask Open-Source head and neck Surface-Guided radiation therapy system},

author = {Youssef Ben Bouchta and Mark Gardner and Chandrima Sengupta and Julia Johnson and Paul Keall},

doi = {10.1016/j.phro.2024.100541},

issn = {2405-6316},

year  = {2024},

date = {2024-01-00},

journal = {Physics and Imaging in Radiation Oncology},

volume = {29},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Lombardo2024,

title = {Real-time motion management in MRI-guided radiotherapy: Current status and AI-enabled prospects},

author = {Elia Lombardo and Jennifer Dhont and Denis Page and Cristina Garibaldi and Luise A. Künzel and Coen Hurkmans and Rob H.N. Tijssen and Chiara Paganelli and Paul Z.Y. Liu and Paul J. Keall and Marco Riboldi and Christopher Kurz and Guillaume Landry and Davide Cusumano and Marco Fusella and Lorenzo Placidi},

doi = {10.1016/j.radonc.2023.109970},

issn = {0167-8140},

year  = {2024},

date = {2024-01-00},

journal = {Radiotherapy and Oncology},

volume = {190},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Sengupta2024b,

title = {The first clinical implementation of real-time 6 degree-of-freedom image-guided radiotherapy for liver SABR patients},

author = {Chandrima Sengupta and Doan Trang Nguyen and Trevor Moodie and Daniel Mason and Jianjie Luo and Trent Causer and Sau Fan Liu and Elizabeth Brown and Lauren Inskip and Maryam Hazem and Menglei Chao and Tim Wang and Yoo Y. Lee and Kirsten van Gysen and Emma Sullivan and Eireann Cosgriff and Prabhakar Ramachandran and Per Poulsen and Jeremy Booth and Ricky O'Brien and Peter Greer and Paul Keall},

doi = {10.1016/j.radonc.2023.110031},

issn = {0167-8140},

year  = {2024},

date = {2024-01-00},

journal = {Radiotherapy and Oncology},

volume = {190},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Lombardo2024b,

title = {Real-time motion management in MRI-guided radiotherapy: Current status and AI-enabled prospects},

author = {Elia Lombardo and Jennifer Dhont and Denis Page and Cristina Garibaldi and Luise A. Künzel and Coen Hurkmans and Rob H.N. Tijssen and Chiara Paganelli and Paul Z.Y. Liu and Paul J. Keall and Marco Riboldi and Christopher Kurz and Guillaume Landry and Davide Cusumano and Marco Fusella and Lorenzo Placidi},

doi = {10.1016/j.radonc.2023.109970},

issn = {0167-8140},

year  = {2024},

date = {2024-01-00},

journal = {Radiotherapy and Oncology},

volume = {190},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Sengupta2024c,

title = {The first clinical implementation of real-time 6 degree-of-freedom image-guided radiotherapy for liver SABR patients},

author = {Chandrima Sengupta and Doan Trang Nguyen and Trevor Moodie and Daniel Mason and Jianjie Luo and Trent Causer and Sau Fan Liu and Elizabeth Brown and Lauren Inskip and Maryam Hazem and Menglei Chao and Tim Wang and Yoo Y. Lee and Kirsten van Gysen and Emma Sullivan and Eireann Cosgriff and Prabhakar Ramachandran and Per Poulsen and Jeremy Booth and Ricky O'Brien and Peter Greer and Paul Keall},

doi = {10.1016/j.radonc.2023.110031},

issn = {0167-8140},

year  = {2024},

date = {2024-01-00},

journal = {Radiotherapy and Oncology},

volume = {190},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Shan2024,

title = {Image Reconstruction With B₀ Inhomogeneity Using a Deep Unrolled Network on an Open-Bore MRI-Linac},

author = {Shanshan Shan and Yang Gao and David Waddington and Hongli Chen and Brendan Whelan and Paul Liu and Yaohui Wang and Chunyi Liu and Hongping Gan and Mingyuan Gao and Feng Liu},

doi = {10.1109/tim.2024.3481545},

issn = {1557-9662},

year  = {2024},

date = {2024-00-00},

journal = {IEEE Trans. Instrum. Meas.},

volume = {73},

pages = {1--9},

publisher = {Institute of Electrical and Electronics Engineers (IEEE)},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Brighi2023b,

title = {The MANGO study: a prospective investigation of oxygen enhanced and blood-oxygen level dependent MRI as imaging biomarkers of hypoxia in glioblastoma},

author = {Caterina Brighi and David E. J. Waddington and Paul J. Keall and Jeremy Booth and Kieran O’Brien and Shona Silvester and Jonathon Parkinson and Marco Mueller and Jackie Yim and Dale L. Bailey and Michael Back and James Drummond},

doi = {10.3389/fonc.2023.1306164},

issn = {2234-943X},

year  = {2023},

date = {2023-12-19},

journal = {Front. Oncol.},

volume = {13},

publisher = {Frontiers Media SA},

abstract = {BackgroundGlioblastoma (GBM) is the most aggressive type of brain cancer, with a 5-year survival rate of ~5% and most tumours recurring locally within months of first-line treatment. Hypoxia is associated with worse clinical outcomes in GBM, as it leads to localized resistance to radiotherapy and subsequent tumour recurrence. Current standard of care treatment does not account for tumour hypoxia, due to the challenges of mapping tumour hypoxia in routine clinical practice. In this clinical study, we aim to investigate the role of oxygen enhanced (OE) and blood-oxygen level dependent (BOLD) MRI as non-invasive imaging biomarkers of hypoxia in GBM, and to evaluate their potential role in dose-painting radiotherapy planning and treatment response assessment.MethodsThe primary endpoint is to evaluate the quantitative and spatial correlation between OE and BOLD MRI measurements and [18F]MISO values of uptake in the tumour. The secondary endpoints are to evaluate the repeatability of MRI biomarkers of hypoxia in a test-retest study, to estimate the potential clinical benefits of using MRI biomarkers of hypoxia to guide dose-painting radiotherapy, and to evaluate the ability of MRI biomarkers of hypoxia to assess treatment response. Twenty newly diagnosed GBM patients will be enrolled in this study. Patients will undergo standard of care treatment while receiving additional OE/BOLD MRI and [18F]MISO PET scans at several timepoints during treatment. The ability of OE/BOLD MRI to map hypoxic tumour regions will be evaluated by assessing spatial and quantitative correlations with areas of hypoxic tumour identified via [18F]MISO PET imaging.DiscussionMANGO (Magnetic resonance imaging of hypoxia for radiation treatment guidance in glioblastoma multiforme) is a diagnostic/prognostic study investigating the role of imaging biomarkers of hypoxia in GBM management. The study will generate a large amount of longitudinal multimodal MRI and PET imaging data that could be used to unveil dynamic changes in tumour physiology that currently limit treatment efficacy, thereby providing a means to develop more effective and personalised treatments.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Byrne2023,

title = {BRAVEHeart: a randomised trial comparing the accuracy of Breathe Well and RPM for deep inspiration breath hold breast cancer radiotherapy},

author = {Hilary L. Byrne and Elisabeth Steiner and Jeremy Booth and Gillian Lamoury and Marita Morgia and Kylie Richardson and Leigh Ambrose and Kuldeep Makhija and Cameron Stanton and Benjamin Zwan and Regina Bromley and John Atyeo and Shona Silvester and Natalie Plant and Paul Keall},

doi = {10.1186/s13063-023-07072-y},

issn = {1745-6215},

year  = {2023},

date = {2023-12-00},

journal = {Trials},

volume = {24},

number = {1},

publisher = {Springer Science and Business Media LLC},

abstract = {Abstract

                Background

                Deep inspiration breath hold (DIBH) reduces radiotherapy cardiac dose for left-sided breast cancer patients. The primary aim of the BRAVEHeart (Breast Radiotherapy Audio Visual Enhancement for sparing the Heart) trial is to assess the accuracy and usability of a novel device, Breathe Well, for DIBH guidance for left-sided breast cancer patients. Breathe Well will be compared to an adapted widely available monitoring system, the Real-time Position Management system (RPM).

              

                Methods

                BRAVEHeart is a single institution prospective randomised trial of two DIBH devices. BRAVEHeart will assess the DIBH accuracy for Breathe Well and RPM during left-sided breast cancer radiotherapy. After informed consent has been obtained, 40 patients will be randomised into two equal groups, the experimental arm (Breathe Well) and the control arm (RPM with in-house modification of an added patient screen). The primary hypothesis of BRAVEHeart is that the accuracy of Breathe Well in maintaining the position of the chest during DIBH is superior to the RPM system. Accuracy will be measured by comparing chest wall motion extracted from images acquired of the treatment field during breast radiotherapy for patients treated using the Breathe Well system and those using the RPM system.

              

                Discussion

                The Breathe Well device uses a depth camera to monitor the chest surface while the RPM system monitors a block on the patient’s abdomen. The hypothesis of this trial is that the chest surface is a better surrogate for the internal chest wall motion used as a measure of treatment accuracy. The Breathe Well device aims to deliver an easy-to-use implementation of surface monitoring. The findings from the study will help inform the technology choice for other centres performing DIBH.

              

                Trial registration

                ClinicalTrials.govNCT02881203. Registered on 26 August 2016.

              },

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Debrot2023,

title = {Nano X Image Guidance in radiation therapy: feasibility study protocol for cone beam computed tomography imaging with gravity-induced motion},

author = {Emily Debrot and Paul Liu and Mark Gardner and Soo Min Heng and Chin Hwa Chan and Stephanie Corde and Simon Downes and Michael Jackson and Paul Keall},

doi = {10.1186/s40814-023-01340-z},

issn = {2055-5784},

year  = {2023},

date = {2023-12-00},

journal = {Pilot Feasibility Stud},

volume = {9},

number = {1},

publisher = {Springer Science and Business Media LLC},

abstract = {Abstract

                Background

                This paper describes the protocol for the Nano X Image Guidance (Nano X IG) trial, a single-institution, clinical imaging study. The Nano X is a prototype fixed-beam radiotherapy system developed to investigate the feasibility of a low-cost, compact radiotherapy system to increase global access to radiation therapy. This study aims to assess the feasibility of volumetric image guidance with cone beam computed tomography (CBCT) acquired during horizontal patient rotation on the Nano X radiotherapy system.

              

                Methods

                In the Nano X IG study, we will determine whether radiotherapy image guidance can be performed with the Nano X radiotherapy system where the patient is horizontally rotated while scan projections are acquired. We will acquire both conventional CBCT scans and Nano X CBCT scans for 30 patients aged 18 and above and receiving radiotherapy for head/neck or upper abdomen cancers. For each patient, a panel of experts will assess the image quality of Nano X CBCT scans against conventional CBCT scans. Each patient will receive two Nano X CBCT scans to determine the image quality reproducibility, the extent and reproducibility of patient motion and assess patient tolerance.

              

                Discussion

                Fixed-beam radiotherapy systems have the potential to help ease the current shortfall and increase global access to radiotherapy treatment. Advances in image guidance could facilitate fixed-beam radiotherapy using horizontal patient rotation. The efficacy of this radiotherapy approach is dependent on our ability to image and adapt to motion due to rotation and for patients to tolerate rotation during treatment.

              

                Trial registration

                ClinicalTrials.gov, NCT04488224. Registered on 27 July 2020.



              },

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Dillon2023,

title = {X-ray source arrays for volumetric imaging during radiotherapy treatment},

author = {Owen Dillon and Tess Reynolds and Ricky T. O’Brien},

doi = {10.1038/s41598-023-36708-x},

issn = {2045-2322},

year  = {2023},

date = {2023-12-00},

journal = {Sci Rep},

volume = {13},

number = {1},

publisher = {Springer Science and Business Media LLC},

abstract = {AbstractThis work presents a novel hardware configuration for radiotherapy systems to enable fast 3D X-ray imaging before and during treatment delivery. Standard external beam radiotherapy linear accelerators (linacs) have a single X-ray source and detector located at ± 90° from the treatment beam respectively. The entire system can be rotated around the patient acquiring multiple 2D X-ray images to create a 3D cone-beam Computed Tomography (CBCT) image before treatment delivery to ensure the tumour and surrounding organs align with the treatment plan. Scanning with a single source is slow relative to patient respiration or breath holds and cannot be performed during treatment delivery, limiting treatment delivery accuracy in the presence of patient motion and excluding some patients from concentrated treatment plans that would be otherwise expected to have improved outcomes. This simulation study investigated whether recent advances in carbon nanotube (CNT) field emission source arrays, high frame rate (60 Hz) flat panel detectors and compressed sensing reconstruction algorithms could circumvent imaging limitations of current linacs. We investigated a novel hardware configuration incorporating source arrays and high frame rate detectors into an otherwise standard linac. We investigated four potential pre-treatment scan protocols that could be achieved in a 17 s breath hold or 2–10 1 s breath holds. Finally, we demonstrated for the first time volumetric X-ray imaging during treatment delivery by using source arrays, high frame rate detectors and compressed sensing. Image quality was assessed quantitatively over the CBCT geometric field of view as well as across each axis through the tumour centroid. Our results demonstrate that source array imaging enables larger volumes to be imaged with acquisitions as short as 1 s albeit with reduced image quality arising from lower photon flux and shorter imaging arcs.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Trada2023,

title = {Changes in serial multiparametric MRI and FDG-PET/CT functional imaging during radiation therapy can predict treatment response in patients with head and neck cancer},

author = {Yuvnik Trada and Paul Keall and Michael Jameson and Daniel Moses and Peter Lin and Phillip Chlap and Lois Holloway and Myo Min and Dion Forstner and Allan Fowler and Mark T. Lee},

doi = {10.1007/s00330-023-09843-2},

issn = {1432-1084},

year  = {2023},

date = {2023-12-00},

journal = {Eur Radiol},

volume = {33},

number = {12},

pages = {8788--8799},

publisher = {Springer Science and Business Media LLC},

abstract = {Abstract

                Objectives

                To test if tumour changes measured using combination of diffusion-weighted imaging (DWI) MRI and FDG-PET/CT performed serially during radiotherapy (RT) in mucosal head and neck carcinoma can predict treatment response.

              

                Methods

                Fifty-five patients from two prospective imaging biomarker studies were analysed. FDG-PET/CT was performed at baseline, during RT (week 3), and post RT (3 months). DWI was performed at baseline, during RT (weeks 2, 3, 5, 6), and post RT (1 and 3 months). The ADCmean from DWI and FDG-PET parameters SUVmax, SUVmean, metabolic tumour volume (MTV), and total lesion glycolysis (TLG) were measured. Absolute and relative change (%∆) in DWI and PET parameters were correlated to 1-year local recurrence. Patients were categorised into favourable, mixed, and unfavourable imaging response using optimal cut-off (OC) values of DWI and FDG-PET parameters and correlated to local control.

              

                Results

                The 1-year local, regional, and distant recurrence rates were 18.2% (10/55), 7.3% (4/55), and 12.7% (7/55), respectively. ∆Week 3 ADCmean (AUC 0.825, p = 0.003; OC ∆ > 24.4%) and ∆MTV (AUC 0.833, p = 0.001; OC ∆ > 50.4%) were the best predictors of local recurrence. Week 3 was the optimal time point for assessing DWI imaging response. Using a combination of ∆ADCmean and ∆MTV improved the strength of correlation to local recurrence (p ≤ 0.001). In patients who underwent both week 3 MRI and FDG-PET/CT, significant differences in local recurrence rates were seen between patients with favourable (0%), mixed (17%), and unfavourable (78%) combined imaging response.

              

                Conclusions

                Changes in mid-treatment DWI and FDG-PET/CT imaging can predict treatment response and could be utilised in the design of future adaptive clinical trials.

              

                Clinical relevance statement

                Our study shows the complementary information provided by two functional imaging modalities for mid-treatment response prediction in patients with head and neck cancer.

              

                Key Points

                •FDG-PET/CT and DWI MRI changes in tumour during radiotherapy in head and neck cancer can predict treatment response.

                •Combination of FDG-PET/CT and DWI parameters improved correlation to clinical outcome.

                •Week 3 was the optimal time point for DWI MRI imaging response assessment.

              },

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Brighi2023,

title = {Comparison between [68Ga]Ga-PSMA-617 and [18F]FET PET as Imaging Biomarkers in Adult Recurrent Glioblastoma},

author = {Caterina Brighi and Simon Puttick and Amanda Woods and Paul Keall and Paul A. Tooney and David E. J. Waddington and Vicki Sproule and Stephen Rose and Michael Fay},

doi = {10.3390/ijms242216208},

issn = {1422-0067},

year  = {2023},

date = {2023-11-00},

journal = {IJMS},

volume = {24},

number = {22},

publisher = {MDPI AG},

abstract = {The aim of this prospective clinical study was to evaluate the potential of the prostate specific membrane antigen (PSMA) targeting ligand, [68Ga]-PSMA–Glu–NH–CO–NH–Lys-2-naphthyl-L-Ala-cyclohexane-DOTA ([68Ga]Ga-PSMA-617) as a positron emission tomography (PET) imaging biomarker in recurrent glioblastoma patients. Patients underwent [68Ga]Ga-PSMA-617 and O-(2-[18F]-fluoroethyl)-L-tyrosine ([18F]FET) PET scans on two separate days. [68Ga]Ga-PSMA-617 tumour selectivity was assessed by comparing tumour volume delineation and by assessing the intra-patient correlation between tumour uptake on [68Ga]Ga-PSMA-617 and [18F]FET PET images. [68Ga]Ga-PSMA-617 tumour specificity was evaluated by comparing its tumour-to-brain ratio (TBR) with [18F]FET TBR and its tumour volume with the magnetic resonance imaging (MRI) contrast-enhancing (CE) tumour volume. Ten patients were recruited in this study. [68Ga]Ga-PSMA-617-avid tumour volume was larger than the [18F]FET tumour volume (p = 0.063). There was a positive intra-patient correlation (median Pearson r = 0.51; p < 0.0001) between [68Ga]Ga-PSMA-617 and [18F]FET in the tumour volume. [68Ga]Ga-PSMA-617 had significantly higher TBR (p = 0.002) than [18F]FET. The [68Ga]Ga-PSMA-617-avid tumour volume was larger than the CE tumour volume (p = 0.0039). Overall, accumulation of [68Ga]-Ga-PSMA-617 beyond [18F]FET-avid tumour regions suggests the presence of neoangiogenesis in tumour regions that are not overly metabolically active yet. Higher tumour specificity suggests that [68Ga]-Ga-PSMA-617 could be a better imaging biomarker for recurrent tumour delineation and secondary treatment planning than [18F]FET and CE MRI.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Lombardo2023,

title = {Experimental comparison of linear regression and LSTM motion prediction models for MLC‐tracking on an MRI‐linac},

author = {Elia Lombardo and Paul Z. Y. Liu and David E. J. Waddington and James Grover and Brendan Whelan and Esther Wong and Michael Reiner and Stefanie Corradini and Claus Belka and Marco Riboldi and Christopher Kurz and Guillaume Landry and Paul J. Keall},

doi = {10.1002/mp.16770},

issn = {2473-4209},

year  = {2023},

date = {2023-11-00},

journal = {Medical Physics},

volume = {50},

number = {11},

pages = {7083--7092},

publisher = {Wiley},

abstract = {AbstractBackgroundMagnetic resonance imaging (MRI)‐guided radiotherapy with multileaf collimator (MLC)‐tracking is a promising technique for intra‐fractional motion management, achieving high dose conformality without prolonging treatment times. To improve beam‐target alignment, the geometric error due to system latency should be reduced by using temporal prediction.PurposeTo experimentally compare linear regression (LR) and long‐short‐term memory (LSTM) motion prediction models for MLC‐tracking on an MRI‐linac using multiple patient‐derived traces with different complexities.MethodsExperiments were performed on a prototype 1.0 T MRI‐linac capable of MLC‐tracking. A motion phantom was programmed to move a target in superior‐inferior (SI) direction according to eight lung cancer patient respiratory motion traces. Target centroid positions were localized from sagittal 2D cine MRIs acquired at 4 Hz using a template matching algorithm. The centroid positions were input to one of four motion prediction models. We used (1) a LSTM network which had been optimized in a previous study on patient data from another cohort (offline LSTM). We also used (2) the same LSTM model as a starting point for continuous re‐optimization of its weights during the experiment based on recent motion (offline+online LSTM). Furthermore, we implemented (3) a continuously updated LR model, which was solely based on recent motion (online LR). Finally, we used (4) the last available target centroid without any changes as a baseline (no‐predictor). The predictions of the models were used to shift the MLC aperture in real‐time. An electronic portal imaging device (EPID) was used to visualize the target and MLC aperture during the experiments. Based on the EPID frames, the root‐mean‐square error (RMSE) between the target and the MLC aperture positions was used to assess the performance of the different motion predictors. Each combination of motion trace and prediction model was repeated twice to test stability, for a total of 64 experiments.ResultsThe end‐to‐end latency of the system was measured to be (389 ± 15) ms and was successfully mitigated by both LR and LSTM models. The offline+online LSTM was found to outperform the other models for all investigated motion traces. It obtained a median RMSE over all traces of (2.8 ± 1.3) mm, compared to the (3.2 ± 1.9) mm of the offline LSTM, the (3.3 ± 1.4) mm of the online LR and the (4.4 ± 2.4) mm when using the no‐predictor. According to statistical tests, differences were significant (p‐value <0.05) among all models in a pair‐wise comparison, but for the offline LSTM and online LR pair. The offline+online LSTM was found to be more reproducible than the offline LSTM and the online LR with a maximum deviation in RMSE between two measurements of 10%.ConclusionsThis study represents the first experimental comparison of different prediction models for MRI‐guided MLC‐tracking using several patient‐derived respiratory motion traces. We have shown that among the investigated models, continuously re‐optimized LSTM networks are the most promising to account for the end‐to‐end system latency in MRI‐guided radiotherapy with MLC‐tracking.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Shan2023,

title = {Distortion‐corrected image reconstruction with deep learning on an MRI‐Linac},

author = {Shanshan Shan and Yang Gao and Paul Z. Y. Liu and Brendan Whelan and Hongfu Sun and Bin Dong and Feng Liu and David E. J. Waddington},

doi = {10.1002/mrm.29684},

issn = {1522-2594},

year  = {2023},

date = {2023-09-00},

journal = {Magnetic Resonance in Med},

volume = {90},

number = {3},

pages = {963--977},

publisher = {Wiley},

abstract = {PurposeMRI is increasingly utilized for image‐guided radiotherapy due to its outstanding soft‐tissue contrast and lack of ionizing radiation. However, geometric distortions caused by gradient nonlinearities (GNLs) limit anatomical accuracy, potentially compromising the quality of tumor treatments. In addition, slow MR acquisition and reconstruction limit the potential for effective image guidance. Here, we demonstrate a deep learning‐based method that rapidly reconstructs distortion‐corrected images from raw k‐space data for MR‐guided radiotherapy applications.MethodsWe leverage recent advances in interpretable unrolling networks to develop a Distortion‐Corrected Reconstruction Network (DCReconNet) that applies convolutional neural networks (CNNs) to learn effective regularizations and nonuniform fast Fourier transforms for GNL‐encoding. DCReconNet was trained on a public MR brain dataset from 11 healthy volunteers for fully sampled and accelerated techniques, including parallel imaging (PI) and compressed sensing (CS). The performance of DCReconNet was tested on phantom, brain, pelvis, and lung images acquired on a 1.0T MRI‐Linac. The DCReconNet, CS‐, PI‐and UNet‐based reconstructed image quality was measured by structural similarity (SSIM) and RMS error (RMSE) for numerical comparisons. The computation time and residual distortion for each method were also reported.ResultsImaging results demonstrated that DCReconNet better preserves image structures compared to CS‐ and PI‐based reconstruction methods. DCReconNet resulted in the highest SSIM (0.95 median value) and lowest RMSE (<0.04) on simulated brain images with four times acceleration. DCReconNet is over 10‐times faster than iterative, regularized reconstruction methods.ConclusionsDCReconNet provides fast and geometrically accurate image reconstruction and has the potential for MRI‐guided radiotherapy applications.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Trada2023b,

title = {Mid-treatment 18F-FDG PET imaging changes in parotid gland correlates to radiation-induced xerostomia},

author = {Yuvnik Trada and Mark T. Lee and Michael G. Jameson and Phillip Chlap and Paul Keall and Daniel Moses and Peter Lin and Allan Fowler},

doi = {10.1016/j.radonc.2023.109745},

issn = {0167-8140},

year  = {2023},

date = {2023-09-00},

journal = {Radiotherapy and Oncology},

volume = {186},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Whelan2023,

title = {ParticlePhaseSpace: A python package for streamlined

import, analysis, and export of particle phase space data},

author = {Brendan Whelan and Leo Esnault},

doi = {10.21105/joss.05375},

issn = {2475-9066},

year  = {2023},

date = {2023-09-00},

journal = {JOSS},

volume = {8},

number = {89},

publisher = {The Open Journal},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Richardson2023,

title = {Genitourinary Quality-of-Life Comparison Between Urethral Sparing Prostate Stereotactic Body Radiation Therapy Monotherapy and Virtual High-Dose-Rate Brachytherapy Boost},

author = {Matthew Richardson and Mark Sidhom and Paul Keall and Lucy Leigh and Helen Ball and Joseph Bucci and Sarah Gallagher and Peter Greer and Amy J. Hayden and Andrew Kneebone and David Pryor and Shankar Siva and Jarad Martin},

doi = {10.1016/j.ijrobp.2023.02.049},

issn = {0360-3016},

year  = {2023},

date = {2023-08-00},

journal = {International Journal of Radiation Oncology*Biology*Physics},

volume = {116},

number = {5},

pages = {1069--1078},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Hindley2023,

title = {A patient-specific deep learning framework for 3D motion estimation and volumetric imaging during lung cancer radiotherapy},

author = {Nicholas Hindley and Chun-Chien Shieh and Paul Keall},

doi = {10.1088/1361-6560/ace1d0},

issn = {1361-6560},

year  = {2023},

date = {2023-07-21},

journal = {Phys. Med. Biol.},

volume = {68},

number = {14},

publisher = {IOP Publishing},

abstract = {Abstract

               

                  Objective. Respiration introduces a constant source of irregular motion that poses a significant challenge for the precise irradiation of thoracic and abdominal cancers. Current real-time motion management strategies require dedicated systems that are not available in most radiotherapy centers. We sought to develop a system that estimates and visualises the impact of respiratory motion in 3D given the 2D images acquired on a standard linear accelerator. Approach. In this paper we introduce Voxelmap, a patient-specific deep learning framework that achieves 3D motion estimation and volumetric imaging using the data and resources available in standard clinical settings. Here we perform a simulation study of this framework using imaging data from two lung cancer patients. Main results. Using 2D images as input and 3D–3D Elastix registrations as ground-truth, Voxelmap was able to continuously predict 3D tumor motion with mean errors of 0.1 ± 0.5, −0.6 ± 0.8, and 0.0 ± 0.2 mm along the left–right, superior–inferior, and anterior–posterior axes respectively. Voxelmap also predicted 3D thoracoabdominal motion with mean errors of −0.1 ± 0.3, −0.1 ± 0.6, and −0.2 ± 0.2 mm respectively. Moreover, volumetric imaging was achieved with mean average error 0.0003, root-mean-squared error 0.0007, structural similarity 1.0 and peak-signal-to-noise ratio 65.8. Significance. The results of this study demonstrate the possibility of achieving 3D motion estimation and volumetric imaging during lung cancer treatments on a standard linear accelerator.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Gardner2023,

title = {Realistic CT data augmentation for accurate deep‐learning based segmentation of head and neck tumors in kV images acquired during radiation therapy},

author = {Mark Gardner and Youssef Ben Bouchta and Adam Mylonas and Marco Mueller and Chen Cheng and Phillip Chlap and Robert Finnegan and Jonathan Sykes and Paul J Keall and Doan Trang Nguyen},

doi = {10.1002/mp.16388},

issn = {2473-4209},

year  = {2023},

date = {2023-07-00},

journal = {Medical Physics},

volume = {50},

number = {7},

pages = {4206--4219},

publisher = {Wiley},

abstract = {AbstractBackgroundUsing radiation therapy (RT) to treat head and neck (H&N) cancers requires precise targeting of the tumor to avoid damaging the surrounding healthy organs. Immobilisation masks and planning target volume margins are used to attempt to mitigate patient motion during treatment, however patient motion can still occur. Patient motion during RT can lead to decreased treatment effectiveness and a higher chance of treatment related side effects. Tracking tumor motion would enable motion compensation during RT, leading to more accurate dose delivery.PurposeThe purpose of this paper is to develop a method to detect and segment the tumor in kV images acquired during RT. Unlike previous tumor segmentation methods for kV images, in this paper, a process for generating realistic and synthetic CT deformations was developed to augment the training data and make the segmentation method robust to patient motion. Detecting the tumor in 2D kV images is a necessary step toward 3D tracking of the tumor position during treatment.MethodIn this paper, a conditional generative adversarial network (cGAN) is presented that can detect and segment the gross tumor volume (GTV) in kV images acquired during H&N RT. Retrospective data from 15 H&N cancer patients obtained from the Cancer Imaging Archive were used to train and test patient‐specific cGANs. The training data consisted of digitally reconstructed radiographs (DRRs) generated from each patient's planning CT and contoured GTV. Training data was augmented by using synthetically deformed CTs to generate additional DRRs (in total 39 600 DRRs per patient or 25 200 DRRs for nasopharyngeal patients) containing realistic patient motion. The method for deforming the CTs was a novel deformation method based on simulating head rotation and internal tumor motion. The testing dataset consisted of 1080 DRRs for each patient, obtained by deforming the planning CT and GTV at different magnitudes to the training data.The accuracy of the generated segmentations was evaluated by measuring the segmentation centroid error, Dice similarity coefficient (DSC) and mean surface distance (MSD). This paper evaluated the hypothesis that when patient motion occurs, using a cGAN to segment the GTV would create a more accurate segmentation than no‐tracking segmentations from the original contoured GTV, the current standard‐of‐care. This hypothesis was tested using the 1‐tailed Mann‐Whitney U‐test.ResultsThe magnitude of our cGAN segmentation centroid error was (mean ± standard deviation) 1.1 ± 0.8 mm and the DSC and MSD values were 0.90 ± 0.03 and 1.6 ± 0.5 mm, respectively. Our cGAN segmentation method reduced the segmentation centroid error (p < 0.001), and MSD (p = 0.031) when compared to the no‐tracking segmentation, but did not significantly increase the DSC (p = 0.294).ConclusionsThe accuracy of our cGAN segmentation method demonstrates the feasibility of this method for H&N cancer patients during RT. Accurate tumor segmentation of H&N tumors would allow for intrafraction monitoring methods to compensate for tumor motion during treatment, ensuring more accurate dose delivery and enabling better H&N cancer patient outcomes.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Chrystall2023,

title = {Deep learning enables MV-based real-time image guided radiation therapy for prostate cancer patients},

author = {Danielle Chrystall and Adam Mylonas and Emily Hewson and Jarad Martin and Paul Keall and Jeremy Booth and Doan Trang Nguyen},

doi = {10.1088/1361-6560/acc77c},

issn = {1361-6560},

year  = {2023},

date = {2023-05-07},

journal = {Phys. Med. Biol.},

volume = {68},

number = {9},

publisher = {IOP Publishing},

abstract = {Abstract

               

                  Objective. Using MV images for real-time image guided radiation therapy (IGRT) is ideal as it does not require additional imaging equipment, adds no additional imaging dose and provides motion data in the treatment beam frame of reference. However, accurate tracking using MV images is challenging due to low contrast and modulated fields. Here, a novel real-time marker tracking system based on a convolutional neural network (CNN) classifier was developed and evaluated on retrospectively acquired patient data for MV-based IGRT for prostate cancer patients. Approach. MV images, acquired from 29 volumetric modulated arc therapy (VMAT) prostate cancer patients treated in a multi-institutional clinical trial, were used to train and evaluate a CNN-based marker tracking system. The CNN was trained using labelled MV images from 9 prostate cancer patients (35 fractions) with implanted markers. CNN performance was evaluated on an independent cohort of unseen MV images from 20 patients (78 fractions), using a Precision–Recall curve (PRC), area under the PRC plot (AUC) and sensitivity and specificity. The accuracy of the tracking system was evaluated on the same unseen dataset and quantified by calculating mean absolute (±1 SD) and [1st, 99th] percentiles of the geometric tracking error in treatment beam co-ordinates using manual identification as the ground truth. Main results. The CNN had an AUC of 0.99, sensitivity of 98.31% and specificity of 99.87%. The mean absolute geometric tracking error was 0.30 ± 0.27 and 0.35 ± 0.31 mm in the lateral and superior–inferior directions of the MV images, respectively. The [1st, 99th] percentiles of the error were [−1.03, 0.90] and [−1.12, 1.12] mm in the lateral and SI directions, respectively. Significance. The high classification performance on unseen MV images demonstrates the CNN can successfully identify implanted prostate markers. Furthermore, the sub-millimetre accuracy and precision of the marker tracking system demonstrates potential for adaptation to real-time applications.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Ahmed2023,

title = {Evaluation of deep learning based implanted fiducial markers tracking in pancreatic cancer patients},

author = {Abdella M Ahmed and Maegan Gargett and Levi Madden and Adam Mylonas and Danielle Chrystall and Ryan Brown and Adam Briggs and Trang Nguyen and Paul Keall and Andrew Kneebone and George Hruby and Jeremy Booth},

doi = {10.1088/2057-1976/acb550},

issn = {2057-1976},

year  = {2023},

date = {2023-05-01},

journal = {Biomed. Phys. Eng. Express},

volume = {9},

number = {3},

publisher = {IOP Publishing},

abstract = {Abstract

               Real-time target position verification during pancreas stereotactic body radiation therapy (SBRT) is important for the detection of unplanned tumour motions. Fast and accurate fiducial marker segmentation is a Requirement of real-time marker-based verification. Deep learning (DL) segmentation techniques are ideal because they don’t require additional learning imaging or prior marker information (e.g., shape, orientation). In this study, we evaluated three DL frameworks for marker tracking applied to pancreatic cancer patient data. The DL frameworks evaluated were (1) a convolutional neural network (CNN) classifier with sliding window, (2) a pretrained you-only-look-once (YOLO) version-4 architecture, and (3) a hybrid CNN-YOLO. Intrafraction kV images collected during pancreas SBRT treatments were used as training data (44 fractions, 2017 frames). All patients had 1-4 implanted fiducial markers. Each model was evaluated on unseen kV images (42 fractions, 2517 frames). The ground truth was calculated from manual segmentation and triangulation of markers in orthogonal paired kV/MV images. The sensitivity, specificity, and area under the precision-recall curve (AUC) were calculated. In addition, the mean-absolute-error (MAE), root-mean-square-error (RMSE) and standard-error-of-mean (SEM) were calculated for the centroid of the markers predicted by the models, relative to the ground truth. The sensitivity and specificity of the CNN model were 99.41% and 99.69%, respectively. The AUC was 0.9998. The average precision of the YOLO model for different values of recall was 96.49%. The MAE of the three models in the left-right, superior-inferior, and anterior-posterior directions were under 0.88 ± 0.11 mm, and the RMSE were under 1.09 ± 0.12 mm. The detection times per frame on a GPU were 48.3, 22.9, and 17.1 milliseconds for the CNN, YOLO, and CNN-YOLO, respectively. The results demonstrate submillimeter accuracy of marker position predicted by DL models compared to the ground truth. The marker detection time was fast enough to meet the requirements for real-time application.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Trada2023c,

title = {Impact of tumour region of interest delineation method for mid-treatment FDG-PET response prediction in head and neck squamous cell carcinoma undergoing radiotherapy},

author = {Yuvnik Trada and Peter Lin and Mark T. Lee and Michael G. Jameson and Phillip Chlap and Paul Keall and Daniel Moses and Allan Fowler},

doi = {10.21037/qims-22-798},

issn = {2223-4306},

year  = {2023},

date = {2023-05-00},

journal = {Quant Imaging Med Surg},

volume = {13},

number = {5},

pages = {2822--2836},

publisher = {AME Publishing Company},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Reynolds2023,

title = {Technical note: Extended longitudinal and lateral 3D imaging with a continuous dual‐isocenter CBCT scan},

author = {Tess Reynolds and Sepideh Hatamikia and Yiqun Q. Ma and Owen Dillon and Grace Gang and J. Webster Stayman and Ricky O'Brien},

doi = {10.1002/mp.16234},

issn = {2473-4209},

year  = {2023},

date = {2023-04-00},

journal = {Medical Physics},

volume = {50},

number = {4},

pages = {2372--2379},

publisher = {Wiley},

abstract = {AbstractBackgroundThe clinical benefits of intraoperative cone beam CT (CBCT) during orthopedic procedures include (1) improved accuracy for procedures involving the placement of hardware and (2) providing immediate surgical verification.PurposeOrthopedic interventions often involve long and wide anatomical sites (e.g., lower extremities). Therefore, in order to ensure that the clinical benefits are available to all orthopedic procedures, we investigate the feasibility of a novel imaging trajectory to simultaneously expand the CBCT field‐of‐view longitudinally and laterally.MethodsA continuous dual‐isocenter imaging trajectory was implemented on a clinical robotic CBCT system using additional real‐time control hardware. The trajectory consisted of 200° circular arcs separated by alternating lateral and longitudinal table translations. Due to hardware constraints, the direction of rotation (clockwise/anticlockwise) and lateral table translation (left/right) was reversed every 400°. X‐ray projections were continuously acquired at 15 frames/s throughout all movements. A whole‐body phantom was used to verify the trajectory. As comparator, a series of conventional large volume acquisitions were stitched together. Image quality was quantified using Root Mean Square Deviation (RMSD), Mean Absolute Percentage Deviation (MAPD), Structural Similarity Index Metric (SSIM) and Contrast‐to‐Noise Ratio (CNR).ResultsThe imaging volume produced by the continuous dual‐isocenter trajectory had dimensions of L = 95 cm × W = 45 cm × H = 45 cm. This enabled the hips to the feet of the whole‐body phantom to be captured in approximately half the imaging dose and acquisition time of the 11 stitched conventional acquisitions required to match the longitudinal and lateral imaging dimensions. Compared to the stitched conventional images, the continuous dual‐isocenter acquisition had RMSD of 4.84, MAPD of 6.58% and SSIM of 0.99. The CNR of the continuous dual‐isocenter and stitched conventional acquisitions were 1.998 and 1.999, respectively.ConclusionExtended longitudinal and lateral intraoperative volumetric imaging is feasible on clinical robotic CBCT systems.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Waddington2023,

title = {Real‐time radial reconstruction with domain transform manifold learning for MRI‐guided radiotherapy},

author = {David E. J. Waddington and Nicholas Hindley and Neha Koonjoo and Christopher Chiu and Tess Reynolds and Paul Z. Y. Liu and Bo Zhu and Danyal Bhutto and Chiara Paganelli and Paul J. Keall and Matthew S. Rosen},

doi = {10.1002/mp.16224},

issn = {2473-4209},

year  = {2023},

date = {2023-04-00},

journal = {Medical Physics},

volume = {50},

number = {4},

pages = {1962--1974},

publisher = {Wiley},

abstract = {AbstractBackgroundMRI‐guidance techniques that dynamically adapt radiation beams to follow tumor motion in real time will lead to more accurate cancer treatments and reduced collateral healthy tissue damage. The gold‐standard for reconstruction of undersampled MR data is compressed sensing (CS) which is computationally slow and limits the rate that images can be available for real‐time adaptation.PurposeOnce trained, neural networks can be used to accurately reconstruct raw MRI data with minimal latency. Here, we test the suitability of deep‐learning‐based image reconstruction for real‐time tracking applications on MRI‐Linacs.MethodsWe use automated transform by manifold approximation (AUTOMAP), a generalized framework that maps raw MR signal to the target image domain, to rapidly reconstruct images from undersampled radial k‐space data. The AUTOMAP neural network was trained to reconstruct images from a golden‐angle radial acquisition, a benchmark for motion‐sensitive imaging, on lung cancer patient data and generic images from ImageNet. Model training was subsequently augmented with motion‐encoded k‐space data derived from videos in the YouTube‐8M dataset to encourage motion robust reconstruction.ResultsAUTOMAP models fine‐tuned on retrospectively acquired lung cancer patient data reconstructed radial k‐space with equivalent accuracy to CS but with much shorter processing times. Validation of motion‐trained models with a virtual dynamic lung tumor phantom showed that the generalized motion properties learned from YouTube lead to improved target tracking accuracy.ConclusionAUTOMAP can achieve real‐time, accurate reconstruction of radial data. These findings imply that neural‐network‐based reconstruction is potentially superior to alternative approaches for real‐time image guidance applications.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Yamamoto2023,

title = {Four-Dimensional Computed Tomography Ventilation Image-Guided Lung Functional Avoidance Radiation Therapy: A Single-Arm Prospective Pilot Clinical Trial},

author = {Tokihiro Yamamoto and Sven Kabus and Matthieu Bal and Paul J. Keall and Angel Moran and Cari Wright and Stanley H. Benedict and Devin Holland and Nichole Mahaffey and Lihong Qi and Megan E. Daly},

doi = {10.1016/j.ijrobp.2022.11.026},

issn = {0360-3016},

year  = {2023},

date = {2023-04-00},

journal = {International Journal of Radiation Oncology*Biology*Physics},

volume = {115},

number = {5},

pages = {1144--1154},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Yamamoto2023b,

title = {Four-Dimensional Computed Tomography Ventilation Image-Guided Lung Functional Avoidance Radiation Therapy: A Single-Arm Prospective Pilot Clinical Trial},

author = {Tokihiro Yamamoto and Sven Kabus and Matthieu Bal and Paul J. Keall and Angel Moran and Cari Wright and Stanley H. Benedict and Devin Holland and Nichole Mahaffey and Lihong Qi and Megan E. Daly},

doi = {10.1016/j.ijrobp.2022.11.026},

issn = {0360-3016},

year  = {2023},

date = {2023-04-00},

journal = {International Journal of Radiation Oncology*Biology*Physics},

volume = {115},

number = {5},

pages = {1144--1154},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Blake2023,

title = {Thoracic motion‐compensated cone‐beam computed tomography in under 20 seconds on a fast‐rotating linac: A simulation study},

author = {Samuel J. Blake and Owen Dillon and Hilary L. Byrne and Ricky T. O'Brien},

doi = {10.1002/acm2.13909},

issn = {1526-9914},

year  = {2023},

date = {2023-03-00},

journal = {J Applied Clin Med Phys},

volume = {24},

number = {3},

publisher = {Wiley},

abstract = {AbstractBackgroundRapid kV cone‐beam computed tomography (CBCT) scans are achievable in under 20 s on select linear accelerator systems to generate volumetric images in three dimensions (3D). Daily pre‐treatment four‐dimensional CBCT (4DCBCT) is recommended in image‐guided lung radiotherapy to mitigate the detrimental effects of respiratory motion on treatment quality.PurposeTo demonstrate the potential for thoracic 4DCBCT reconstruction using projection data that was simulated using a clinical rapid 3DCBCT acquisition protocol.MethodsWe simulated conventional (1320 projections over 4 min) and rapid (491 projections over 16.6 s) CBCT acquisitions using 4D computed tomography (CT) volumes of 14 lung cancer patients. Conventional acquisition data were reconstructed using the 4D Feldkamp‐Davis‐Kress (FDK) algorithm. Rapid acquisition data were reconstructed using 3DFDK, 4DFDK, and Motion‐Compensated FDK (MCFDK). Image quality was evaluated using Contrast‐to‐Noise Ratio (CNR), Tissue Interface Width (TIW), Root‐Mean‐Square Error (RMSE), and Structural SIMilarity (SSIM).ResultsThe conventional acquisition 4DFDK reconstructions had median phase averaged CNR, TIW, RMSE, and SSIM of 2.96, 8.02 mm, 83.5, and 0.54, respectively. The rapid acquisition 3DFDK reconstructions had median CNR, TIW, RMSE, and SSIM of 2.99, 13.6 mm, 112, and 0.44 respectively. The rapid acquisition MCFDK reconstructions had median phase averaged CNR, TIW, RMSE, and SSIM of 2.98, 10.2 mm, 103, and 0.46, respectively. Rapid acquisition 4DFDK reconstruction quality was insufficient for any practical use due to sparse angular projection sampling.ConclusionsResults suggest that 4D motion‐compensated reconstruction of rapid acquisition thoracic CBCT data are feasible with image quality approaching conventional acquisition CBCT data reconstructed using standard 4DFDK.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Whelan2022,

title = {TopasOpt: An open‐source library for optimization with Topas Monte Carlo},

author = {Brendan Whelan and Billy W. Loo and Jinghui Wang and Paul Keall},

doi = {10.1002/mp.16126},

issn = {2473-4209},

year  = {2023},

date = {2023-02-00},

journal = {Medical Physics},

volume = {50},

number = {2},

pages = {1121--1131},

publisher = {Wiley},

abstract = {AbstractPurposeTo describe and test TopasOpt: a free, open‐source and extensible library for performing mathematical optimization of Monte Carlo simulations in Topas.MethodsTopasOpt enables any Topas model to be transformed into an optimization problem, and any parameter within the model to be treated as an optimization variable. Three case studies are presented. The starting model consists of a 10 MeV electron beam striking a tungsten target. The resulting bremsstrahlung X‐ray spectrum is collimated by a primary and secondary collimator before being scored in a water tank. In the first case study (electron phase space optimization), five parameters describing the electron beam were treated as optimization variables and assigned a random starting value. An objective function was defined based on differences of depth‐dose and profiles in water between the original (ground truth) model and a given model generated by TopasOpt. The problem was solved using Bayesian Optimization and the Nelder‐Mead method. One hundred iterations were run in each case. In the second case study, (collimator geometry optimization), this process was repeated, but three geometric parameters defining the secondary collimator were treated as optimization variables and assigned random starting values, and forty iterations were run. In the third case study, the optimization was repeated with different number of primary particles to study the effect of noise on convergence.ResultsFor case 1 (phase space optimization), both optimization algorithms successfully minimized the objective function, with absolute mean differences in profile dose of 0.4% (Bayesian) and 0.3% (Nelder‐Mead) and 0.2% in depth‐dose for both algorithms. The beam energy was recovered to within 1%, however some parameters had relative errors of up to 171% – a result consistent with the known X‐ray dose is insensitivity to many electron beam parameters. For case 2 (geometry optimization), absolute mean differences in profile dose were 0.6% (Bayesian) and 0.9% (Nelder‐Mead), and 0.5% and 0.9% in depth‐dose. The maximum percentage error in any parameter was 9% with Bayesian Optimization and 28% with Nelder‐Mead. Finally, the Bayesian Optimization algorithm was demonstrated to be robust to moderate levels of noise; when the standard deviation of the objective function was 16% of the mean, the maximum error in any parameter value was 16%, and the absolute mean difference in dose was 0.9% (profile) and 0.8% (depth‐dose).ConclusionsAn open‐source library for optimization with Topas Monte Carlo has been developed, tested, and released. This tool will improve accuracy and efficiency in any situation in which the optimal value of a parameter in a Monte Carlo simulation is unknown. Applications for this tool include (1) The design of new components (2) Reverse engineering of models based on limited experimental or published data, and (3) Tuning of Monte Carlo “hyper parameters” such as variance reduction, physics settings, or scoring parameters.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Whelan2022b,

title = {TopasOpt: An open‐source library for optimization with Topas Monte Carlo},

author = {Brendan Whelan and Billy W. Loo and Jinghui Wang and Paul Keall},

doi = {10.1002/mp.16126},

issn = {2473-4209},

year  = {2023},

date = {2023-02-00},

journal = {Medical Physics},

volume = {50},

number = {2},

pages = {1121--1131},

publisher = {Wiley},

abstract = {AbstractPurposeTo describe and test TopasOpt: a free, open‐source and extensible library for performing mathematical optimization of Monte Carlo simulations in Topas.MethodsTopasOpt enables any Topas model to be transformed into an optimization problem, and any parameter within the model to be treated as an optimization variable. Three case studies are presented. The starting model consists of a 10 MeV electron beam striking a tungsten target. The resulting bremsstrahlung X‐ray spectrum is collimated by a primary and secondary collimator before being scored in a water tank. In the first case study (electron phase space optimization), five parameters describing the electron beam were treated as optimization variables and assigned a random starting value. An objective function was defined based on differences of depth‐dose and profiles in water between the original (ground truth) model and a given model generated by TopasOpt. The problem was solved using Bayesian Optimization and the Nelder‐Mead method. One hundred iterations were run in each case. In the second case study, (collimator geometry optimization), this process was repeated, but three geometric parameters defining the secondary collimator were treated as optimization variables and assigned random starting values, and forty iterations were run. In the third case study, the optimization was repeated with different number of primary particles to study the effect of noise on convergence.ResultsFor case 1 (phase space optimization), both optimization algorithms successfully minimized the objective function, with absolute mean differences in profile dose of 0.4% (Bayesian) and 0.3% (Nelder‐Mead) and 0.2% in depth‐dose for both algorithms. The beam energy was recovered to within 1%, however some parameters had relative errors of up to 171% – a result consistent with the known X‐ray dose is insensitivity to many electron beam parameters. For case 2 (geometry optimization), absolute mean differences in profile dose were 0.6% (Bayesian) and 0.9% (Nelder‐Mead), and 0.5% and 0.9% in depth‐dose. The maximum percentage error in any parameter was 9% with Bayesian Optimization and 28% with Nelder‐Mead. Finally, the Bayesian Optimization algorithm was demonstrated to be robust to moderate levels of noise; when the standard deviation of the objective function was 16% of the mean, the maximum error in any parameter value was 16%, and the absolute mean difference in dose was 0.9% (profile) and 0.8% (depth‐dose).ConclusionsAn open‐source library for optimization with Topas Monte Carlo has been developed, tested, and released. This tool will improve accuracy and efficiency in any situation in which the optimal value of a parameter in a Monte Carlo simulation is unknown. Applications for this tool include (1) The design of new components (2) Reverse engineering of models based on limited experimental or published data, and (3) Tuning of Monte Carlo “hyper parameters” such as variance reduction, physics settings, or scoring parameters.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Carr2023,

title = {A method for beam’s eye view breath-hold monitoring during breast volumetric modulated arc therapy},

author = {Michael A. Carr and Maegan Gargett and Cameron Stanton and Benjamin Zwan and Hilary L. Byrne and Jeremy T. Booth},

doi = {10.1016/j.phro.2023.100419},

issn = {2405-6316},

year  = {2023},

date = {2023-01-00},

journal = {Physics and Imaging in Radiation Oncology},

volume = {25},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Liu2023,

title = {Rapid distortion correction enables accurate magnetic resonance imaging-guided real-time adaptive radiotherapy},

author = {Paul Z. Y Liu and Shanshan Shan and David Waddington and Brendan Whelan and Bin Dong and Gary Liney and Paul Keall},

doi = {10.1016/j.phro.2023.100414},

issn = {2405-6316},

year  = {2023},

date = {2023-01-00},

journal = {Physics and Imaging in Radiation Oncology},

volume = {25},

publisher = {Elsevier BV},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Sengupta2022,

title = {The dosimetric error due to uncorrected tumor rotation during real‐time adaptive prostate stereotactic body radiation therapy},

author = {Chandrima Sengupta and Simon Skouboe and Thomas Ravkilde and Per Rugaard Poulsen and Doan Trang Nguyen and Peter B. Greer and Trevor Moodie and Nicholas Hardcastle and Amy J. Hayden and Sandra Turner and Shankar Siva and Keen‐Hun Tai and Jarad Martin and Jeremy T. Booth and Ricky O'Brien and Paul J. Keall},

doi = {10.1002/mp.16094},

issn = {2473-4209},

year  = {2023},

date = {2023-01-00},

journal = {Medical Physics},

volume = {50},

number = {1},

pages = {20--29},

publisher = {Wiley},

abstract = {AbstractBackgroundDuring prostate stereotactic body radiation therapy (SBRT), prostate tumor translational motion may deteriorate the planned dose distribution. Most of the major advances in motion management to date have focused on correcting this one aspect of the tumor motion, translation. However, large prostate rotation up to 30° has been measured. As the technological innovation evolves toward delivering increasingly precise radiotherapy, it is important to quantify the clinical benefit of translational and rotational motion correction over translational motion correction alone.PurposeThe purpose of this work was to quantify the dosimetric impact of intrafractional dynamic rotation of the prostate measured with a six degrees‐of‐freedom tumor motion monitoring technology.MethodsThe delivered dose was reconstructed including (a) translational and rotational motion and (b) only translational motion of the tumor for 32 prostate cancer patients recruited on a 5‐fraction prostate SBRT clinical trial. Patients on the trial received 7.25 Gy in a treatment fraction. A 5 mm clinical target volume (CTV) to planning target volume (PTV) margin was applied in all directions except the posterior direction where a 3 mm expansion was used. Prostate intrafractional translational motion was managed using a gating strategy, and any translation above the gating threshold was corrected by applying an equivalent couch shift. The residual translational motion is denoted as . Prostate intrafractional rotational motion  was recorded but not corrected. The dose differences from the planned dose due to  + , ΔD( + ) and due to  alone, ΔD(), were then determined for CTV D98, PTV D95, bladder V6Gy, and rectum V6Gy. The residual dose error due to uncorrected rotation,  was then quantified:  = ΔD( + ) ‐ ΔD().ResultsFractional data analysis shows that the dose differences from the plan (both ΔD( + ) and ΔD()) for CTV D98 was less than 5% in all treatment fractions. ΔD( + ) was larger than 5% in one fraction for PTV D95, in one fraction for bladder V6Gy, and in five fractions for rectum V6Gy. Uncorrected rotation,  induced residual dose error, , resulted in less dose to CTV and PTV in 43% and 59% treatment fractions, respectively, and more dose to bladder and rectum in 51% and 53% treatment fractions, respectively. The cumulative dose over five fractions, ∑D( + ) and ∑D(), was always within 5% of the planned dose for all four structures for every patient.ConclusionsThe dosimetric impact of tumor rotation on a large prostate cancer patient cohort was quantified in this study. These results suggest that the standard 3–5 mm CTV‐PTV margin was sufficient to account for the intrafraction prostate rotation observed for this cohort of patients, provided an appropriate gating threshold was applied to correct for translational motion. Residual dose errors due to uncorrected prostate rotation were small in magnitude, which may be corrected using different treatment adaptation strategies to further improve the dosimetric accuracy.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Sengupta2022b,

title = {The dosimetric error due to uncorrected tumor rotation during real‐time adaptive prostate stereotactic body radiation therapy},

author = {Chandrima Sengupta and Simon Skouboe and Thomas Ravkilde and Per Rugaard Poulsen and Doan Trang Nguyen and Peter B. Greer and Trevor Moodie and Nicholas Hardcastle and Amy J. Hayden and Sandra Turner and Shankar Siva and Keen‐Hun Tai and Jarad Martin and Jeremy T. Booth and Ricky O'Brien and Paul J. Keall},

doi = {10.1002/mp.16094},

issn = {2473-4209},

year  = {2023},

date = {2023-01-00},

journal = {Medical Physics},

volume = {50},

number = {1},

pages = {20--29},

publisher = {Wiley},

abstract = {AbstractBackgroundDuring prostate stereotactic body radiation therapy (SBRT), prostate tumor translational motion may deteriorate the planned dose distribution. Most of the major advances in motion management to date have focused on correcting this one aspect of the tumor motion, translation. However, large prostate rotation up to 30° has been measured. As the technological innovation evolves toward delivering increasingly precise radiotherapy, it is important to quantify the clinical benefit of translational and rotational motion correction over translational motion correction alone.PurposeThe purpose of this work was to quantify the dosimetric impact of intrafractional dynamic rotation of the prostate measured with a six degrees‐of‐freedom tumor motion monitoring technology.MethodsThe delivered dose was reconstructed including (a) translational and rotational motion and (b) only translational motion of the tumor for 32 prostate cancer patients recruited on a 5‐fraction prostate SBRT clinical trial. Patients on the trial received 7.25 Gy in a treatment fraction. A 5 mm clinical target volume (CTV) to planning target volume (PTV) margin was applied in all directions except the posterior direction where a 3 mm expansion was used. Prostate intrafractional translational motion was managed using a gating strategy, and any translation above the gating threshold was corrected by applying an equivalent couch shift. The residual translational motion is denoted as . Prostate intrafractional rotational motion  was recorded but not corrected. The dose differences from the planned dose due to  + , ΔD( + ) and due to  alone, ΔD(), were then determined for CTV D98, PTV D95, bladder V6Gy, and rectum V6Gy. The residual dose error due to uncorrected rotation,  was then quantified:  = ΔD( + ) ‐ ΔD().ResultsFractional data analysis shows that the dose differences from the plan (both ΔD( + ) and ΔD()) for CTV D98 was less than 5% in all treatment fractions. ΔD( + ) was larger than 5% in one fraction for PTV D95, in one fraction for bladder V6Gy, and in five fractions for rectum V6Gy. Uncorrected rotation,  induced residual dose error, , resulted in less dose to CTV and PTV in 43% and 59% treatment fractions, respectively, and more dose to bladder and rectum in 51% and 53% treatment fractions, respectively. The cumulative dose over five fractions, ∑D( + ) and ∑D(), was always within 5% of the planned dose for all four structures for every patient.ConclusionsThe dosimetric impact of tumor rotation on a large prostate cancer patient cohort was quantified in this study. These results suggest that the standard 3–5 mm CTV‐PTV margin was sufficient to account for the intrafraction prostate rotation observed for this cohort of patients, provided an appropriate gating threshold was applied to correct for translational motion. Residual dose errors due to uncorrected prostate rotation were small in magnitude, which may be corrected using different treatment adaptation strategies to further improve the dosimetric accuracy.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}