Physics in medicine and biology最新文献

{"title":"Multi-institution investigations of online daily adaptive proton strategies for head and neck cancer patients.","authors":"Evangelia Choulilitsa, Mislav Bobić, Brian A Winey, Harald Paganetti, Antony John Lomax, Francesca Albertini","doi":"10.1088/1361-6560/adbb51","DOIUrl":"https://doi.org/10.1088/1361-6560/adbb51","url":null,"abstract":"<p><strong>Objective: </strong>Fast computation of daily reoptimization is key for an efficient online adaptive proton therapy workflow. Various approaches aim to expedite this process, often compromising daily dose. This study compares MGH's online dose reoptimization approach, PSI's online replanning workflow and a full reoptimization adaptive workflow for head and neck cancer (H&N) patients.&#xD;&#xD;Approach:10 H&N patients (PSI:5, MGH:5) with daily CBCTs were included. Synthetic CTs were created by deforming the planning CT to each CBCT. Targets and OARs were deformed on daily images. Three adaptive approaches were investigated: i) an online dose reoptimization approach modifying the fluence of a subset of beamlets, ii) full reoptimization adaptive workflow modifying the fluence of all beamlets, and iii) a full online replanning approach, allowing the optimizer to modify both fluence and position of all beamlets. Two non-adapted (NA) scenarios were simulated by recalculating the original plan on the daily image using: Monte Carlo for NAMGH and raycasting algorithm for NAPSI.&#xD; &#xD;Main results:All adaptive scenarios from both institutions achieved the prescribed daily target dose, with further improvements from online replanning. For all patients, low-dose CTV D98% shows mean daily deviations of -2.2%, -1.1%, and 0.4% for workflows i, ii, and iii, respectively. For the online adaptive scenarios, plan optimization averages 2.2 minutes for iii) and 2.4 for i) while the full dose reoptimization requires 72 minutes. The OAMGH20% dose reoptimization approach produced results comparable to online replanning for most patients and fractions. However, for one patient, differences up to 11% in low-dose CTV D98% occurred.&#xD;&#xD;Significance:Despite significant anatomical changes, all three adaptive approaches ensure target coverage without compromising OAR sparing. Our data suggests 20% dose reoptimization suffices, for most cases, yielding comparable results to online replanning with a marginal time increase due to Monte Carlo. For optimal daily adaptation, a rapid online replanning is preferable.&#xD.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143524138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A comparative analysis of image harmonization techniques in mitigating differences in CT acquisition and reconstruction.","authors":"Anil Yadav, Spencer Welland, John M Hoffman, Grace Hyun J Kim, Matthew S Brown, Ashley E Prosper, Denise R Aberle, Michael F McNitt-Gray, William Hsu","doi":"10.1088/1361-6560/adabad","DOIUrl":"10.1088/1361-6560/adabad","url":null,"abstract":"<p><p><i>Objective</i>. The study aims to systematically characterize the effect of CT parameter variations on images and lung radiomic and deep features, and to evaluate the ability of different image harmonization methods to mitigate the observed variations.<i>Approach</i>. A retrospective in-house sinogram dataset of 100 low-dose chest CT scans was reconstructed by varying radiation dose (100%, 25%, 10%) and reconstruction kernels (smooth, medium, sharp). A set of image processing, convolutional neural network (CNNs), and generative adversarial network-based (GANs) methods were trained to harmonize all image conditions to a reference condition (100% dose, medium kernel). Harmonized scans were evaluated for image similarity using peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and learned perceptual image patch similarity (LPIPS), and for the reproducibility of radiomic and deep features using concordance correlation coefficient (CCC).<i>Main Results</i>. CNNs consistently yielded higher image similarity metrics amongst others; for Sharp/10%, which exhibited the poorest visual similarity, PSNR increased from a mean ± CI of 17.763 ± 0.492 to 31.925 ± 0.571, SSIM from 0.219 ± 0.009 to 0.754 ± 0.017, and LPIPS decreased from 0.490 ± 0.005 to 0.275 ± 0.016. Texture-based radiomic features exhibited a greater degree of variability across conditions, i.e. a CCC of 0.500 ± 0.332, compared to intensity-based features (0.972 ± 0.045). GANs achieved the highest CCC (0.969 ± 0.009 for radiomic and 0.841 ± 0.070 for deep features) amongst others. CNNs are suitable if downstream applications necessitate visual interpretation of images, whereas GANs are better alternatives for generating reproducible quantitative image features needed for machine learning applications.<i>Significance</i>. Understanding the efficacy of harmonization in addressing multi-parameter variability is crucial for optimizing diagnostic accuracy and a critical step toward building generalizable models suitable for clinical use.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11866762/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143009972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DRACO: differentiable reconstruction for arbitrary CBCT orbits.","authors":"Chengze Ye, Linda-Sophie Schneider, Yipeng Sun, Mareike Thies, Siyuan Mei, Andreas Maier","doi":"10.1088/1361-6560/adbb50","DOIUrl":"https://doi.org/10.1088/1361-6560/adbb50","url":null,"abstract":"<p><strong>Objective: </strong>This study introduces a novel method for reconstructing cone beam computed tomography (CBCT) images for arbitrary orbits, addressing the computational and memory challenges associated with traditional iterative reconstruction algorithms.</p><p><strong>Approach: </strong>The proposed method employs a differentiable shift-variant filtered backprojection neural network, optimized for arbitrary trajectories. By integrating known operators into the learning model, the approach minimizes the number of trainable parameters while enhancing model interpretability. This framework adapts seamlessly to specific orbit geometries, including non-continuous trajectories such as circular-plus-arc or sinusoidal paths, enabling faster and more accurate CBCT reconstructions.</p><p><strong>Main results: </strong>Experimental validation demonstrates that the method significantly accelerates reconstruction, reducing computation time by over 97% compared to conventional iterative algorithms. It achieves superior or comparable image quality with reduced noise, as evidenced by a 38.6% reduction in mean squared error, a 7.7% increase in peak signal-to-noise ratio, and a 5.0% improvement in the structural similarity index measure. The flexibility and robustness of the approach are confirmed through its ability to handle data from diverse scan geometries.</p><p><strong>Significance: </strong>This method represents a significant advancement in interventional medical imaging, particularly for robotic C-arm CT systems, enabling real-time, high-quality CBCT reconstructions for customized orbits. It offers a transformative solution for clinical applications requiring computational efficiency and precision in imaging. Code Availability: Code is available at https://github.com/ChengzeYe/Defrise-and-Clack-reconstruction.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143524137","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CycleH-CUT: an unsupervised medical image translation method based on cycle consistency and hybrid contrastive learning.","authors":"Weiwei Jiang, Yingyu Qin, Xiaoyan Wang, Qiuju Chen, Qiu Guan, Minhua Lu","doi":"10.1088/1361-6560/adb2d7","DOIUrl":"10.1088/1361-6560/adb2d7","url":null,"abstract":"<p><p>Unsupervised medical image translation tasks are challenging due to the difficulty of obtaining perfectly paired medical images. CycleGAN-based methods have proven effective in unpaired medical image translation. However, these methods can produce artifacts in the generated medical images. To address this issue, we propose an unsupervised network based on cycle consistency and hybrid contrastive unpaired translation (CycleH-CUT). CycleH-CUT consists of two CUT (H-CUT) networks. In the H-CUT network, a query-selected attention mechanism is adopted to select queries with important features. The boosted contrastive learning loss is employed to reweight all negative patches via the optimal transport strategy. We further apply spectral normalization to improve training stability, allowing the generator to extract complex features. On the basis of the H-CUT network, a new CycleH-CUT framework is proposed to integrate contrastive learning and cycle consistency. Two H-CUT networks are used to reconstruct the generated images back to the source domain, facilitating effective translation between unpaired medical images. We conduct extensive experiments on three public datasets (BraTS, OASIS3, and IXI) and a private Spinal Column dataset to demonstrate the effectiveness of CycleH-CUT and H-CUT. Specifically, CycleH-CUT achieves an average SSIM of 0.926 in the BraTS dataset, an average SSIM of 0.796 on the OASIS3 dataset, an average SSIM of 0.932 on the IXI dataset, and an average SSIM of 0.890 on the private Spinal Column dataset.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143256367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"3DΠ: three-dimensional positron imaging, a novel total-body PET scanner using xenon-doped liquid argon scintillator.","authors":"Azam Zabihi, Xinran Li, Alejandro Ramirez, Iftikhar Ahmad, Manuel Dionisio Da Rocha Rolo, Davide Franco, Federico Gabriele, Cristiano Galbiati, Michela Lai, Daniel R Marlow, Andrew Renshaw, Shawn Westerdale, Masayuki Wada","doi":"10.1088/1361-6560/adbaac","DOIUrl":"10.1088/1361-6560/adbaac","url":null,"abstract":"<p><strong>Objective: </strong>This paper introduces a novel PET imaging methodology called 3-dimensional positron imaging (3DΠ), which integrates total-body (TB) coverage, time-of-flight (TOF) technology, ultra-low dose imaging capabilities, and ultra-fast readout electronics inspired by emerging technology from the DarkSide collaboration.</p><p><strong>Approach: </strong>The study evaluates the performance of 3DΠ using Monte Carlo simulations based on NEMA NU 2-2018 protocols. The methodology employs a homogenous, monolithic scintillator composed of liquid argon (LAr) doped with xenon (Xe) with silicon photomultipliers (SiPM) operating at cryogenic temperatures.</p><p><strong>Main results: </strong>Significant enhancements in system performance are observed, with the 3DΠ system achieving a noise equivalent count rate (NECR) of 3.2 Mcps which is approximately two times higher than uEXPLORER's peak NECR (1.5 Mcps) at 17.3 (kBq/mL). Spatial resolution measurements show an average FWHM of 2.7 mm across both axial positions. The system exhibits superior sensitivity, with values reaching 373 kcps/MBq with a line source at the center of the field of view. Additionally, 3DΠ achieves a TOF resolution of 151 ps at 5.3 kBq/mL, highlighting its potential to produce high-quality images with reduced noise levels.</p><p><strong>Significance: </strong>The study underscores the potential of 3DΠ in improving PET imaging performance, offering the potential for shorter scan times and reduced radiation exposure for patients. The Xe-doped LAr offers advantages such as fast scintillation, enhanced light yield, and cost-effectiveness. Future research will focus on optimizing system geometry and further refining reconstruction algorithms to exploit the strengths of 3DΠ for clinical applications.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143516494","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Metal artifacts correction based on a physics-informed nonlinear sinogram completion model.","authors":"Shuqiong Fan, Mengfei Li, Chuwen Huang, Xiaojuan Deng, Hongwei Li","doi":"10.1088/1361-6560/adbaad","DOIUrl":"https://doi.org/10.1088/1361-6560/adbaad","url":null,"abstract":"<p><strong>Objective: </strong>Metal artifacts seriously deteriorate CT image quality. Current metal artifacts reduction methods suffer from insufficient correction or easily introduce secondary artifacts. To better suppress metal artifacts, we propose a sinogram completion approach extracting and utilizing useful information that contained in the corrupted metal trace projections.</p><p><strong>Approach: </strong>Our method mainly contains two stages: sinogram interpolation by an improved normalization technique for initial correction and physics-informed nonlinear sinogram decomposition for further improvement. In the first stage, different from the popular normalized metal artifact reduction method, we propose a more meaningful normalization scheme for the interpolation procedure. In the second stage, instead of performing a linear sinogram decomposition as done in the physics-informed sinogram completion method, we introduce a nonlinear decomposition model that can accurately separate the sinogram into metal and non-metal contributions by better modeling the physical scanning process. The interpolated sinogram and physics-informed correction compensate each other to reach the optimal correction results.</p><p><strong>Main results: </strong>Experimental results on simulated and real data indicate that, in terms of both structures preservation and detail recovery, the proposed PNSC method achieves very competitive performance for metal artifacts reduction compared to existing methods.</p><p><strong>Significance: </strong>According to our knowledge, it's for the first time that a nonlinear sinogram decomposition model is proposed in the literature for metal artifacts correction. It might motivate further research exploring this idea for various sinogram processing tasks.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143516514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Self-supervised U-transformer network with mask reconstruction for metal artifact reduction.","authors":"Fanning Kong, Zaifeng Shi, Huaisheng Cao, Yudong Hao, Qingjie Cao","doi":"10.1088/1361-6560/adbaae","DOIUrl":"https://doi.org/10.1088/1361-6560/adbaae","url":null,"abstract":"<p><strong>Objective: </strong>Metal artifacts severely damaged human tissue information from the computed tomography (CT) image, posing significant challenges to disease diagnosis. Deep learning (DL) has been widely explored for the metal artifact reduction (MAR) task. Nevertheless, paired metal artifact CT datasets suitable for training do not exist in reality. Although the synthetic CT image dataset provides additional training data, the trained networks still generalize poorly to real metal artifact data.&#xD;Approach. A self-supervised U-shaped Transformer network (SUTransNet) is proposed to focus on model generalizability enhancement in MAR tasks. This framework consists of a self-supervised mask reconstruction pre-text task and a down-stream task. In the pre-text task, the CT images are randomly corrupted by masks. They are recovered with themselves as the label, aiming at acquiring the artifacts and tissue structure of the actual physical situation. Down-stream task fine-tunes MAR target through labeled images. Utilizing the multi-layer long-range feature extraction capabilities of the Transformer efficiently captures features of metal artifacts. The incorporation of the MAR bottleneck allows for the distinction of metal artifact features through cross-channel self-attention.&#xD;Main result. Experiments demonstrate that the framework maintains strong generalization ability in the MAR task, effectively preserving tissue details while suppressing metal artifacts. The results achieved a peak signal-to-noise ratio (PSNR) of 43.86 dB and a structural similarity index (SSIM) of 0.9863 while ensuring the efficiency of the model inference. In addition, the Dice coefficient and Mean Intersection over Union (MIoU) are improved by 11.70% and 9.51% in the segmentation of the MAR image, respectively.&#xD;Significance. The combination of unlabeled real-artifact CT images and labeled synthetic-artifact CT images facilitates a self-supervised learning process that positively contributes to model generalizability.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143516518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated hallucination detection for synthetic CT images used in MR-only radiotherapy workflows.","authors":"Abdul K Parchur, Mohammad Zarenia, Colette Gage, Eric S Paulson, Ergun Ahunbay","doi":"10.1088/1361-6560/adb5eb","DOIUrl":"10.1088/1361-6560/adb5eb","url":null,"abstract":"<p><p><i>Objective</i>. Artificial intelligence (AI)-generated synthetic CT (sCT) images have become commercially available to provide electron densities and reference anatomies in MR-only radiotherapy workflows. However, hallucinations (false regions of bone or air) introduced in AI-generated sCT images may affect the accuracy of dose calculation and patient setup verification. We developed a tool to detect bone hallucinations and/or inaccuracies in AI-generated pelvic sCT images used in MR-only workflows.<i>Approach</i>. A deep learning auto segmentation (DLAS) model was trained to auto-segment bone on MR images. The model was implemented with a 3D SegResNet network architecture using the MONAI framework with a training dataset of 86 Dixon MR image sets paired with their corresponding ground truth contours derived from planning CT images deformed to the MR images. The model performance was then assessed on an independent testing dataset (<i>n</i>= 10).<i>Main results</i>. The DLAS model-based hallucination screener identified hallucinations in bone structures using daily MR images and accurately flagged these regions on sCT images. The sensitivity of the screener is adjustable based on the distance of discrepancies between bone regions derived from sCT to bone contours generated by the DLAS. The average specificity of the DLAS model was 0.78, 0.93 and 0.98 for distance parameters of 0.8, 1.0 and 1.2 cm, respectively. The screener identified false high-density hallucination regions in the abdomen of AI-generated sCT images for all testing patients, highlighting potential issues with the training data used for the AI sCT model.<i>Significance</i>. A hallucination screener for AI-generated pelvic sCT images was developed and implemented for routine clinical use. The screener serves as an important quality assurance tool for MR-only radiotherapy workflows. By identifying potential AI-generated errors, the hallucination screener may improve the safety and accuracy of sCT images used for dose calculation and image guidance.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143414142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Retrospective study on the resonance of thermoacoustic emissions and their possible biological implications in cats treated with electron FLASH beams.","authors":"Julie Lascaud, Martin Rädler, Carla Rohrer Bley, Marie-Catherine Vozenin, Katia Parodi","doi":"10.1088/1361-6560/adb679","DOIUrl":"10.1088/1361-6560/adb679","url":null,"abstract":"<p><p><i>Objective.</i>Radiotherapy delivered at an ultra-high dose rate (UHDR) is a promising cancer treatment. In the last years, it has been shown to selectively reduce toxicity in healthy tissue by triggering the so-called FLASH effect achieved through specific temporal dose fractionation. However, the increase of the instantaneous dose rate results in the production of stronger thermoacoustic emissions for microsecond or shorter pulsed ionizing beams, which could potentially impact the treatment outcomes. Focusing on scenarios expected to create the highest acoustic intensities, the objectives of this work were to assess whether acoustic resonance can theoretically occur<i>in vivo</i>and how it could be mitigated in cases where it would influence the biological response.<i>Approach.</i>Thermoacoustic emissions were retrospectively simulated from post-treatment x-ray computed tomography scans of cats irradiated with a single high dose of electron FLASH to treat squamous carcinoma of the nasal planum. The peak dose, pressure intensity and location of the acoustic resonance were assessed for different beam positioning and reproduced for three animals.<i>Main results.</i>Irradiation of nasal planum in cats using a frontal electron beam results in pressure hot spots due to acoustic resonance that are observed in the vicinity of the rostral maxillary bone. The pressure distribution is mostly influenced by the anatomy (i.e. geometry and heterogeneous composition of the irradiated object), whereas its intensity largely depends on the irradiation setup. While further experimental investigation is needed to understand and mitigate potential associated risks, our results underline that acoustic phenomena so far neglected in conventional radiotherapy may need to be accounted for when using UHDR delivery.<i>Significance.</i>We show that specific irradiation scenarios can induce geometry-dependent thermoacoustic resonances<i>in vivo</i>which may be of sufficient magnitude to induce biological effects and impact the outcomes of FLASH radiotherapy.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143425898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A deep learning model for inter-fraction head and neck anatomical changes in proton therapy.","authors":"Tiberiu Burlacu, Mischa Hoogeman, Danny Lathouwers, Zoltán Perkó","doi":"10.1088/1361-6560/adba39","DOIUrl":"https://doi.org/10.1088/1361-6560/adba39","url":null,"abstract":"<p><p><b>Objective:</b>To assess the performance of a probabilistic deep learning based algorithm for predicting inter-fraction anatomical changes in head and neck patients.&#xD;&#xD;<b>Approach:</b>A probabilistic daily anatomy model for head and neck patients (DAM_HN) is built on the variational autoencoder architecture. The model approximates the generative joint conditional probability distribution of the repeat computed tomography (rCT) images and their corresponding masks on the planning CT images (pCT) and their masks. The model outputs deformation vector fields, which are used to produce possible rCTs and associated masks. The dataset is composed of 93 patients (i.e., 315 pCT - rCT pairs), 9 (i.e., 27 pairs) of which were set aside for final testing. The performance of the model is assessed based on the reconstruction accuracy and the generative performance for the set aside patients. &#xD;&#xD;<b>Main results:</b>The model achieves a DICE score of 0.83 and an image similarity score (NCC) of 0.60 on the test set. The generated parotid glands, spinal cord and constrictor muscle volume change distributions and center of mass shift distributions were also assessed. For all organs, the medians of the distributions are close to the true ones, and the distributions are broad enough to encompass the real observed changes. Moreover, the generated images display anatomical changes in line with the literature reported ones, such as the medial shifts of the parotids glands. &#xD;&#xD;<b>Significance:</b>DAM_HNis capable of generating realistic anatomies observed during the course of the treatment and has applications in anatomical robust optimization, treatment planning based on plan library approaches and robustness evaluation against inter-fractional changes.</p>","PeriodicalId":20185,"journal":{"name":"Physics in medicine and biology","volume":" ","pages":""},"PeriodicalIF":3.3,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143502922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}