• 学术搜索
  • 科研智能体
    • Research Labs
    • AI 阅读
    • AI 文库
    • 深度研究
    • 学者亮点
  • 学术资源
    • AI2000
    • 期刊/会议
    • 学者库
    • 学术API
      MCP&Skill
    • 溯源树
    • 数据集
  • 知识沉淀
    • 学术空间
订阅小程序
旧版功能
aminer vip
开通会员低至0.73元/天
一次搞定AI科研
立即登录
  • English
  • 联系方式
    Journal of applied clinical medical physics

    Journal of applied clinical medical physics

    JournalISSN 1526-9914eISSN 1526-9914中科院 医学 4区

    年发文量

    研究主题

    论文(5021)

    排序
    1Evaluation of GPU‐accelerated Monte Carlo Beam and Agility MLC Modeling in the Monaco Treatment Planning System
    Fasco van Ommen, Sima Sajadi Shahrbabaki, Thomas A. Foppen,Jordi Saez,Victor Hernandez,Bram van Asselen,Sara L. Hackett

    Abstract Background Treatment planning systems (TPSs) are essential in modern radiotherapy. Accurate beam modeling, especially of multileaf collimators (MLCs), is crucial, as their complex designs often introduce systematic uncertainties in planned and delivered doses. Thorough evaluation of the beam model and MLCs must be performed before clinical introduction. Purpose This study evaluates a new GPU‐accelerated Monte Carlo (MC) beam and MLC model (VSM2.0) from Elekta, with a focus on aspects of the MLCs that were previously shown to be particularly challenging to model. Methods Measurements were conducted on two Elekta Versa HD linacs with Agility MLCs. All calculations were performed in Monaco v6.2.3 using VSM1.6 and VSM2.0. Open field dose distributions were measured and calculated in a water phantom and compared using a gamma analysis (1%/1 mm gamma criterion). MLC characteristics including the leaf tip penumbra, measured with film, the dosimetric leaf gap (DLG) and the tongue‐and‐groove effect, as measured via sweeping gap fields, were compared with calculations from both beam models. These measurements also highlighted some challenges associated with MLC model verification. Finally, patient treatment plans with a wide range of target volumes and modulation were measured and compared against calculations (gamma analysis: 3%/2 mm gamma criterion). Results The VSM2.0 model showed excellent agreement with measurements for open field dose distributions, with gamma pass rates for almost all fields above 90%. The clinical treatment plans showed gamma pass rates above 94%. The VSM1.6 model showed similar accuracy. However, the VSM2.0 modeled leaf transmission more accurately than VSM1.6 and improved representation of the tongue‐and‐groove effect across different gap widths and offsets. By contrast, the maximum deviations from the measured DLGs were 0.15 mm and 0.43 mm for VSM1.6 and 2.0 respectively, largely due to user‐definable tuning of parameters for VSM1.6, which allowed the VSM1.6 model to closely match linac performance. Conclusions The new GPU‐based single phase‐space beam model shows excellent agreement with measurements and improves simulation of complex aspects of the leaf geometry. Reducing customization simplifies the commissioning process of planning systems, thereby potentially reducing associated risks, but the possibility to tune the leaf offset per beam model may further improve the model accuracy.

    2026
    引用
    AI阅读
    加入学术空间
    2Cardiac Substructure Dose in Thoracic VMAT Without Explicit Constraints: FMO Versus SPO on Urt‐tpois with Cross‐platform Reference to Eclipse PO
    Chun Wu, Zhenmei Cao, Tingting Duan, Lixiang Han, Bin Wang,Weijie Lei

    Abstract Background Cardiac substructure doses are increasingly recognized as more specific predictors of radiation‐induced cardiac toxicity than whole‐heart metrics, yet these substructures are often left unconstrained in routine thoracic volumetric‐modulated arc therapy (VMAT) planning due to contouring burden or low prioritization. Purpose This study aims to evaluate, in the absence of explicit cardiac substructure constraints, whether optimizer selection is associated with differences in cardiac substructure doses within a single treatment planning system (TPS), and to provide a descriptive cross‐system reference against a widely used clinical platform. Methods Thirty patients (10 each across esophageal, breast, and non‐small cell lung cancers) initially treated with clinically delivered Photon Optimizer (PO) plans on the Eclipse platform were retrospectively replanned using two optimization algorithms on the uRT‐TPOIS platform: Fluence Map Optimization (FMO) and Stochastic Platform Optimization (SPO). No explicit dose constraints were applied to any cardiac substructure. Prespecified primary endpoints were mean heart dose (MHD), mean dose to the left anterior descending artery (LAD Dmean), mean dose to the left ventricle (LV Dmean), and mean lung dose (MLD), with the latter serving as a pulmonary toxicity reference. All other dosimetric parameters were analyzed as secondary/exploratory endpoints. Paired comparisons used the Wilcoxon signed‐rank test, with Benjamini–Hochberg false discovery rate correction applied separately within each cancer type and pairwise comparison for primary endpoints. Secondary endpoints were exploratory and not adjusted for multiplicity. Results Intra‐platform comparison within uRT‐TPOIS showed that SPO was associated with lower MHD, LAD Dmean, and MLD than FMO in esophageal cancer, and lower MHD, LAD Dmean, and LV Dmean in breast cancer ( q < 0.05). In NSCLC, no significant differences were observed for the primary endpoints. In secondary analyses, SPO was associated with shorter beam‐on time across all sites and improved target coverage in NSCLC. In descriptive cross‐system comparisons with Eclipse PO, SPO was associated with lower primary endpoint doses in esophageal and breast cancers, whereas in NSCLC, only MHD differed significantly, favoring PO. Exploratory analyses suggested additional differences in other cardiac substructures, but these were not adjusted for multiplicity and should be considered hypothesis‐generating. Conclusion Under the specific uRT‐TPOIS protocol and without explicit cardiac substructure constraints, SPO was associated with lower primary endpoint doses than FMO in esophageal and breast cancers, but not in NSCLC. Cross‐system comparisons were descriptive and confounded by platform‐level factors. Whether explicit cardiac substructure constraints provide incremental benefit remains to be tested in dedicated within‐platform studies.

    2026
    引用
    AI阅读
    加入学术空间
    3Comparative Assessment of Respiratory‐induced Tumor Motion Using Fluoroscopy and Four‐dimensional CT of Solitary Pulmonary Lesions
    Masato Kato,Takahiro Kato,Yuki Narita,Masao Murakami

    Abstract Background Four‐dimensional computed tomography (4DCT) is widely used to quantify respiratory‐induced tumor motion in thoracic radiotherapy. However, 4DCT captures discretized respiratory phases and assumes stable breathing, potentially introducing reconstruction‐related uncertainties. Fluoroscopy enables real‐time observation across multiple breathing cycles, yet direct large‐scale comparisons between fluoroscopy‐ and 4DCT‐derived motion measurements remain scarce. Purpose To directly compare respiratory‐induced tumor motion measured by fluoroscopy and 4DCT in patients with solitary pulmonary lesions undergoing proton therapy planning under free breathing without abdominal compression. Methods This retrospective study included 191 patients with solitary pulmonary lesions who underwent sequential fluoroscopy and 4DCT under free‐breathing conditions as part of proton therapy planning. Tumor motion was evaluated in the craniocaudal direction and stratified by lung lobe. Furthermore, diaphragmatic motion was analyzed in 131 cases. Spearman's rank correlation coefficient was used to assess correlations between motions measured by the two modalities. Measurement differences were defined as ΔD (4DCT minus fluoroscopy) and analyzed according to fluoroscopy‐measured respiratory displacement. For subgroup analyses, cases were categorized into < 5 mm, ≥ 5 to ≤ 10 mm, and > 10 mm motion groups based on fluoroscopy‐measured respiratory displacement. For Bland–Altman analyses, motion magnitude was defined as the mean of the fluoroscopy‐ and 4DCT‐derived respiratory displacement measurements, and intermodality agreement was assessed within each motion category. Results The correlation coefficients for tumor motion in the upper‐, middle‐, and lower‐lobes were 0.61 ( p < 0.01), 0.76 ( p < 0.01), and 0.55 ( p < 0.01), respectively. The corresponding correlation coefficients for diaphragmatic motion were lower: 0.40 ( p < 0.01) on the left and 0.38 ( p < 0.01) on the right. The median |ΔD| values (interquartile range) were 1.3 mm (0.7–1.9 mm), 2.4 mm (1.0–4.1 mm), and 4.7 mm (2.6–6.8 mm) for the < 5 mm, ≥ 5 to ≤ 10 mm, and > 10 mm groups, respectively. Overall, the mean difference between 4DCT and fluoroscopy was small (−0.2 mm; 95% confidence interval, −0.9 to 0.6 mm), with 95% limits of agreement of −10.2 to 9.9 mm. The limits of agreement widened with increasing respiratory motion. Conclusions In this single‐institution study of patients undergoing proton therapy planning, respiratory motion measurements derived from fluoroscopy and 4DCT showed larger absolute differences for lower‐lobe tumors (median |ΔD| = 4.20 mm) than for upper‐ and middle‐lobe tumors (1.70 and 1.90 mm, respectively). These findings indicate that respiratory motion estimated by 4DCT alone should be interpreted with caution in such cases. When feasible and when the lesion is fluoroscopically visible, fluoroscopic assessment may serve as a useful consistency check to complement 4DCT‐based motion evaluation.

    2026
    引用
    AI阅读
    加入学术空间
    4Leveraging Longitudinal ACR CT Dose Index Registry Data to Improve Interpretation of Institutional CT Dose Performance
    Azmul H. Siddique, Mike Weaver,Gary Ge, Leslie Anaskevich,Jie Zhang

    Abstract Background The American College of Radiology CT Dose Index Registry (DIR) is a national benchmarking program that enables institutions to compare CT dose metrics with aggregated reference distributions to support CT radiation dose optimization and quality assurance. In practice, benchmarking is typically reviewed at discrete reporting intervals, which can be limited by benchmark shifts, transient variation, or incomplete reference data. Purpose This study evaluates how longitudinal DIR analysis improves interpretation of institutional CT dose performance and supports protocol optimizations. Methods Longitudinal DIR data from 2020 to 2025 were analyzed for 10 common CT exams across 21 reporting quarters for both adult and pediatric. CTDI vol , DLP, and SSDE distributions (25th, 50th, 75th percentiles) were compared with exam‐specific DIR reference percentiles. Exams with n ≥ 30 per reporting interval were included for analysis. Exceedance rates, defined as the median value of facility metric exceeding the DIR 75th percentile (Diagnostic Reference Level [DRL]), were evaluated across exams and reporting intervals. Linear mixed‐effects (LME) regression was used to evaluate longitudinal trends and identify exams demonstrating statistically significant dose drift. Coefficient of variation (CV) and F‐test were performed to assess year‐to‐year variability. Results Four of ten adult exams remained within DRL across all intervals for both CTDI vol and DLP. CT Neck with IV contrast exceeded the CTDI vol and DLP DRLs in 19/21 intervals, and CT C‐spine without IV contrast exceeded the DLP DRL in 19/21 intervals. Pediatric data demonstrated exceedance rates greater than 50% for both CTDI vol and DLP in those same exams. For adult exams with seven available DIR SSDE benchmarks, five remained within the SSDE DRL across all intervals. LME analysis identified a significant increase in institutional dose metrics in 2025 ( β 2 = 0.139; p = 0.001), with CT C‐Spine contributing largest positive shift (30.56%) for adults. Facility median variability (CV) ranged from 4.6%–127%, whereas DIR 75th percentile variability ranged from 2.8% to 21.9%. Conclusion Longitudinal analysis of ACR DIR executive summary data extends isolated quarterly review by contextualizing exceedances over time, accounting for benchmark variability, and distinguishing persistent exam behavior from transient effects. This supports improved interpretation of institutional CT dose performance and targeted exam triage.

    2026
    引用
    AI阅读
    加入学术空间
    5Attention Xception U‐Net (axunet): an Explainable Combination of CNN and Self‐attention for Reliable Glioma Segmentation in MRI
    Farzan Moodi,Fereshteh Khodadadi Shoushtari,Gelareh Valizadeh, Dornaz Mazinani,Hanieh Mobarak Salari,Sedigheh Sina,Hamidreza Saligheh Rad

    Abstract Background Gliomas are the most common primary malignant brain tumors and exhibit highly heterogeneous morphology and infiltrative growth patterns. Accurate delineation of tumor subregions in magnetic resonance imaging (MRI) is therefore critical for diagnosis, treatment planning, and longitudinal monitoring, yet manual segmentation remains time‐consuming and subject to inter‐observer variability. Purpose Accurate glioma segmentation in magnetic resonance imaging (MRI) is essential for diagnosis, treatment planning, and reproducible tumor delineation. However, for clinical adoption, automated segmentation systems must provide not only high accuracy but also robustness and interpretability. This study aimed to develop an explainable and reliable deep learning framework for glioma segmentation and to evaluate its performance on both internal and external datasets. Method We used the BraTS 2021 training dataset, which contains 1251 multi‐institutional pre‐operative glioma MRI cases. After comparing multiple sequence combinations, T1ce, T2, and Fluid Attenuated Inversion Recovery (FLAIR) were selected as the final three‐channel input. Preprocessing included tumor‐oriented slice selection, cropping, Min‐Max normalization, and augmentation. We developed AXUNet, a UNet‐based architecture that integrates memory‐efficient Xception blocks with a self‐attention module combining kernel‐based pixel attention and dot‐product channel attention. Internal ablation analysis and Gradient‐weighted Class Activation Mapping (Grad‐CAM)‐based visualization were also performed. Results On the BraTS 2021 test partition, AXUNet achieved the highest mean Dice score (93.73), outperforming UNet (93.26), Inception‐UNet (90.88), Attention Gate UNet (AG‐UNet) (90.38), Attention ResUNet (AResUNet) (92.80), and Xception‐UNet (93.24). Region‐wise Dice scores were 92.59 for Whole Tumor (WT), 86.81 for Tumor Core (TC), and 84.89 for Enhancing tumor (ET). Although AResUNet achieved a slightly higher ET score (84.92), AXUNet demonstrated the strongest performance in mean Dice, WT, and TC. To further assess generalizability, AXUNet was independently evaluated on 219 previously unseen BraTS 2023 cases through the Synapse platform, achieving Dice scores of 86.29% for WT, 78.53% for TC, and 69.00% for ET. Conclusion AXUNet demonstrates that combining Xception‐based encoding with self‐attention can improve glioma segmentation while also supporting explainability and external robustness, positioning it as a promising framework for MRI‐based glioma segmentation. Its clinical relevance, and the practical value of its interpretability outputs, remain to be established through further, dedicated validation beyond Dice‐based benchmarking.

    2026
    引用
    AI阅读
    加入学术空间
    立即登录,查看全部 5021 篇论文

    高被引作者

    作者引用发文
    Stephen F. Kry163732
    Timothy D Solberg110532
    Peter A. Balter98834
    Vladimir Feygelman96627
    Sotirios Stathakis93328
    Laurence E. Court92251
    Fang-Fang Yin76123
    Geoffrey Zhang75523
    Richard A Popple75033
    Indra J Das71524

    高产作者

    作者引用发文
    Michael D. Mills25160
    Yi Rong56359
    Laurence E. Court92251
    Jay Burmeister30543
    Damodar Pokhrel52343
    Mitsuhiro Nakamura34542
    Peter A. Balter98834
    Ping Xia63434
    Takashi Mizowaki29033
    Richard A Popple75033

    相关期刊

    Igaku butsuri : Nihon Igaku Butsuri Gakkai kikanshi = Japanese journal of medical physics : an official journal of Japan Society of Medical Physics
    Nihon Hoshasen Gijutsu Gakkai zasshi
    Journal of medical physics
    Radiological Physics and Technology
    Physics and imaging in radiation oncology
    EJNMMI Physics
    Medical Physics
    Physica Medica
    PHYSICS & IMAGinG in RADIATION ONCOLOGY
    Physica Medica-European Journal of Medical Physics