The linear-quadratic model is one of the key tools in radiation biology and physics. It provides a simple relationship between cell survival and delivered dose: , and has been used extensively to analyse and predict responses to ionising radiation both in vitro and in vivo. Despite its ubiquity, there remain questions about its interpretation and wider applicability—Is it a convenient empirical fit or representative of some deeper mechanistic behaviour? Does a model of single-cell survival in vitro really correspond to clinical tissue responses? Is it applicable at very high and very low doses? Here, we review these issues, discussing current usage of the LQ model, its historical context, what we now know about its mechanistic underpinnings, and the potential challenges and confounding factors that arise when trying to apply it across a range of systems.
The aim of the Institute of Physics and Engineering in Medicine (IPEM) is to promote the advancement of physics and engineering applied to medicine and biology for the public benefit. Its members are professionals working in healthcare, education, industry and research.
IPEM publishes scientific journals and books and organises conferences to disseminate knowledge and support members in their development. It sets and advises on standards for the practice, education and training of scientists and engineers working in healthcare to secure an effective and appropriate workforce.
ISSN: 1361-6560
The international journal of biomedical physics and engineering, published by IOP Publishing on behalf of the Institute of Physics and Engineering in Medicine (IPEM).
Browse IPEM-IOP ebooks Series in Physics and Engineering in Medicine and Biology.
Open all abstracts, in this tab
Stephen Joseph McMahon 2019 Phys. Med. Biol. 64 01TR01
Wayne D Newhauser and Rui Zhang 2015 Phys. Med. Biol. 60 R155
The physics of proton therapy has advanced considerably since it was proposed in 1946. Today analytical equations and numerical simulation methods are available to predict and characterize many aspects of proton therapy. This article reviews the basic aspects of the physics of proton therapy, including proton interaction mechanisms, proton transport calculations, the determination of dose from therapeutic and stray radiations, and shielding design. The article discusses underlying processes as well as selected practical experimental and theoretical methods. We conclude by briefly speculating on possible future areas of research of relevance to the physics of proton therapy.
Conor K McGarry et al 2020 Phys. Med. Biol. 65 23TR01
Tissue mimicking materials (TMMs), typically contained within phantoms, have been used for many decades in both imaging and therapeutic applications. This review investigates the specifications that are typically being used in development of the latest TMMs. The imaging modalities that have been investigated focus around CT, mammography, SPECT, PET, MRI and ultrasound. Therapeutic applications discussed within the review include radiotherapy, thermal therapy and surgical applications. A number of modalities were not reviewed including optical spectroscopy, optical imaging and planar x-rays. The emergence of image guided interventions and multimodality imaging have placed an increasing demand on the number of specifications on the latest TMMs. Material specification standards are available in some imaging areas such as ultrasound. It is recommended that this should be replicated for other imaging and therapeutic modalities. Materials used within phantoms have been reviewed for a series of imaging and therapeutic applications with the potential to become a testbed for cross-fertilization of materials across modalities. Deformation, texture, multimodality imaging and perfusion are common themes that are currently under development.
Shaoyan Pan et al 2023 Phys. Med. Biol. 68 105004
Objective. Artificial intelligence (AI) methods have gained popularity in medical imaging research. The size and scope of the training image datasets needed for successful AI model deployment does not always have the desired scale. In this paper, we introduce a medical image synthesis framework aimed at addressing the challenge of limited training datasets for AI models. Approach. The proposed 2D image synthesis framework is based on a diffusion model using a Swin-transformer-based network. This model consists of a forward Gaussian noise process and a reverse process using the transformer-based diffusion model for denoising. Training data includes four image datasets: chest x-rays, heart MRI, pelvic CT, and abdomen CT. We evaluated the authenticity, quality, and diversity of the synthetic images using visual Turing assessments conducted by three medical physicists, and four quantitative evaluations: the Inception score (IS), Fréchet Inception Distance score (FID), feature similarity and diversity score (DS, indicating diversity similarity) between the synthetic and true images. To leverage the framework value for training AI models, we conducted COVID-19 classification tasks using real images, synthetic images, and mixtures of both images. Main results. Visual Turing assessments showed an average accuracy of 0.64 (accuracy converging to indicates a better realistic visual appearance of the synthetic images), sensitivity of 0.79, and specificity of 0.50. Average quantitative accuracy obtained from all datasets were IS = 2.28, FID = 37.27, FDS = 0.20, and DS = 0.86. For the COVID-19 classification task, the baseline network obtained an accuracy of 0.88 using a pure real dataset, 0.89 using a pure synthetic dataset, and 0.93 using a dataset mixed of real and synthetic data. Significance. A image synthesis framework was demonstrated for medical image synthesis, which can generate high-quality medical images of different imaging modalities with the purpose of supplementing existing training sets for AI model deployment. This method has potential applications in many data-driven medical imaging research.
Mats Danielsson et al 2021 Phys. Med. Biol. 66 03TR01
The introduction of photon-counting detectors is expected to be the next major breakthrough in clinical x-ray computed tomography (CT). During the last decade, there has been considerable research activity in the field of photon-counting CT, in terms of both hardware development and theoretical understanding of the factors affecting image quality. In this article, we review the recent progress in this field with the intent of highlighting the relationship between detector design considerations and the resulting image quality. We discuss detector design choices such as converter material, pixel size, and readout electronics design, and then elucidate their impact on detector performance in terms of dose efficiency, spatial resolution, and energy resolution. Furthermore, we give an overview of data processing, reconstruction methods and metrics of imaging performance; outline clinical applications; and discuss potential future developments.
Stefan Gundacker and Arjan Heering 2020 Phys. Med. Biol. 65 17TR01
The silicon photomultiplier (SiPM) is an established device of choice for a variety of applications, e.g. in time of flight positron emission tomography (TOF-PET), lifetime fluorescence spectroscopy, distance measurements in LIDAR applications, astrophysics, quantum-cryptography and related applications as well as in high energy physics (HEP).
To fully utilize the exceptional performances of the SiPM, in particular its sensitivity down to single photon detection, the dynamic range and its intrinsically fast timing properties, a qualitative description and understanding of the main SiPM parameters and properties is necessary. These analyses consider the structure and the electrical model of a single photon avalanche diode (SPAD) and the integration in an array of SPADs, i.e. the SiPM. The discussion will include the front-end readout and the comparison between analog-SiPMs, where the array of SPADs is connected in parallel, and the digital SiPM, where each SPAD is read out and digitized by its own electronic channel.
For several applications a further complete phenomenological view on SiPMs is necessary, defining several SiPM intrinsic parameters, i.e. gain fluctuation, afterpulsing, excess noise, dark count rate, prompt and delayed optical crosstalk, single photon time resolution (SPTR), photon detection effieciency (PDE) etc. These qualities of SiPMs influence directly and indirectly the time and energy resolution, for example in PET and HEP. This complete overview of all parameters allows one to draw solid conclusions on how best performances can be achieved for the various needs of the different applications.
Luca Brombal et al 2024 Phys. Med. Biol. 69 075027
Following the rapid, but independent, diffusion of x-ray spectral and phase-contrast systems, this work demonstrates the first combination of spectral and phase-contrast computed tomography (CT) obtained by using the edge-illumination technique and a CdTe small-pixel (62 μm) spectral detector. A theoretical model is introduced, starting from a standard attenuation-based spectral decomposition and leading to spectral phase-contrast material decomposition. Each step of the model is followed by quantification of accuracy and sensitivity on experimental data of a test phantom containing different solutions with known concentrations. An example of a micro CT application (20 μm voxel size) on an iodine-perfused ex vivo murine model is reported. The work demonstrates that spectral-phase contrast combines the advantages of spectral imaging, i.e. high-Z material discrimination capability, and phase-contrast imaging, i.e. soft tissue sensitivity, yielding simultaneously mass density maps of water, calcium, and iodine with an accuracy of 1.1%, 3.5%, and 1.9% (root mean square errors), respectively. Results also show a 9-fold increase in the signal-to-noise ratio of the water channel when compared to standard spectral decomposition. The application to the murine model revealed the potential of the technique in the simultaneous 3D visualization of soft tissue, bone, and vasculature. While being implemented by using a broad spectrum (pink beam) at a synchrotron radiation facility (Elettra, Trieste, Italy), the proposed experimental setup can be readily translated to compact laboratory systems including conventional x-ray tubes.
Lu Liu et al 2021 Phys. Med. Biol. 66 11TR01
Deep learning (DL) has become widely used for medical image segmentation in recent years. However, despite these advances, there are still problems for which DL-based segmentation fails. Recently, some DL approaches had a breakthrough by using anatomical information which is the crucial cue for manual segmentation. In this paper, we provide a review of anatomy-aided DL for medical image segmentation which covers systematically summarized anatomical information categories and corresponding representation methods. We address known and potentially solvable challenges in anatomy-aided DL and present a categorized methodology overview on using anatomical information with DL from over 70 papers. Finally, we discuss the strengths and limitations of the current anatomy-aided DL approaches and suggest potential future work.
Lena Nenoff et al 2023 Phys. Med. Biol. 68 24TR01
Deformable image registration (DIR) is a versatile tool used in many applications in radiotherapy (RT). DIR algorithms have been implemented in many commercial treatment planning systems providing accessible and easy-to-use solutions. However, the geometric uncertainty of DIR can be large and difficult to quantify, resulting in barriers to clinical practice. Currently, there is no agreement in the RT community on how to quantify these uncertainties and determine thresholds that distinguish a good DIR result from a poor one. This review summarises the current literature on sources of DIR uncertainties and their impact on RT applications. Recommendations are provided on how to handle these uncertainties for patient-specific use, commissioning, and research. Recommendations are also provided for developers and vendors to help users to understand DIR uncertainties and make the application of DIR in RT safer and more reliable.
Amirhossein Sanaat et al 2022 Phys. Med. Biol. 67 155021
Organ-specific PET scanners have been developed to provide both high spatial resolution and sensitivity, although the deployment of several dedicated PET scanners at the same center is costly and space-consuming. Active-PET is a multifunctional PET scanner design exploiting the advantages of two different types of detector modules and mechanical arms mechanisms enabling repositioning of the detectors to allow the implementation of different geometries/configurations. Active-PET can be used for different applications, including brain, axilla, breast, prostate, whole-body, preclinical and pediatrics imaging, cell tracking, and image guidance for therapy. Monte Carlo techniques were used to simulate a PET scanner with two sets of high resolution and high sensitivity pixelated Lutetium Oxyorthoscilicate (LSO(Ce)) detector blocks (24 for each group, overall 48 detector modules for each ring), one with large pixel size (4 × 4 mm2) and crystal thickness (20 mm), and another one with small pixel size (2 × 2 mm2) and thickness (10 mm). Each row of detector modules is connected to a linear motor that can displace the detectors forward and backward along the radial axis to achieve variable gantry diameter in order to image the target subject at the optimal/desired resolution and/or sensitivity. At the center of the field-of-view, the highest sensitivity (15.98 kcps MBq−1) was achieved by the scanner with a small gantry and high-sensitivity detectors while the best spatial resolution was obtained by the scanner with a small gantry and high-resolution detectors (2.2 mm, 2.3 mm, 2.5 mm FWHM for tangential, radial, and axial, respectively). The configuration with large-bore (combination of high-resolution and high-sensitivity detectors) achieved better performance and provided higher image quality compared to the Biograph mCT as reflected by the 3D Hoffman brain phantom simulation study. We introduced the concept of a non-static PET scanner capable of switching between large and small field-of-view as well as high-resolution and high-sensitivity imaging.
Open all abstracts, in this tab
Didier Lustermans et al 2024 Phys. Med. Biol. 69 105018
Objective. Newer cone-beam computed tomography (CBCT) imaging systems offer reconstruction algorithms including metal artifact reduction (MAR) and extended field-of-view (eFoV) techniques to improve image quality. In this study a new CBCT imager, the new Varian HyperSight CBCT, is compared to fan-beam CT and two CBCT imagers installed in a ring-gantry and C-arm linear accelerator, respectively. Approach. The image quality was assessed for HyperSight CBCT which uses new hardware, including a large-size flat panel detector, and improved image reconstruction algorithms. The decrease of metal artifacts was quantified (structural similarity index measure (SSIM) and root-mean-squared error (RMSE)) when applying MAR reconstruction and iterative reconstruction for a dental and spine region using a head-and-neck phantom. The geometry and CT number accuracy of the eFoV reconstruction was evaluated outside the standard field-of-view (sFoV) on a large 3D-printed chest phantom. Phantom size dependency of CT numbers was evaluated on three cylindrical phantoms of increasing diameter. Signal-to-noise and contrast-to-noise were quantified on an abdominal phantom. Main results. In phantoms with streak artifacts, MAR showed comparable results for HyperSight CBCT and CT, with MAR increasing the SSIM (0.97–0.99) and decreasing the RMSE (62–55 HU) compared to iterative reconstruction without MAR. In addition, HyperSight CBCT showed better geometrical accuracy in the eFoV than CT (Jaccard Conformity Index increase of 0.02–0.03). However, the CT number accuracy outside the sFoV was lower than for CT. The maximum CT number variation between different phantom sizes was lower for the HyperSight CBCT imager (∼100 HU) compared to the two other CBCT imagers (∼200 HU), but not fully comparable to CT (∼50 HU). Significance. This study demonstrated the imaging performance of the new HyperSight CBCT imager and the potential of applying this CBCT system in more advanced scenarios by comparing the quality against fan-beam CT.
Amir Reza Zekavat et al 2024 Phys. Med. Biol. 69 105017
Objective. To report on a micro computed tomography (micro-CT) system capable of x-ray phase contrast imaging and of increasing spatial resolution at constant magnification. Approach. The micro-CT system implements the edge illumination (EI) method, which relies on two absorbing masks with periodically spaced transmitting apertures in the beam path; these split the beam into an array of beamlets and provide sensitivity to the beamlets' directionality, i.e. refraction. In EI, spatial resolution depends on the width of the beamlets rather than on the source/detector point spread function (PSF), meaning that resolution can be increased by decreasing the mask apertures, without changing the source/detector PSF or the magnification. Main results. We have designed a dedicated mask featuring multiple bands with differently sized apertures and used this to demonstrate that resolution is a tuneable parameter in our system, by showing that increasingly small apertures deliver increasingly detailed images. Phase contrast images of a bar pattern-based resolution phantom and a biological sample (a mouse embryo) were obtained at multiple resolutions. Significance. The new micro-CT system could find application in areas where phase contrast is already known to provide superior image quality, while the added tuneable resolution functionality could enable more sophisticated analyses in these applications, e.g. by scanning samples at multiple scales.
Charles Pageot et al 2024 Phys. Med. Biol. 69 105016
Objective. Beam current transformers (BCT) are promising detectors for real-time beam monitoring in ultra-high dose rate (UHDR) electron radiotherapy. However, previous studies have reported a significant sensitivity of the BCT signal to changes in source-to-surface distance (SSD), field size, and phantom material which have until now been attributed to the fluctuating levels of electrons backscattered within the BCT. The purpose of this study is to evaluate this hypothesis, with the goal of understanding and mitigating the variations in BCT signal due to changes in irradiation conditions. Approach. Monte Carlo simulations and experimental measurements were conducted with a UHDR-capable intra-operative electron linear accelerator to analyze the impact of backscattered electrons on BCT signal. The potential influence of charge accumulation in media as a mechanism affecting BCT signal perturbation was further investigated by examining the effects of phantom conductivity and electrical grounding. Finally, the effectiveness of Faraday shielding to mitigate BCT signal variations is evaluated. Main Results. Monte Carlo simulations indicated that the fraction of electrons backscattered in water and on the collimator plastic at 6 and 9 MeV is lower than 1%, suggesting that backscattered electrons alone cannot account for the observed BCT signal variations. However, our experimental measurements confirmed previous findings of BCT response variation up to 15% for different field diameters. A significant impact of phantom type on BCT response was also observed, with variations in BCT signal as high as 14.1% when comparing measurements in water and solid water. The introduction of a Faraday shield to our applicators effectively mitigated the dependencies of BCT signal on SSD, field size, and phantom material. Significance. Our results indicate that variations in BCT signal as a function of SSD, field size, and phantom material are likely driven by an electric field originating in dielectric materials exposed to the UHDR electron beam. Strategies such as Faraday shielding were shown to effectively prevent these electric fields from affecting BCT signal, enabling reliable BCT-based electron UHDR beam monitoring.
Mingzhe Hu et al 2024 Phys. Med. Biol. 69 10TR01
This review paper aims to serve as a comprehensive guide and instructional resource for researchers seeking to effectively implement language models in medical imaging research. First, we presented the fundamental principles and evolution of language models, dedicating particular attention to large language models. We then reviewed the current literature on how language models are being used to improve medical imaging, emphasizing a range of applications such as image captioning, report generation, report classification, findings extraction, visual question response systems, interpretable diagnosis and so on. Notably, the capabilities of ChatGPT were spotlighted for researchers to explore its further applications. Furthermore, we covered the advantageous impacts of accurate and efficient language models in medical imaging analysis, such as the enhancement of clinical workflow efficiency, reduction of diagnostic errors, and assistance of clinicians in providing timely and accurate diagnoses. Overall, our goal is to have better integration of language models with medical imaging, thereby inspiring new ideas and innovations. It is our aspiration that this review can serve as a useful resource for researchers in this field, stimulating continued investigative and innovative pursuits of the application of language models in medical imaging.
Taku Inaniwa et al 2024 Phys. Med. Biol. 69 105015
Objective. To investigate the effect of redistribution and reoxygenation on the 3-year tumor control probability (TCP) of patients with stage I non-small cell lung cancer (NSCLC) treated with carbon-ion radiotherapy. Approach. A meta-analysis of published clinical data of 233 NSCLC patients treated by carbon-ion radiotherapy under 18-, 9-, 4-, and single-fraction schedules was conducted. The linear-quadratic (LQ)-based cell-survival model incorporating the radiobiological 5Rs, radiosensitivity, repopulation, repair, redistribution, and reoxygenation, was developed to reproduce the clinical TCP data. Redistribution and reoxygenation were regarded together as a single phenomenon and termed 'resensitization' in the model. The optimum interval time between fractions was investigated for each fraction schedule using the determined model parameters. Main results. The clinical TCP data for 18-, 9-, and 4-fraction schedules were reasonably reproduced by the model without the resensitization effect, whereas its incorporation was essential to reproduce the TCP data for all fraction schedules including the single fraction. The curative dose for the single-fraction schedule was estimated to be 49.0 Gy (RBE), which corresponds to the clinically adopted dose prescription of 50.0 Gy (RBE). For 18-, 9-, and 4-fraction schedules, a 2-to-3-day interval is required to maximize the resensitization effect during the time interval. In contrast, the single-fraction schedule cannot benefit from the resensitization effect, and the shorter treatment time is preferable to reduce the effect of sub-lethal damage repair during the treatment. Significance. The LQ-based cell-survival model incorporating the radiobiological 5Rs was developed and used to evaluate the effect of the resensitization on clinical results of NSCLC patients treated with hypo-fractionated carbon-ion radiotherapy. The incorporation of the resensitization into the cell-survival model improves the reproducibility to the clinical TCP data. A shorter treatment time is preferable in the single-fraction schedule, while a 2-to-3-day interval between fractions is preferable in the multi-fraction schedules for effective treatments.
Open all abstracts, in this tab
Mingzhe Hu et al 2024 Phys. Med. Biol. 69 10TR01
This review paper aims to serve as a comprehensive guide and instructional resource for researchers seeking to effectively implement language models in medical imaging research. First, we presented the fundamental principles and evolution of language models, dedicating particular attention to large language models. We then reviewed the current literature on how language models are being used to improve medical imaging, emphasizing a range of applications such as image captioning, report generation, report classification, findings extraction, visual question response systems, interpretable diagnosis and so on. Notably, the capabilities of ChatGPT were spotlighted for researchers to explore its further applications. Furthermore, we covered the advantageous impacts of accurate and efficient language models in medical imaging analysis, such as the enhancement of clinical workflow efficiency, reduction of diagnostic errors, and assistance of clinicians in providing timely and accurate diagnoses. Overall, our goal is to have better integration of language models with medical imaging, thereby inspiring new ideas and innovations. It is our aspiration that this review can serve as a useful resource for researchers in this field, stimulating continued investigative and innovative pursuits of the application of language models in medical imaging.
Christian P Karger et al 2024 Phys. Med. Biol. 69 06TR01
Modern radiotherapy delivers highly conformal dose distributions to irregularly shaped target volumes while sparing the surrounding normal tissue. Due to the complex planning and delivery techniques, dose verification and validation of the whole treatment workflow by end-to-end tests became much more important and polymer gel dosimeters are one of the few possibilities to capture the delivered dose distribution in 3D. The basic principles and formulations of gel dosimetry and its evaluation methods are described and the available studies validating device-specific geometrical parameters as well as the dose delivery by advanced radiotherapy techniques, such as 3D-CRT/IMRT and stereotactic radiosurgery treatments, the treatment of moving targets, online-adaptive magnetic resonance-guided radiotherapy as well as proton and ion beam treatments, are reviewed. The present status and limitations as well as future challenges of polymer gel dosimetry for the validation of complex radiotherapy techniques are discussed.
Dong Sik Kim 2024 Phys. Med. Biol. 69 03TR01
Objective. The noise characteristics of digital x-ray imaging devices are determined by contributions such as photon noise, electronic noise, and fixed pattern noise, and can be evaluated from measuring the noise power spectrum (NPS), which is the power spectral density of the noise. Hence, accurately measuring NPS is important in developing detectors for acquiring low-noise digital x-ray images. To make accurate measurements, it is necessary to understand NPS, identify problems that may arise, and know how to process the obtained x-ray images. Approach. The primitive concept of NPS is first introduced with a periodogram-based estimate and its bias and variance are discussed. In measuring NPS based on the IEC62220 standards, various issues, such as the fixed pattern noise, high-precision estimates, and lag corrections, are summarized with simulation examples. Main results. High-precision estimates can be provided for an appropriate number of samples extracted from x-ray images while compromising spectral resolution. Depending on medical imaging systems, by eliminating the influence of fixed pattern noise, NPS, which represents only photon and electronic noise, can be efficiently measured. For NPS measurements in dynamic detectors, an appropriate lag correction technique can be selected depending on the emitted x-rays and image acquisition process. Significance. Various issues in measuring NPS are reviewed and summarized for accurately evaluating the noise performance of digital x-ray imaging devices.
Lena Nenoff et al 2023 Phys. Med. Biol. 68 24TR01
Deformable image registration (DIR) is a versatile tool used in many applications in radiotherapy (RT). DIR algorithms have been implemented in many commercial treatment planning systems providing accessible and easy-to-use solutions. However, the geometric uncertainty of DIR can be large and difficult to quantify, resulting in barriers to clinical practice. Currently, there is no agreement in the RT community on how to quantify these uncertainties and determine thresholds that distinguish a good DIR result from a poor one. This review summarises the current literature on sources of DIR uncertainties and their impact on RT applications. Recommendations are provided on how to handle these uncertainties for patient-specific use, commissioning, and research. Recommendations are also provided for developers and vendors to help users to understand DIR uncertainties and make the application of DIR in RT safer and more reliable.
Jia-wei Li et al 2023 Phys. Med. Biol. 68 23TR01
Breast cancer, which is the most common type of malignant tumor among humans, is a leading cause of death in females. Standard treatment strategies, including neoadjuvant chemotherapy, surgery, postoperative chemotherapy, targeted therapy, endocrine therapy, and radiotherapy, are tailored for individual patients. Such personalized therapies have tremendously reduced the threat of breast cancer in females. Furthermore, early imaging screening plays an important role in reducing the treatment cycle and improving breast cancer prognosis. The recent innovative revolution in artificial intelligence (AI) has aided radiologists in the early and accurate diagnosis of breast cancer. In this review, we introduce the necessity of incorporating AI into breast imaging and the applications of AI in mammography, ultrasonography, magnetic resonance imaging, and positron emission tomography/computed tomography based on published articles since 1994. Moreover, the challenges of AI in breast imaging are discussed.
Open all abstracts, in this tab
Tang et al
Objective: This study aims to address the limitations of traditional methods for calculating linear energy transfer (LET), a critical component in assessing relative biological effectiveness (RBE). Currently, Monte Carlo (MC) simulation, the gold-standard for accuracy, is resource-intensive and slow for dose optimization, while the speedier analytical approximation has compromised accuracy. Our objective was to prototype a deep-learning-based model for calculating dose-averaged LET (LETd) using patient anatomy and dose-to-water (DW) data, facilitating real-time biological dose evaluation and LET optimization within proton treatment planning systems.
Approach: 275 4-field prostate proton Stereotactic Body Radiotherapy (SBRT) plans were analyzed, rendering a total of 1100 fields. Those were randomly split into 880, 110, and 110 fields for training, validation, and testing. A 3D Cascaded UNet model, along with data processing and inference pipelines, was developed to generate patient-specific LETd distributions from CT images and DW. The accuracy of the LETd of the test dataset was evaluated against MC-generated ground truth through voxel-based mean absolute error (MAE) and gamma analysis.
Main Results: The proposed model accurately inferred LETd distributions for each proton field in the test dataset. A single-field LETd calculation took around 100 ms with trained models running on a NVidia A100 GPU. The selected model yielded an average MAE of 0.94±0.14 MeV/cm and a gamma passing rate of 97.4% ± 1.3% when applied to the test dataset, with the largest discrepancy at the edge of fields where the dose gradient was the largest and counting statistics was the lowest.
Significance: This study demonstrates that deep-learning-based models can efficiently calculate LETd with high accuracy as a fast-forward approach. The model shows great potential to be utilized for optimizing the RBE of proton treatment plans. Future efforts will focus on enhancing the model's performance and evaluating its adaptability to different clinical scenarios.
Safari et al
Objective: This study developed an unsupervised motion artifact reduction method for MRI images of patients with brain tumors. The proposed novel design uses multi-parametric multicenter contrast-enhanced T1W (ceT1W) and T2-FLAIR MRI images.
Approach: The proposed framework included two generators, two discriminators, and two feature extractor networks. A 3-fold cross-validation was used to train and fine-tune the hyperparameters of the proposed model using 230 brain MRI images with tumors, which were then tested on 148 patients' in-vivo datasets. An ablation was performed to evaluate the model's compartments. Our model was compared with Pix2pix and CycleGAN. Six evaluation metrics were reported, including normalized mean squared error (NMSE), structural similarity index (SSIM), multi-scale-SSIM (MS-SSIM), peak signal-to-noise ratio (PSNR), visual information fidelity (VIF), and multi-scale gradient magnitude similarity deviation (MS-GMSD). Artifact reduction and consistency of tumor regions, image contrast, and sharpness were evaluated by three evaluators using Likert scales and compared with ANOVA and Tukey's HSD tests. 
Main results: On average, our method outperforms comparative models to remove heavy motion artifacts with the lowest NMSE (18.34±5.07%) and MS-GMSD (0.07±0.03) for heavy motion artifact level. Additionally, our method creates motion-free images with the highest SSIM (0.93±0.04), PSNR (30.63±4.96), and VIF (0.45±0.05) values, along with comparable MS-SSIM (0.96±0.31). Similarly, our method outperformed comparative models in removing in-vivo motion artifacts for different distortion levels except for MS- SSIM and VIF, which have comparable performance with CycleGAN. Moreover, our method had a consistent performance for different artifact levels. For the heavy level of motion artifacts, our method got the highest Likert scores of 2.82±0.52, 1.88±0.71, and 1.02±0.14 (p-values<<0.0001) for our method, CycleGAN, and Pix2pix respectively. Similar trends were also found for other motion artifact levels.
Significance: Our proposed unsupervised method was demonstrated to reduce motion artifacts from the ceT1W brain images under a multi-parametric framework.
Hardt et al
Objective: Recently, a new and promising approach for range verification was proposed. This method requires the use of two different ion species. Due to their equal magnetic rigidity, fully ionized carbon and helium ions can be simultaneously accelerated in accelerators like synchrotrons. At sufficiently high treatment energies, helium ions can exit the patient distally, reaching approximately three times the range
of carbon ions at an equal energy per nucleon. Therefore, the proposal involves adding a small helium fluence to the carbon ion beam and utilizing helium as an online range probe during radiation therapy. This work aims to develop a software framework for treatment planning and motion verification in range-guided radiation therapy using mixed carbon-helium beams. Approach: The developed framework is
based on the open-source treatment planning toolkit matRad. Dose distributions and helium radiographs were simulated using the open-source Monte Carlo package TOPAS. Beam delivery system parameters were obtained from the Heidelberg Ion Therapy Center, and imaging detectors along with reconstruction were facilitated by ProtonVDA. Methods for reconstructing the most likely patient positioning error scenarios and the motion phase of 4DCT are presented for prostate and lung cancer sites. Main results: The developed framework provides the capability to calculate and optimize treatment plans for mixed carbon-helium ion therapy. It can simulate the treatment process and generate helium radiographs for simulated patient geometry, including small beam views. Furthermore, motion reconstruction based on these radiographs seems possible with preliminary validation. Significance: The developed framework can be applied for further experimental work with the promising mixed carbon-helium ion implementation of range-guided radiotherapy. It offers opportunities for adaptation in particle therapy, improving dose accumulation, and enabling patient anatomy reconstruction during radiotherapy.
Shao et al
Objective: Dynamic cone-beam computed tomography (CBCT) can capture high-spatial-resolution, time-varying images for motion monitoring, patient setup, and adaptive planning of radiotherapy. However, dynamic CBCT reconstruction is an extremely ill-posed spatiotemporal inverse problem, as each CBCT volume in the dynamic sequence is only captured by one or a few X-ray projections, due to the slow gantry rotation speed and the fast anatomical motion (e.g., breathing). 
Approach: We developed a machine learning-based technique, prior-model-free spatiotemporal implicit neural representation (PMF-STINR), to reconstruct dynamic CBCTs from sequentially acquired X-ray projections. PMF-STINR employs a joint image reconstruction and registration approach to address the under-sampling challenge, enabling dynamic CBCT reconstruction from singular X-ray projections. Specifically, PMF-STINR uses spatial implicit neural representations to reconstruct a reference CBCT volume, and it applies temporal INR to represent the intra-scan dynamic motion with respect to the reference CBCT to yield dynamic CBCTs. PMF-STINR couples the temporal INR with a learning-based B-spline motion model to capture time-varying deformable motion during the reconstruction. Compared with the previous methods, the spatial INR, the temporal INR, and the B-spline model of PMF-STINR are all learned on the fly during reconstruction in a one-shot fashion, without using any patient-specific prior knowledge or motion sorting/binning. 
Main results: PMF-STINR was evaluated via digital phantom simulations, physical phantom measurements, and a multi-institutional patient dataset featuring various imaging protocols (half-fan/full-fan, full sampling/sparse sampling, different energy and mAs settings, etc.). The results showed that the one-shot learning-based PMF-STINR can accurately and robustly reconstruct dynamic CBCTs and capture highly irregular motion with high temporal (~0.1s) resolution and sub-millimeter accuracy.
Significance: PMF-STINR can reconstruct dynamic CBCTs and solve the intra-scan motion problem from conventional 3D CBCT scans without using any prior anatomical/motion model or motion sorting/binning. It can be a promising tool for motion management by offering richer motion information than traditional 4D-CBCTs.
Pinnock et al
Minimally invasive ablation techniques for renal cancer are becoming more popular due to their low complication rate and rapid recovery period. Despite excellent visualisation, one drawback of the use of computed tomography (CT) in these procedures is the requirement for iodine-based contrast agents, which are associated with adverse reactions and require a higher X-ray dose. The purpose of this work is to examine the use of time information to generate synthetic contrast enhanced images at arbitrary points after contrast agent injection from non-contrast CT images acquired during renal cryoablation cases. To achieve this, we propose a new method of conditioning generative adversarial networks with normalised time stamps and demonstrate that the use of a HyperNetwork is feasible for this task, generating images of competitive quality compared to standard generative modelling techniques. We also show that reducing the receptive field can help tackle challenges in interventional CT data, offering significantly better image quality as well as better performance when generating images for a downstream segmentation task. Lastly, we show that all proposed models are robust enough to perform inference on unseen intra-procedural data, while also improving needle artefacts and generalising contrast enhancement to other clinically relevant regions and features.
Trending on Altmetric
Open all abstracts, in this tab
Didier Lustermans et al 2024 Phys. Med. Biol. 69 105018
Objective. Newer cone-beam computed tomography (CBCT) imaging systems offer reconstruction algorithms including metal artifact reduction (MAR) and extended field-of-view (eFoV) techniques to improve image quality. In this study a new CBCT imager, the new Varian HyperSight CBCT, is compared to fan-beam CT and two CBCT imagers installed in a ring-gantry and C-arm linear accelerator, respectively. Approach. The image quality was assessed for HyperSight CBCT which uses new hardware, including a large-size flat panel detector, and improved image reconstruction algorithms. The decrease of metal artifacts was quantified (structural similarity index measure (SSIM) and root-mean-squared error (RMSE)) when applying MAR reconstruction and iterative reconstruction for a dental and spine region using a head-and-neck phantom. The geometry and CT number accuracy of the eFoV reconstruction was evaluated outside the standard field-of-view (sFoV) on a large 3D-printed chest phantom. Phantom size dependency of CT numbers was evaluated on three cylindrical phantoms of increasing diameter. Signal-to-noise and contrast-to-noise were quantified on an abdominal phantom. Main results. In phantoms with streak artifacts, MAR showed comparable results for HyperSight CBCT and CT, with MAR increasing the SSIM (0.97–0.99) and decreasing the RMSE (62–55 HU) compared to iterative reconstruction without MAR. In addition, HyperSight CBCT showed better geometrical accuracy in the eFoV than CT (Jaccard Conformity Index increase of 0.02–0.03). However, the CT number accuracy outside the sFoV was lower than for CT. The maximum CT number variation between different phantom sizes was lower for the HyperSight CBCT imager (∼100 HU) compared to the two other CBCT imagers (∼200 HU), but not fully comparable to CT (∼50 HU). Significance. This study demonstrated the imaging performance of the new HyperSight CBCT imager and the potential of applying this CBCT system in more advanced scenarios by comparing the quality against fan-beam CT.
Amir Reza Zekavat et al 2024 Phys. Med. Biol. 69 105017
Objective. To report on a micro computed tomography (micro-CT) system capable of x-ray phase contrast imaging and of increasing spatial resolution at constant magnification. Approach. The micro-CT system implements the edge illumination (EI) method, which relies on two absorbing masks with periodically spaced transmitting apertures in the beam path; these split the beam into an array of beamlets and provide sensitivity to the beamlets' directionality, i.e. refraction. In EI, spatial resolution depends on the width of the beamlets rather than on the source/detector point spread function (PSF), meaning that resolution can be increased by decreasing the mask apertures, without changing the source/detector PSF or the magnification. Main results. We have designed a dedicated mask featuring multiple bands with differently sized apertures and used this to demonstrate that resolution is a tuneable parameter in our system, by showing that increasingly small apertures deliver increasingly detailed images. Phase contrast images of a bar pattern-based resolution phantom and a biological sample (a mouse embryo) were obtained at multiple resolutions. Significance. The new micro-CT system could find application in areas where phase contrast is already known to provide superior image quality, while the added tuneable resolution functionality could enable more sophisticated analyses in these applications, e.g. by scanning samples at multiple scales.
Charles Pageot et al 2024 Phys. Med. Biol. 69 105016
Objective. Beam current transformers (BCT) are promising detectors for real-time beam monitoring in ultra-high dose rate (UHDR) electron radiotherapy. However, previous studies have reported a significant sensitivity of the BCT signal to changes in source-to-surface distance (SSD), field size, and phantom material which have until now been attributed to the fluctuating levels of electrons backscattered within the BCT. The purpose of this study is to evaluate this hypothesis, with the goal of understanding and mitigating the variations in BCT signal due to changes in irradiation conditions. Approach. Monte Carlo simulations and experimental measurements were conducted with a UHDR-capable intra-operative electron linear accelerator to analyze the impact of backscattered electrons on BCT signal. The potential influence of charge accumulation in media as a mechanism affecting BCT signal perturbation was further investigated by examining the effects of phantom conductivity and electrical grounding. Finally, the effectiveness of Faraday shielding to mitigate BCT signal variations is evaluated. Main Results. Monte Carlo simulations indicated that the fraction of electrons backscattered in water and on the collimator plastic at 6 and 9 MeV is lower than 1%, suggesting that backscattered electrons alone cannot account for the observed BCT signal variations. However, our experimental measurements confirmed previous findings of BCT response variation up to 15% for different field diameters. A significant impact of phantom type on BCT response was also observed, with variations in BCT signal as high as 14.1% when comparing measurements in water and solid water. The introduction of a Faraday shield to our applicators effectively mitigated the dependencies of BCT signal on SSD, field size, and phantom material. Significance. Our results indicate that variations in BCT signal as a function of SSD, field size, and phantom material are likely driven by an electric field originating in dielectric materials exposed to the UHDR electron beam. Strategies such as Faraday shielding were shown to effectively prevent these electric fields from affecting BCT signal, enabling reliable BCT-based electron UHDR beam monitoring.
Xueyan Tang et al 2024 Phys. Med. Biol.
Objective: This study aims to address the limitations of traditional methods for calculating linear energy transfer (LET), a critical component in assessing relative biological effectiveness (RBE). Currently, Monte Carlo (MC) simulation, the gold-standard for accuracy, is resource-intensive and slow for dose optimization, while the speedier analytical approximation has compromised accuracy. Our objective was to prototype a deep-learning-based model for calculating dose-averaged LET (LETd) using patient anatomy and dose-to-water (DW) data, facilitating real-time biological dose evaluation and LET optimization within proton treatment planning systems.
Approach: 275 4-field prostate proton Stereotactic Body Radiotherapy (SBRT) plans were analyzed, rendering a total of 1100 fields. Those were randomly split into 880, 110, and 110 fields for training, validation, and testing. A 3D Cascaded UNet model, along with data processing and inference pipelines, was developed to generate patient-specific LETd distributions from CT images and DW. The accuracy of the LETd of the test dataset was evaluated against MC-generated ground truth through voxel-based mean absolute error (MAE) and gamma analysis.
Main Results: The proposed model accurately inferred LETd distributions for each proton field in the test dataset. A single-field LETd calculation took around 100 ms with trained models running on a NVidia A100 GPU. The selected model yielded an average MAE of 0.94±0.14 MeV/cm and a gamma passing rate of 97.4% ± 1.3% when applied to the test dataset, with the largest discrepancy at the edge of fields where the dose gradient was the largest and counting statistics was the lowest.
Significance: This study demonstrates that deep-learning-based models can efficiently calculate LETd with high accuracy as a fast-forward approach. The model shows great potential to be utilized for optimizing the RBE of proton treatment plans. Future efforts will focus on enhancing the model's performance and evaluating its adaptability to different clinical scenarios.
Mingzhe Hu et al 2024 Phys. Med. Biol. 69 10TR01
This review paper aims to serve as a comprehensive guide and instructional resource for researchers seeking to effectively implement language models in medical imaging research. First, we presented the fundamental principles and evolution of language models, dedicating particular attention to large language models. We then reviewed the current literature on how language models are being used to improve medical imaging, emphasizing a range of applications such as image captioning, report generation, report classification, findings extraction, visual question response systems, interpretable diagnosis and so on. Notably, the capabilities of ChatGPT were spotlighted for researchers to explore its further applications. Furthermore, we covered the advantageous impacts of accurate and efficient language models in medical imaging analysis, such as the enhancement of clinical workflow efficiency, reduction of diagnostic errors, and assistance of clinicians in providing timely and accurate diagnoses. Overall, our goal is to have better integration of language models with medical imaging, thereby inspiring new ideas and innovations. It is our aspiration that this review can serve as a useful resource for researchers in this field, stimulating continued investigative and innovative pursuits of the application of language models in medical imaging.
Hua-Chieh Shao et al 2024 Phys. Med. Biol.
Objective: Dynamic cone-beam computed tomography (CBCT) can capture high-spatial-resolution, time-varying images for motion monitoring, patient setup, and adaptive planning of radiotherapy. However, dynamic CBCT reconstruction is an extremely ill-posed spatiotemporal inverse problem, as each CBCT volume in the dynamic sequence is only captured by one or a few X-ray projections, due to the slow gantry rotation speed and the fast anatomical motion (e.g., breathing). 
Approach: We developed a machine learning-based technique, prior-model-free spatiotemporal implicit neural representation (PMF-STINR), to reconstruct dynamic CBCTs from sequentially acquired X-ray projections. PMF-STINR employs a joint image reconstruction and registration approach to address the under-sampling challenge, enabling dynamic CBCT reconstruction from singular X-ray projections. Specifically, PMF-STINR uses spatial implicit neural representations to reconstruct a reference CBCT volume, and it applies temporal INR to represent the intra-scan dynamic motion with respect to the reference CBCT to yield dynamic CBCTs. PMF-STINR couples the temporal INR with a learning-based B-spline motion model to capture time-varying deformable motion during the reconstruction. Compared with the previous methods, the spatial INR, the temporal INR, and the B-spline model of PMF-STINR are all learned on the fly during reconstruction in a one-shot fashion, without using any patient-specific prior knowledge or motion sorting/binning. 
Main results: PMF-STINR was evaluated via digital phantom simulations, physical phantom measurements, and a multi-institutional patient dataset featuring various imaging protocols (half-fan/full-fan, full sampling/sparse sampling, different energy and mAs settings, etc.). The results showed that the one-shot learning-based PMF-STINR can accurately and robustly reconstruct dynamic CBCTs and capture highly irregular motion with high temporal (~0.1s) resolution and sub-millimeter accuracy.
Significance: PMF-STINR can reconstruct dynamic CBCTs and solve the intra-scan motion problem from conventional 3D CBCT scans without using any prior anatomical/motion model or motion sorting/binning. It can be a promising tool for motion management by offering richer motion information than traditional 4D-CBCTs.
Mark A Pinnock et al 2024 Phys. Med. Biol.
Minimally invasive ablation techniques for renal cancer are becoming more popular due to their low complication rate and rapid recovery period. Despite excellent visualisation, one drawback of the use of computed tomography (CT) in these procedures is the requirement for iodine-based contrast agents, which are associated with adverse reactions and require a higher X-ray dose. The purpose of this work is to examine the use of time information to generate synthetic contrast enhanced images at arbitrary points after contrast agent injection from non-contrast CT images acquired during renal cryoablation cases. To achieve this, we propose a new method of conditioning generative adversarial networks with normalised time stamps and demonstrate that the use of a HyperNetwork is feasible for this task, generating images of competitive quality compared to standard generative modelling techniques. We also show that reducing the receptive field can help tackle challenges in interventional CT data, offering significantly better image quality as well as better performance when generating images for a downstream segmentation task. Lastly, we show that all proposed models are robust enough to perform inference on unseen intra-procedural data, while also improving needle artefacts and generalising contrast enhancement to other clinically relevant regions and features.
Caleb Sample et al 2024 Phys. Med. Biol. 69 105001
Objective. To improve intravoxel incoherent motion imaging (IVIM) magnetic resonance Imaging quality using a new image denoising technique and model-independent parameterization of the signal versus b-value curve. Approach. IVIM images were acquired for 13 head-and-neck patients prior to radiotherapy. Post-radiotherapy scans were also acquired for five of these patients. Images were denoised prior to parameter fitting using neural blind deconvolution, a method of solving the ill-posed mathematical problem of blind deconvolution using neural networks. The signal decay curve was then quantified in terms of several area under the curve (AUC) parameters. Improvements in image quality were assessed using blind image quality metrics, total variation (TV), and the correlations between parameter changes in parotid glands with radiotherapy dose levels. The validity of blur kernel predictions was assessed by the testing the method's ability to recover artificial 'pseudokernels'. AUC parameters were compared with monoexponential, biexponential, and triexponential model parameters in terms of their correlations with dose, contrast-to-noise (CNR) around parotid glands, and relative importance via principal component analysis. Main results. Image denoising improved blind image quality metrics, smoothed the signal versus b-value curve, and strengthened correlations between IVIM parameters and dose levels. Image TV was reduced and parameter CNRs generally increased following denoising. AUC parameters were more correlated with dose and had higher relative importance than exponential model parameters. Significance. IVIM parameters have high variability in the literature and perfusion-related parameters are difficult to interpret. Describing the signal versus b-value curve with model-independent parameters like the AUC and preprocessing images with denoising techniques could potentially benefit IVIM image parameterization in terms of reproducibility and functional utility.
Yongshun Xu et al 2024 Phys. Med. Biol. 69 105012
Objective. Cardiac computed tomography (CT) is widely used for diagnosis of cardiovascular disease, the leading cause of morbidity and mortality in the world. Diagnostic performance depends strongly on the temporal resolution of the CT images. To image the beating heart, one can reduce the scanning time by acquiring limited-angle projections. However, this leads to increased image noise and limited-angle-related artifacts. The goal of this paper is to reconstruct high quality cardiac CT images from limited-angle projections. Approach. The ability to reconstruct high quality images from limited-angle projections is highly desirable and remains a major challenge. With the development of deep learning networks, such as U-Net and transformer networks, progresses have been reached on image reconstruction and processing. Here we propose a hybrid model based on the U-Net and Swin-transformer (U-Swin) networks. The U-Net has the potential to restore structural information due to missing projection data and related artifacts, then the Swin-transformer can gather a detailed global feature distribution. Main results. Using synthetic XCAT and clinical cardiac COCA datasets, we demonstrate that our proposed method outperforms the state-of-the-art deep learning-based methods. Significance. It has a great potential to freeze the beating heart with a higher temporal resolution.
Sébastien Quetin et al 2024 Phys. Med. Biol. 69 105011
Objective. Monte Carlo (MC) simulations are the benchmark for accurate radiotherapy dose calculations, notably in patient-specific high dose rate brachytherapy (HDR BT), in cases where considering tissue heterogeneities is critical. However, the lengthy computational time limits the practical application of MC simulations. Prior research used deep learning (DL) for dose prediction as an alternative to MC simulations. While accurate dose predictions akin to MC were attained, graphics processing unit limitations constrained these predictions to large voxels of 3 mm × 3 mm × 3 mm. This study aimed to enable dose predictions as accurate as MC simulations in 1 mm × 1 mm × 1 mm voxels within a clinically acceptable timeframe. Approach. Computed tomography scans of 98 breast cancer patients treated with Iridium-192-based HDR BT were used: 70 for training, 14 for validation, and 14 for testing. A new cropping strategy based on the distance to the seed was devised to reduce the volume size, enabling efficient training of 3D DL models using 1 mm × 1 mm × 1 mm dose grids. Additionally, novel DL architecture with layer-level fusion were proposed to predict MC simulated dose to medium-in-medium (Dm,m). These architectures fuse information from TG-43 dose to water-in-water (Dw,w) with patient tissue composition at the layer-level. Different inputs describing patient body composition were investigated. Main results. The proposed approach demonstrated state-of-the-art performance, on par with the MC Dm,m maps, but 300 times faster. The mean absolute percent error for dosimetric indices between the MC and DL-predicted complete treatment plans was 0.17% ± 0.15% for the planning target volume V100, 0.30% ± 0.32% for the skin D2cc, 0.82% ± 0.79% for the lung D2cc, 0.34% ± 0.29% for the chest wall D2cc and 1.08% ± 0.98% for the heart D2cc. Significance. Unlike the time-consuming MC simulations, the proposed novel strategy efficiently converts TG-43 Dw,w maps into precise Dm,m maps at high resolution, enabling clinical integration.