Medical physics最新文献

{"title":"Response characterization of radiochromic OC-1 films in photon, electron, and proton beams","authors":"Qinghao Chen, Xiandong Zhao, Jufri Setianegara, Yao Hao, Tianyu Zhao, Tiezhi Zhang, Arash Darafsheh","doi":"10.1002/mp.17356","DOIUrl":"10.1002/mp.17356","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <h3> Background</h3>\u0000 \u0000 <p>Radiochromic film (RCF) dosimeters with their high spatial resolution and tissue equivalent properties are conveniently used for two-dimensional and small-field dosimetry. OC-1 is a new model of RCF dosimeter that was commercially introduced recently. Due to its novelty there is a need to characterize its response in various radiation beam types.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Purpose</h3>\u0000 \u0000 <p>To study the response of OC-1 RCFs to megavoltage clinical x-ray, electron, and proton beams, as well as kilovoltage x-ray beams used in a small animal research irradiator.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Materials and methods</h3>\u0000 \u0000 <p>OC-1 RCFs were cut into ∼4 × 4 cm<sup>2</sup> pieces. RCF samples were irradiated at various dose levels in the range 0.5–120 Gy using different modalities; a small animal radiation research platform (SARRP) (220 kVp), a medical linear accelerator (6 MV, 10 MV, 15 MV, 6 MV FFF, 10 MV FFF photon beams, as well as 6 and 20 MeV electron beams), and a gantry-mounted proton therapy synchrocyclotron. In order to study any dependency on the fractionation scheme, same dose was delivered at several fractions to a set of films. Different dose rates in the range 200–600 MU/min were delivered to a set of films to investigate any dose rate dependency. The films were scanned pre-irradiation and at 48 h post-irradiation using a flatbed scanner. The net optical density (OD) was measured for red, green, and blue color channel for each film. The orientation dependency was studied by scanning the films at eight different orientations. In order to study the temporal evolution of the response of the films, film samples were irradiated at 10 and 50 Gy using 6 MV photon beams and were scanned upon irradiation at certain time intervals up to 3 months. The spectral response of the films were studied over the visible range using a spectrometer.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Results</h3>\u0000 \u0000 <p>For megavoltage photon, electron, and plateau region of the proton beams, we did not observe a significant dependency on the beam quality, dose rate, and fractionation scheme. At the kV beam, an unusual over-response was observed in the films’ net OD. An orientation dependency in the response of the films with a sinusoidal trend was observed. The response of the films increased with time following a double or triple exponential trend. The spectral absorption peaks were blue-shifted with dose.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Conclusion</h3>\u0000 \u0000 ","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 11","pages":"8584-8596"},"PeriodicalIF":3.2,"publicationDate":"2024-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/mp.17356","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142074870","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transformer- and joint learning-based dual-domain networks for undersampled MRI segmentation","authors":"Jizhong Duan,&nbsp;Zhenyu Huang,&nbsp;Yunshuang Xie,&nbsp;Junfeng Wang,&nbsp;Yu Liu","doi":"10.1002/mp.17358","DOIUrl":"10.1002/mp.17358","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <h3> Background</h3>\u0000 \u0000 <p>Recently, magnetic resonance imaging (MRI) has become a crucial medical imaging technology widely used in clinical practice. However, MRI faces challenges such as the lengthy acquisition time for k-space data and the need for time-consuming manual annotation by radiologists. Traditionally, these challenges have been addressed individually through undersampled MRI reconstruction and automatic segmentation algorithms. Whether undersampled MRI segmentation can be enhanced by treating undersampled MRI reconstruction and segmentation as an end-to-end task, trained simultaneously, rather than as serial tasks should be explored.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Purpose</h3>\u0000 \u0000 <p>We introduce a novel Transformer- and Joint Learning-based Dual-domain Network (TJLD-Net) for undersampled MRI segmentation.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Methods</h3>\u0000 \u0000 <p>This method significantly enhances feature recognition in the segmentation process by fully utilizing the rich detail obtained during the image reconstruction phase. Consequently, the method can achieve precise and reliable image segmentation even with undersampled k-space data. Additionally, it incorporates an attention mechanism for feature enhancement, which improves the representation of shared features by learning the contextual information in MR images.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Results</h3>\u0000 \u0000 <p>Simulation experiments demonstrate that the segmentation performance of TJLD-Net on three datasets is significantly higher than that of the joint model (RecSeg) and six baseline models (where reconstruction and segmentation are regarded as serial tasks). On the CHAOS dataset, the Dice scores of TJLD-Net are, on average, 9.87%, 2.17%, 1.90%, 1.80%, 9.60%, 0.80%, and 6.50% higher than those of the seven compared models. On the ATLAS challenge dataset, the average Dice scores of TJLD-Net improve by 4.23%, 5.63%, 2.30%, 1.53%, 3.57%, 0.93%, and 6.60%. Similarly, on the SKM-TEA dataset, the average Dice scores of TJLD-Net improve by 4.73%, 12.80%, 14.83%, 8.67%, 4.53%, 11.60%, and 12.10%. The novel TJLD-Net model provides a promising solution for undersampled MRI segmentation, overcoming the poor performance issues encountered by automated segmentation algorithms in low-quality accelerated imaging.</p>\u0000 </section>\u0000 </div>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 11","pages":"8108-8123"},"PeriodicalIF":3.2,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142019977","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impact of breathing signal-guided dose modulation on step-and-shoot 4D CT image reconstruction","authors":"Lukas Wimmert,&nbsp;Annette Schwarz,&nbsp;Tobias Gauer,&nbsp;Christian Hofmann,&nbsp;Jannis Dickmann,&nbsp;Thilo Sentker,&nbsp;Rene Werner","doi":"10.1002/mp.17360","DOIUrl":"10.1002/mp.17360","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Breathing signal-guided 4D CT sequence scanning such as the intelligent 4D CT (i4DCT) approach reduces imaging artifacts compared to conventional 4D CT. By design, i4DCT captures entire breathing cycles during beam-on periods, leading to redundant projection data and increased radiation exposure to patients exhibiting prolonged exhalation phases. A recently proposed breathing-guided dose modulation (DM) algorithm promises to lower the imaging dose by temporarily reducing the CT tube current, but the impact on image reconstruction and the resulting images have not been investigated.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;We evaluate the impact of breathing signal-guided DM on 4D CT image reconstruction and corresponding images.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;This study is designed as a comparative and retrospective analysis based on 104 4D CT datasets. Each dataset underwent retrospective reconstruction twice: (a) utilizing the acquired clinical projection data for reconstruction, which yields reference image data, and (b) excluding projections acquired during potential DM phases from image reconstruction, resulting in DM-affected image data. Resulting images underwent automatic organ segmentation (lung/liver). (Dis)Similarity of reference and DM-affected images were quantified by the Dice coefficient of the entire organ masks and the organ overlaps within the DM-affected slices. Further, for lung cases, (a) and (b) were deformably registered and median magnitudes of the obtained displacement field were computed. Eventually, for 17 lung cases, gross tumor volumes (GTV) were recontoured on both (a) and (b). Target volume similarity was quantified by the Hausdorff distance.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;DM resulted in a median imaging dose reduction of 15.4% (interquartile range [IQR]: 11.3%–19.9%) for the present patient cohort. Dice coefficients for lung (&lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;n&lt;/mi&gt;\u0000 &lt;mo&gt;=&lt;/mo&gt;\u0000 &lt;mn&gt;73&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$n=73$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;) and liver (&lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;n&lt;/mi&gt;\u0000 &lt;mo&gt;=&lt;/mo&gt;\u0000 &lt;mn&gt;31&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$n=31$&lt;/annotation&gt;\u0000 ","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 10","pages":"7119-7126"},"PeriodicalIF":3.2,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/mp.17360","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142019975","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Addressing intra- and inter-institution variability of a radiomic framework based on Apparent Diffusion Coefficient in prostate cancer","authors":"Letizia Morelli,&nbsp;Chiara Paganelli,&nbsp;Giulia Marvaso,&nbsp;Giovanni Parrella,&nbsp;Simone Annunziata,&nbsp;Maria Giulia Vicini,&nbsp;Mattia Zaffaroni,&nbsp;Matteo Pepa,&nbsp;Paul Eugene Summers,&nbsp;Ottavio De Cobelli,&nbsp;Giuseppe Petralia,&nbsp;Barbara Alicja Jereczek-Fossa,&nbsp;Guido Baroni","doi":"10.1002/mp.17355","DOIUrl":"10.1002/mp.17355","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Prostate cancer (PCa) is a highly heterogeneous disease, making tailored treatment approaches challenging. Magnetic resonance imaging (MRI), notably diffusion-weighted imaging (DWI) and the derived Apparent Diffusion Coefficient (ADC) maps, plays a crucial role in PCa characterization. In this context, radiomics is a very promising approach able to disclose insights from MRI data. However, the sensitivity of radiomic features to MRI settings, encompassing DWI protocols and multicenter variations, requires the development of robust and generalizable models.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;To develop a comprehensive radiomics framework for noninvasive PCa characterization using ADC maps, focusing on identifying reliable imaging biomarkers against intra- and inter-institution variations.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Materials and methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Two patient cohorts, including an internal cohort (118 PCa patients) used for both training (75%) and hold-out testing (25%), and an external cohort (50 PCa patients) for independent testing, were employed in the study. DWI images were acquired with three different DWI protocols on two different MRI scanners: two DWI protocols acquired on a 1.5-T scanner for the internal cohort, and one DWI protocol acquired on a 3-T scanner for the external cohort. One hundred and seven radiomics features (i.e., shape, first order, texture) were extracted from ADC maps of the whole prostate gland. To address variations in DWI protocols and multicenter variability, a dedicated pipeline, including two-way ANOVA, sequential-feature-selection (SFS), and ComBat features harmonization was implemented. Mann–Whitney &lt;i&gt;U&lt;/i&gt;-tests (&lt;i&gt;α&lt;/i&gt; = 0.05) were performed to find statistically significant features dividing patients with different tumor characteristics in terms of Gleason score (GS) and T-stage. Support-Vector-Machine models were then developed to predict GS and T-stage, and the performance was assessed through the area under the curve (AUC) of receiver-operating-characteristic curves.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Downstream of ANOVA, two subsets of 38 and 41 features stable against DWI protocol were identified for GS and T-stage, respectively. Among these, SFS revealed the most predictive features, yielding an AUC of 0.75 (GS) and 0.70 (T-stage) in the hold-out test. Employing ComBat harmonization improved the external-test performance of the GS model, raising AUC from 0.72 to 0.78.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 ","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 11","pages":"8096-8107"},"PeriodicalIF":3.2,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/mp.17355","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142019973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhanced IDOL segmentation framework using personalized hyperspace learning IDOL","authors":"Byong Su Choi,&nbsp;Chris J. Beltran,&nbsp;Sven Olberg,&nbsp;Xiaoying Liang,&nbsp;Bo Lu,&nbsp;Jun Tan,&nbsp;Alessio Parisi,&nbsp;Janet Denbeigh,&nbsp;Sridhar Yaddanapudi,&nbsp;Jin Sung Kim,&nbsp;Keith M. Furutani,&nbsp;Justin C. Park,&nbsp;Bongyong Song","doi":"10.1002/mp.17361","DOIUrl":"10.1002/mp.17361","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Adaptive radiotherapy (ART) workflows have been increasingly adopted to achieve dose escalation and tissue sparing under shifting anatomic conditions, but the necessity of recontouring and the associated time burden hinders a real-time or online ART workflow. In response to this challenge, approaches to auto-segmentation involving deformable image registration, atlas-based segmentation, and deep learning-based segmentation (DLS) have been developed. Despite the particular promise shown by DLS methods, implementing these approaches in a clinical setting remains a challenge, namely due to the difficulty of curating a data set of sufficient size and quality so as to achieve generalizability in a trained model.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;To address this challenge, we have developed an intentional deep overfit learning (IDOL) framework tailored to the auto-segmentation task. However, certain limitations were identified, particularly the insufficiency of the personalized dataset to effectively overfit the model. In this study, we introduce a personalized hyperspace learning (PHL)-IDOL segmentation framework capable of generating datasets that induce the model to overfit specific patient characteristics for medical image segmentation.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;The PHL-IDOL model is trained in two stages. In the first, a conventional, general model is trained with a diverse set of patient data (&lt;i&gt;n&lt;/i&gt; = 100 patients) consisting of CT images and clinical contours. Following this, the general model is tuned with a data set consisting of two components: (a) selection of a subset of the patient data (&lt;i&gt;m&lt;/i&gt; &lt; &lt;i&gt;n&lt;/i&gt;) using the similarity metrics (mean square error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and the universal quality image index (UQI) values); (b) adjust the CT and the clinical contours using a deformed vector generated from the reference patient and the selected patients using (a). After training, the general model, the continual model, the conventional IDOL model, and the proposed PHL-IDOL model were evaluated using the volumetric dice similarity coefficient (VDSC) and the Hausdorff distance 95% (HD95%) computed for 18 structures in 20 test patients.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Implementing the PHL-IDOL framework resulted in improved segmentation performance for each patient. The Dice scores increased from 0.81&lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mo&gt;±&lt;/mo&gt;\u0000 &lt;annotation&gt;$ pm $&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;0.05 wit","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 11","pages":"8568-8583"},"PeriodicalIF":3.2,"publicationDate":"2024-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142019974","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AAPM Task Group Report 325: MRI static magnetic field homogeneity measurement and evaluation procedures – Guidance and resources","authors":"Christina L Brunnquell,&nbsp;Trevor J Andrews,&nbsp;Samuel W Fielden,&nbsp;Kathryn Huff,&nbsp;David W Jordan,&nbsp;Michael T O'Shea,&nbsp;Dustin K Ragan,&nbsp;Michael A Tressler,&nbsp;Travis C Salzillo,&nbsp;Joseph Och","doi":"10.1002/mp.17351","DOIUrl":"10.1002/mp.17351","url":null,"abstract":"<p>Measurement of static magnetic field (B₀) homogeneity is an essential component of routine MRI system evaluation. This report summarizes the work of AAPM Task Group (TG) 325 on vendor-specific methods of B₀ homogeneity measurement and evaluation. TG 325 was charged with producing a set of detailed, step-by-step instructions to implement B₀ homogeneity measurement methods discussed in the American College of Radiology (ACR) MRI Quality Control Manual using specific makes and models of MRI scanners. The TG produced such instructions for as many approaches as was relevant and practical on six currently available vendor platforms including details of software/tools, settings, phantoms, and other experimental details needed for a reproducible protocol. Because edits to these instructions may need to be made as vendors enter and exit the market and change available tools, interfaces, and access levels over time, the step-by-step instructions are published as a living document on the AAPM website. This summary document provides an introduction to B₀ homogeneity testing in MRI and several of the common methods for its measurement and evaluation. A living document on the AAPM website provides vendor-specific step-by-step instructions for performing these tests to facilitate accurate and reproducible B₀ homogeneity evaluation on a routine basis.</p>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 10","pages":"7038-7046"},"PeriodicalIF":3.2,"publicationDate":"2024-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142019940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Semi-supervised abdominal multi-organ segmentation by object-redrawing","authors":"Min Jeong Cho,&nbsp;Jae Sung Lee","doi":"10.1002/mp.17364","DOIUrl":"10.1002/mp.17364","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Multi-organ segmentation is a critical task in medical imaging, with wide-ranging applications in both clinical practice and research. Accurate delineation of organs from high-resolution 3D medical images, such as CT scans, is essential for radiation therapy planning, enhancing treatment outcomes, and minimizing radiation toxicity risks. Additionally, it plays a pivotal role in quantitative image analysis, supporting various medical research studies. Despite its significance, manual segmentation of multiple organs from 3D images is labor-intensive and prone to low reproducibility due to high interoperator variability. Recent advancements in deep learning have led to several automated segmentation methods, yet many rely heavily on labeled data and human anatomy expertise.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;In this study, our primary objective is to address the limitations of existing semi-supervised learning (SSL) methods for abdominal multi-organ segmentation. We aim to introduce a novel SSL approach that leverages unlabeled data to enhance the performance of deep neural networks in segmenting abdominal organs. Specifically, we propose a method that incorporates a redrawing network into the segmentation process to correct errors and improve accuracy.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Our proposed method comprises three interconnected neural networks: a segmentation network for image segmentation, a teacher network for consistency regularization, and a redrawing network for object redrawing. During training, the segmentation network undergoes two rounds of optimization: basic training and readjustment. We adopt the Mean-Teacher model as our baseline SSL approach, utilizing labeled and unlabeled data. However, recognizing significant errors in abdominal multi-organ segmentation using this method alone, we introduce the redrawing network to generate redrawn images based on CT scans, preserving original anatomical information. Our approach is grounded in the generative process hypothesis, encompassing segmentation, drawing, and assembling stages. Correct segmentation is crucial for generating accurate images. In the basic training phase, the segmentation network is trained using both labeled and unlabeled data, incorporating consistency learning to ensure consistent predictions before and after perturbations. The readjustment phase focuses on reducing segmentation errors by optimizing the segmentation network parameters based on the differences between redrawn and original CT images.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;We evaluated our method using two publicly a","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 11","pages":"8334-8347"},"PeriodicalIF":3.2,"publicationDate":"2024-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142019976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Characterizing devices for validation of dose, dose rate, and LET in ultra high dose rate proton irradiations","authors":"Nathan Harrison,&nbsp;Serdar Charyyev,&nbsp;Cristina Oancea,&nbsp;Alexander Stanforth,&nbsp;Edgar Gelover,&nbsp;Shuang Zhou,&nbsp;William S Dynan,&nbsp;Tiezhi Zhang,&nbsp;Steven Biegalski,&nbsp;Liyong Lin","doi":"10.1002/mp.17359","DOIUrl":"10.1002/mp.17359","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <h3> Background</h3>\u0000 \u0000 <p>Ultra high dose rate (UHDR) radiotherapy using ridge filter is a new treatment modality known as conformal FLASH that, when optimized for dose, dose rate (DR), and linear energy transfer (LET), has the potential to reduce damage to healthy tissue without sacrificing tumor killing efficacy via the FLASH effect.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Purpose</h3>\u0000 \u0000 <p>Clinical implementation of conformal FLASH proton therapy has been limited by quality assurance (QA) challenges, which include direct measurement of UHDR and LET. Voxel DR distributions and LET spectra at planning target margins are paramount to the DR/LET-related sparing of organs at risk. We hereby present a methodology to achieve experimental validation of these parameters.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Methods</h3>\u0000 \u0000 <p>Dose, DR, and LET were measured for a conformal FLASH treatment plan involving a 250-MeV proton beam and a 3D-printed ridge filter designed to uniformly irradiate a spherical target. We measured dose and DR simultaneously using a 4D multi-layer strip ionization chamber (MLSIC) under UHDR conditions. Additionally, we developed an “<span>u</span>nder-<span>s</span>ample and <span>re</span>cover (USRe)” technique for a high-resolution pixelated semiconductor detector, Timepix3, to avoid event pile-up and to correct measured LET at high-proton-flux locations without undesirable beam modifications. Confirmation of these measurements was done using a MatriXX PT detector and by Monte Carlo (MC) simulations.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Results</h3>\u0000 \u0000 <p>MC conformal FLASH computed doses had gamma passing rates of &gt;95% (3 mm/3% criteria) when compared to MatriXX PT and MLSIC data. At the lateral margin, DR showed average agreement values within 0.3% of simulation at 100 Gy/s and fluctuations ∼10% at 15 Gy/s. LET spectra in the proximal, lateral, and distal margins had Bhattacharyya distances of &lt;1.3%.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Conclusion</h3>\u0000 \u0000 <p>Our measurements with the MLSIC and Timepix3 detectors shown that the DR distributions for UHDR scenarios and LET spectra using USRe are in agreement with simulations. These results demonstrate that the methodology presented here can be used effectively for the experimental validation and QA of FLASH treatment plans.</p>\u0000 </section>\u0000 </div>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 11","pages":"8411-8422"},"PeriodicalIF":3.2,"publicationDate":"2024-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141997144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Prediction of CD8+T lymphocyte infiltration levels in gastric cancer from contrast-enhanced CT and clinical factors using machine learning","authors":"Wentao Xie,&nbsp;Sheng Jiang,&nbsp;Fangjie Xin,&nbsp;Zinian Jiang,&nbsp;Wenjun Pan,&nbsp;Xiaoming Zhou,&nbsp;Shuai Xiang,&nbsp;Zhenying Xu,&nbsp;Yun Lu,&nbsp;Dongsheng Wang","doi":"10.1002/mp.17350","DOIUrl":"10.1002/mp.17350","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;CD8+ T lymphocyte infiltration is closely associated with the prognosis and immunotherapy response of gastric cancer (GC). For now, the examination of CD8 infiltration levels relies on endoscopic biopsy, which is invasive and unsuitable for longitude assessment during anti-tumor therapy.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;This work aims to develop and validate a noninvasive workflow based on contrast-enhanced CT (CECT) images to evaluate the CD8+ T-cell infiltration profiles of GC.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;GC patients were retrospectively and consecutively enrolled and randomly assigned to the training (validation) or test cohort at a 7:3 ratio. All patients were binary classified into the CD8-high (infiltrated proportion ≥ 20%) or CD8-low group (infiltrated proportion &lt; 20%) group. A total of 1170 radiomics features were extracted from each presurgical CECT series. After feature selection, fifteen radiomics features were transmitted to three independent machine-learning models for the computation of predictive radiological scores. Multilayer perceptron (MLP) was applied to merge the radiological scores with clinical factors. The predictive efficacy of the radiological scores and of the combined model was evaluated by receiver operating characteristic curve, calibration curve, and decision curve analysis in both the training and test cohorts.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;A total of 210 patients were enrolled in this study (mean age: 63.22 ± 8.74 years, 151 men), and were randomly assigned to the training set (&lt;i&gt;n&lt;/i&gt; = 147) or the test set (&lt;i&gt;n&lt;/i&gt; = 63). The merged radiological score was correlated with CD8 infiltration in both the training (&lt;i&gt;p&lt;/i&gt; = 1.8e−10) and test cohorts (&lt;i&gt;p&lt;/i&gt; = 0.00026). The combined model integrating the radiological scores and clinical features achieved an area under the curve (AUC) value of 0.916 (95% CI: 0.872–0.960) in the training set and 0.844 (95% CI: 0.742–0.946) in the test set for classifying CD8-high GCs. The model was well-calibrated and exhibited net benefit over “treat-all” and“treat-none” strategies in decision curve analysis.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Conclusions&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Artificial intelligent systems combining radiological features and clinical factors could accurately predict CD8 infiltration levels of GC, which may benefit personalized treatment of GC in the context of immunotherapy.&lt;/p&gt;\u0000 ","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 10","pages":"7108-7118"},"PeriodicalIF":3.2,"publicationDate":"2024-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/mp.17350","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141997147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dose perturbations at tissue interfaces during parallel linac-MR treatments: The “Lateral Scatter Electron Return Effect” (LS-ERE)","authors":"Stephen Steciw,&nbsp;B. Gino Fallone,&nbsp;Eugene Yip","doi":"10.1002/mp.17363","DOIUrl":"10.1002/mp.17363","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Magnetic resonance (MR) imaging devices have been integrated with medical linear accelerators (linac) in radiation therapy. Both perpendicular linac-MR (LMR-B⊥) and parallel (LMR-B∥) systems exist, where due to the MR's magnetic field dose can be perturbed in the patient. Dose perturbations from the electron return effect (ERE) and electron streaming effects (ESEs) are present in LMR-B⊥ systems, where a dose collimating effect has been observed in LMR-B∥ systems\u0000.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;To report on an asymmetric dose perturbation which is present at the interface between two different materials during treatment in &lt;i&gt;parallel&lt;/i&gt; linac-MR (LMR-B∥) systems. To the best of our knowledge, these asymmetric dose effects, “Lateral Scattered Electron Return Effect” (LS-ERE) have not been previously reported.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;BEAMnrc and EGSnrc Monte Carlo (MC) radiation transport codes were used with the EEMF macro to emulate a 6 FFF beam from the 0.5-T Alberta linac-MR (LMR). Simulations were performed at 0.5 and 1.5 T in several different phantom material–interface combinations and field sizes including from modulated MLC-like fields. MC simulations quantified LS-ERE in patient CT datasets for the head, breast, and lung. LS-ERE cancellation techniques were investigated. LS-ERE asymmetries were quantified by subtracting an antiparallel dose from the parallel dose, dividing by two and normalizing to the global 0-T maximum dose. GafChromic film measurements were made in the 0.5-T Alberta LMR-B∥ system using solid water at the water–air interface to validate MC simulations. ERE was simulated for an emulated LMR-B⊥ system and compared to LMR-B∥ dose perturbations.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;LS-ERE is mostly independent of field size for fields &gt;1 × 1 cm&lt;sup&gt;2&lt;/sup&gt;. For 5 × 5-cm&lt;sup&gt;2&lt;/sup&gt; fields at 0.5T/1.5T, LS-ERE asymmetries are ≤±6.9%/6.9% at bone–air and ≤±9.0%/7.0% at tissue–air for nonair doses, and ≤±4.1%/5.5% at tissue–lung interfaces. LS-ERE increases as the density gradient increases, where the magnitude and extent of LS-ERE are reduced as field strength increases. For a single 5 × 5-cm&lt;sup&gt;2&lt;/sup&gt; field at 0.5T/1.5T, the LS-ERE asymmetry is ≤±10.2%/8.5% at the tissue–air sinus interface for head, ≤±4.2%/5.3% at the spine–lung interface for the lung, and ≤±5.7%/4.9% at the skin–air interface for a breast tangent plan at 0.5T/1.5T. POP fields mostly remove LS-ERE asymmetries, with magnetic field reversal during treatment being the mo","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"51 11","pages":"8506-8523"},"PeriodicalIF":3.2,"publicationDate":"2024-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/mp.17363","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141997145","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}