Physics and Imaging in Radiation Oncology最新文献_第10页

First experimental demonstration of magnetic resonance-guided multileaf collimator tracking for (ultra-)hypofractionated prostate radiotherapy 磁共振引导多叶准直跟踪用于（超）低分割前列腺放疗的首次实验演示

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-09-13 DOI: 10.1016/j.phro.2025.100828

Prescilla Uijtewaal , Pim T.S. Borman , Peter L. Woodhead , Hans C.J. de Boer , Bas W. Raaymakers , Martin F. Fast

Background and purpose:

(Ultra-)hypofractionated radiotherapy is an effective treatment for localized prostate cancer, but intrafraction motion can increase toxicity and/or reduce treatment efficacy. Therefore, motion management is essential. This study explores magnetic resonance imaging (MRI)-guided multileaf collimator (MLC) tracking for 2-fraction prostate radiotherapy on an MR-linac.

Materials and methods:

We compared two MRI-guided MLC centroid tracking workflows, each using a different motion manager to derive and stream target positions to our in-house MLC tracking software. The first workflow relies on interleaved 2D (2.5D) cine-MRI, introducing minimal latency. In contrast, the second workflow utilized 3D cine-MRI, which operates at a relatively lower imaging frequency that introduces more latency.

For experimental validation, we used a motion phantom equipped with an integrated insert that combines film with plastic scintillation dosimetry. A 2x12 Gy 11-beam prostate intensity modulated radiotherapy plan was created for tracking deliveries.

Results:

The signal latency introduced by the motion managers was 0.6 s for 2.5D cine-MRI and 6.3 s for 3D cine-MRI. Despite this latency, MLC tracking effectively restored the planned dose, improving the 2%/2mm local gamma pass-rates from 21% (due to linear drift) to 89% (2.5D) and 91% (3D). Plastic scintillator measurements showed reduced dose deviations at the periphery of the clinical target volume from 13–64% (no tracking) to 0–11% (2.5D) and 2–26% (3D).

Conclusion:

Our experiments demonstrated the technical feasibility of 2.5D and 3D cine-MRI-based MLC tracking on an MR-linac for 2-fraction prostate radiotherapy, with both motion management strategies achieving comparable dosimetric improvements despite the difference in latency.

背景与目的：（超）低分割放疗是治疗局限性前列腺癌的一种有效方法，但术中运动可增加毒性和/或降低治疗效果。因此，运动管理是必不可少的。本研究探讨了磁共振成像（MRI）引导的多叶准直器（MLC）在磁共振直线加速器上对2段前列腺放疗的跟踪。材料和方法：我们比较了两种mri引导的MLC质心跟踪工作流程，每一个都使用不同的运动管理器来导出和传输目标位置到我们内部的MLC跟踪软件。第一个工作流程依赖于交错2D （2.5D）电影mri，引入最小的延迟。相比之下，第二种工作流程使用3D电影mri，其成像频率相对较低，会带来更多延迟。为了实验验证，我们使用了一个装有集成插入物的运动幻影，该插入物结合了薄膜和塑料闪烁剂量测定法。创建2x12 Gy 11束前列腺强度调制放疗计划，用于跟踪分娩。结果：运动管理器引入的信号潜伏期为2.5D电影- mri为0.6 s， 3D电影- mri为6.3 s。尽管存在这种延迟，MLC跟踪有效地恢复了计划剂量，将2%/2mm的局部伽马通过率从21%（由于线性漂移）提高到89% （2.5D）和91% （3D）。塑料闪烁体测量显示，临床靶体积周边的剂量偏差从13-64%（无跟踪）减少到0-11% （2.5D）和2-26% （3D）。结论：我们的实验证明了基于2.5D和3D电影mri的MLC跟踪在MR-linac上用于2段前列腺放疗的技术可行性，尽管延迟不同，但两种运动管理策略都取得了相当的剂量学改善。

{"title":"First experimental demonstration of magnetic resonance-guided multileaf collimator tracking for (ultra-)hypofractionated prostate radiotherapy","authors":"Prescilla Uijtewaal , Pim T.S. Borman , Peter L. Woodhead , Hans C.J. de Boer , Bas W. Raaymakers , Martin F. Fast","doi":"10.1016/j.phro.2025.100828","DOIUrl":"10.1016/j.phro.2025.100828","url":null,"abstract":"<div><h3>Background and purpose:</h3><div>(Ultra-)hypofractionated radiotherapy is an effective treatment for localized prostate cancer, but intrafraction motion can increase toxicity and/or reduce treatment efficacy. Therefore, motion management is essential. This study explores magnetic resonance imaging (MRI)-guided multileaf collimator (MLC) tracking for 2-fraction prostate radiotherapy on an MR-linac.</div></div><div><h3>Materials and methods:</h3><div>We compared two MRI-guided MLC centroid tracking workflows, each using a different motion manager to derive and stream target positions to our in-house MLC tracking software. The first workflow relies on interleaved 2D (2.5D) cine-MRI, introducing minimal latency. In contrast, the second workflow utilized 3D cine-MRI, which operates at a relatively lower imaging frequency that introduces more latency.</div><div>For experimental validation, we used a motion phantom equipped with an integrated insert that combines film with plastic scintillation dosimetry. A 2x12 Gy 11-beam prostate intensity modulated radiotherapy plan was created for tracking deliveries.</div></div><div><h3>Results:</h3><div>The signal latency introduced by the motion managers was 0.6 s for 2.5D cine-MRI and 6.3 s for 3D cine-MRI. Despite this latency, MLC tracking effectively restored the planned dose, improving the 2%/2mm local gamma pass-rates from 21% (due to linear drift) to 89% (2.5D) and 91% (3D). Plastic scintillator measurements showed reduced dose deviations at the periphery of the clinical target volume from 13–64% (no tracking) to 0–11% (2.5D) and 2–26% (3D).</div></div><div><h3>Conclusion:</h3><div>Our experiments demonstrated the technical feasibility of 2.5D and 3D cine-MRI-based MLC tracking on an MR-linac for 2-fraction prostate radiotherapy, with both motion management strategies achieving comparable dosimetric improvements despite the difference in latency.</div></div>","PeriodicalId":36850,"journal":{"name":"Physics and Imaging in Radiation Oncology","volume":"36 ","pages":"Article 100828"},"PeriodicalIF":3.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145268547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Corrigendum to “Proton dose calculation on cone-beam computed tomography using unsupervised 3D deep learning networks”. [Phys Imaging Radiat Oncol 2024;32:100658] “使用无监督3D深度学习网络计算锥束计算机断层扫描的质子剂量”的勘误表。物理成像辐射学报，2024；32:100658]

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-10-15 DOI: 10.1016/j.phro.2025.100849

Casper Dueholm Vestergaard , Ulrik Vindelev Elstrøm , Ludvig Paul Muren , Jintao Ren , Ole Nørrevang , Kenneth Jensen , Vicki Trier Taasti

引用次数: 0

A phantom study of internal target volume and mid-position accuracy in adaptive and conventional four-dimensional computed tomography across regular and irregular motion 自适应和传统四维计算机断层扫描在规则和不规则运动中的内部靶体积和中间位置精度的模拟研究

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 DOI: 10.1016/j.phro.2025.100845

Bart J.J. Kremers , Dave S.C. van Gruijthuijsen , Dominique Reijtenbagh , Jacco L.G. Steenhuijsen , Mariska de Smet , Rob H.N. Tijssen

This technical note evaluates the performance of an adaptive four-dimensional computed tomography (4DCT) acquisition method compared to conventional 4DCT using a motion phantom. Metrics assessed include deviations in volume, CT number, diameter, peak-to-peak amplitude and determination of the internal target volume (ITV) and mid-position. Under regular breathing, most measurements fall within predefined clinical tolerances for all systems. Under irregular motion, the adaptive method showed reduced deviations in ITV and minimal impact on mid-position determination. These findings support the clinical value of adaptive 4DCT in improving motion management and target definition accuracy in radiotherapy planning.

本技术笔记评估了自适应四维计算机断层扫描（4DCT）采集方法的性能，并与使用运动幻象的传统4DCT进行了比较。评估指标包括体积、CT数、直径、峰对峰振幅的偏差，以及内部目标体积（ITV）和中间位置的确定。在正常呼吸的情况下，大多数测量值都在所有系统预定义的临床公差范围内。在不规则运动情况下，自适应方法显示ITV偏差较小，对中间位置确定的影响最小。这些发现支持了适应性4DCT在改善放射治疗计划中的运动管理和靶标定义准确性方面的临床价值。

引用次数: 0

Synthetic computed tomography techniques for adaptive proton therapy in head and neck cancers 头颈癌适应性质子治疗的合成计算机断层扫描技术

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-10-09 DOI: 10.1016/j.phro.2025.100847

Suryakant Kaushik , Nadine Vatterodt , Jakob Ödén , Albin Fredriksson , Stine S. Korreman , Iuliana Toma-Dasu

Head and neck (HN) radiotherapy often requires corrective interventions. This study evaluated three methods for synthetic computed tomography (CT) generation for adaptive HN planning using cone-beam CT (CBCT) images. CBCT images for 15 patients were paired with same-day repeat CT scans and robustly optimized proton plans were recalculated. The anatomy-preserving virtual CT (APvCT) method utilized organs-at-risk as deformation controlling structures. APvCT and conventional virtual CT methods showed lower mean absolute errors in CT number values compared to corrected CBCT; however, all synthetic CT methods were found suitable for proton dose recalculation with gamma passing rates greater than 96.7% (

2 %, 2 mm

).

头颈部放射治疗通常需要矫正干预。本研究评估了三种使用锥形束CT （CBCT）图像生成自适应HN规划的合成计算机断层扫描（CT）方法。15例患者的CBCT图像与同一天重复CT扫描配对，并重新计算稳健优化的质子计划。保持解剖结构的虚拟CT （APvCT）方法利用危险器官作为变形控制结构。与校正后的CBCT相比，APvCT和传统虚拟CT方法的CT值平均绝对误差更小；然而，所有合成CT方法均适用于质子剂量重新计算，伽玛通过率大于96.7% （2%,2mm）。

引用次数: 0

Risk of genitourinary late effects after radiotherapy for prostate cancer associated with early changes in bladder shape 前列腺癌放疗后泌尿生殖系统晚期效应与早期膀胱形状改变的风险

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-10-31 DOI: 10.1016/j.phro.2025.100855

Oscar Casares-Magaz , Renata G. Raidou , Katarina Furmanová , Niclas Pettersson , Vitali Moiseenko , John Einck , Austin Hopper , Rick Knopp , Ludvig P. Muren

Background and purpose

The risk of genitourinary late effects is a major dose-limiting factor in radiotherapy for prostate cancer. By using shape analysis and machine learning, the aim of this study was to evaluate whether bladder shape descriptors from the first week of treatment could identify patients experiencing genitourinary late effects.

Material and methods

From a cohort of 258 prostate cancer patients treated with daily cone-beam computed tomography (CBCT)-guided radiotherapy (prescription doses of 77.4–81.0 Gy), 7 pre-treatment asymptomatic cases experiencing RTOG genitourinary late effects ≥Grade 2 and 21 matched controls were selected. The bladder was manually contoured on each CBCT, and a 17-D vector comprising shape descriptors was used for patient clustering, focusing on bladder contours from the first week of treatment. ANOVA was used to test statistical significance of descriptors across and within clusters.

Results

Of the contours from the first week of treatment, 84 % could be classified in two main clusters with distinct bladder shape characteristics. This cluster stratification remained identical when bladder contours from the entire course of treatment were used. Convexity, elliptic variance and compactness were significantly different between patients with vs. without genitourinary late effects ≥Grade 2 (p < 0.05). Dice Coefficients between predictive models using descriptors of the first week and the voxels’ probability of belonging to the bladder were above 93 ± 6 % (median ± interquartile range).

Conclusion

Bladder shape descriptors in the first week of treatment showed potential to predict the risk of developing genitourinary late effects after radiotherapy for prostate cancer.

背景与目的泌尿生殖系统晚期效应的风险是前列腺癌放疗的一个主要剂量限制因素。通过使用形状分析和机器学习，本研究的目的是评估从治疗第一周开始的膀胱形状描述符是否可以识别患有泌尿生殖系统晚期效应的患者。材料与方法258例接受每日锥束计算机断层扫描（CBCT）引导放射治疗（处方剂量77.4-81.0 Gy）的前列腺癌患者中，选择7例治疗前出现RTOG晚期效应≥2级的无症状患者和21例匹配对照。在每个CBCT上手动绘制膀胱轮廓，并使用包含形状描述符的17-D向量进行患者聚类，重点关注治疗第一周的膀胱轮廓。方差分析用于检验描述符在集群间和集群内的统计显著性。结果治疗后第一周的膀胱外形特征中，有84%的患者可分为两类。当使用整个治疗过程的膀胱轮廓时，这种簇状分层保持相同。泌尿生殖系统晚期效应≥2级患者与非泌尿生殖系统晚期效应患者的凸度、椭圆方差和致密度差异有统计学意义（p < 0.05）。使用第一周描述符的预测模型与属于膀胱的体素概率之间的骰子系数大于93±6%（中位数±四分位数范围）。结论前列腺癌放疗后第一周膀胱形态描述符可预测患者发生泌尿生殖系统晚期反应的风险。

{"title":"Risk of genitourinary late effects after radiotherapy for prostate cancer associated with early changes in bladder shape","authors":"Oscar Casares-Magaz , Renata G. Raidou , Katarina Furmanová , Niclas Pettersson , Vitali Moiseenko , John Einck , Austin Hopper , Rick Knopp , Ludvig P. Muren","doi":"10.1016/j.phro.2025.100855","DOIUrl":"10.1016/j.phro.2025.100855","url":null,"abstract":"<div><h3>Background and purpose</h3><div>The risk of genitourinary late effects is a major dose-limiting factor in radiotherapy for prostate cancer. By using shape analysis and machine learning, the aim of this study was to evaluate whether bladder shape descriptors from the first week of treatment could identify patients experiencing genitourinary late effects.</div></div><div><h3>Material and methods</h3><div>From a cohort of 258 prostate cancer patients treated with daily cone-beam computed tomography (CBCT)-guided radiotherapy (prescription doses of 77.4–81.0 Gy), 7 pre-treatment asymptomatic cases experiencing RTOG genitourinary late effects ≥Grade 2 and 21 matched controls were selected. The bladder was manually contoured on each CBCT, and a 17-D vector comprising shape descriptors was used for patient clustering, focusing on bladder contours from the first week of treatment. ANOVA was used to test statistical significance of descriptors across and within clusters.</div></div><div><h3>Results</h3><div>Of the contours from the first week of treatment, 84 % could be classified in two main clusters with distinct bladder shape characteristics. This cluster stratification remained identical when bladder contours from the entire course of treatment were used. Convexity, elliptic variance and compactness were significantly different between patients with vs. without genitourinary late effects ≥Grade 2 (<em>p</em> < 0.05). Dice Coefficients between predictive models using descriptors of the first week and the voxels’ probability of belonging to the bladder were above 93 ± 6 % (median ± interquartile range).</div></div><div><h3>Conclusion</h3><div>Bladder shape descriptors in the first week of treatment showed potential to predict the risk of developing genitourinary late effects after radiotherapy for prostate cancer.</div></div>","PeriodicalId":36850,"journal":{"name":"Physics and Imaging in Radiation Oncology","volume":"36 ","pages":"Article 100855"},"PeriodicalIF":3.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145417032","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Clinical target volumes for glioma – Automated delineation to improve neuroanatomic consistency 胶质瘤的临床靶体积-自动划定以提高神经解剖一致性

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-11-12 DOI: 10.1016/j.phro.2025.100865

Gregory Buti , Marcela Giovenco , Tugba Yilmaz , Ali Ajdari , Christopher P. Bridge , Gregory C. Sharp , Fredrik Löfman , Helen A. Shih , Thomas Bortfeld

Background and purpose

Delineating clinical target volumes (CTVs) for glioma is challenging as consistency with the neuroanatomy needs to be carefully verified. We developed an automated approach that incorporates tumor infiltration pathways and anatomic barriers to improve the neuroanatomical consistency and efficiency of CTV delineation.

Materials and methods

A deep learning model for brain structure segmentation was developed based on manual delineations of hemispheres, brainstem, cerebellum, optic chiasm, optic nerves, ventricles, and midline on CT images of ninety-nine glioma patients. Brain structures predictions are integrated into a constrained distance transform that defines the CTV as a 15-mm expansion of the gross tumor volume. Connecting structures with white matter tracts allow for expansions across different structure boundaries, e.g., cerebellum and brainstem connecting at the cerebellar peduncles.

Results

Mean (±std) Dice Similarity Coefficient (DSC) for the hemispheres, brainstem, cerebel-lum, chiasm, optic nerves, midline and ventricles were (98.5 ± 0.8)%, (92.5 ± 2.8)%, (96.7 ± 2.2)% (63.9 ± 12.2)%, (83.8 ± 9.0)%, (81.2 ± 7.0) and (91.5 ± 3.9)%. Mean (±std) 95 % Hausdorff distance (HD95) were, in mm, 1.9 ± 2.5, 7.0 ± 5.4, 1.8 ± 1.2, 7.2 ± 3.2, 2.3 ± 1.0, 9.5 ± 10.5, and 3.8 ± 3.1, respectively. Auto-generated CTVs are compared against reference CTVs (15-mm expansion constrained by manually-contoured brain structures). The automatic CTVs showed excellent similarity to the reference CTVs with mean (±std) Surface DSC with 2 mm tolerance and HD95 scores of (95.6 ± 3.4)% and (1.4 ± 1.2) mm, respectively. A physician’s quality assessment reported that the automated method would result in a substantial amount of time saved in 85 % of CTV delineations.

Conclusion

We have successfully incorporated expert knowledge to improve the neuroanatom-ical consistency of automatically-generated CTVs for glioma.

背景和目的描述胶质瘤的临床靶体积（CTVs）具有挑战性，因为需要仔细验证与神经解剖学的一致性。我们开发了一种结合肿瘤浸润路径和解剖屏障的自动化方法，以提高CTV描绘的神经解剖学一致性和效率。材料与方法基于99例胶质瘤患者CT图像的半脑、脑干、小脑、视交叉、视神经、脑室和中线的人工圈定，建立了脑结构分割的深度学习模型。脑结构预测被整合到一个约束距离变换中，该变换将CTV定义为肿瘤总体积扩大15mm。用白质束连接结构允许跨越不同结构边界的扩张，例如，小脑和脑干在小脑梗处连接。结果脑半球、脑干、小脑、交叉、视神经、中线和脑室的平均（±std） Dice Similarity Coefficient （DSC）分别为（98.5±0.8）%、（92.5±2.8）%、（96.7±2.2）%、（63.9±12.2）%、（83.8±9.0）%、（81.2±7.0）%和（91.5±3.9）%。意味着(±std) 95% (HD95)豪斯多夫距离,在毫米,1.9±2.5,7.0±5.4,1.8±1.2,7.2±3.2,2.3±1.0,9.5±10.5,3.8±3.1,分别。将自动生成的ctv与参考ctv进行比较（受手动轮廓脑结构限制的15毫米扩张）。自动CTVs与参考CTVs具有良好的相似性，平均（±std）表面DSC公差为2mm， HD95评分分别为（95.6±3.4）%和（1.4±1.2）mm。一名医生的质量评估报告称，自动化方法将在85%的CTV划定中节省大量时间。结论我们成功地结合了专家知识，提高了神经胶质瘤自动生成cvs的神经解剖学一致性。

{"title":"Clinical target volumes for glioma – Automated delineation to improve neuroanatomic consistency","authors":"Gregory Buti , Marcela Giovenco , Tugba Yilmaz , Ali Ajdari , Christopher P. Bridge , Gregory C. Sharp , Fredrik Löfman , Helen A. Shih , Thomas Bortfeld","doi":"10.1016/j.phro.2025.100865","DOIUrl":"10.1016/j.phro.2025.100865","url":null,"abstract":"<div><h3>Background and purpose</h3><div>Delineating clinical target volumes (CTVs) for glioma is challenging as consistency with the neuroanatomy needs to be carefully verified. We developed an automated approach that incorporates tumor infiltration pathways and anatomic barriers to improve the neuroanatomical consistency and efficiency of CTV delineation.</div></div><div><h3>Materials and methods</h3><div>A deep learning model for brain structure segmentation was developed based on manual delineations of hemispheres, brainstem, cerebellum, optic chiasm, optic nerves, ventricles, and midline on CT images of ninety-nine glioma patients. Brain structures predictions are integrated into a constrained distance transform that defines the CTV as a 15-mm expansion of the gross tumor volume. Connecting structures with white matter tracts allow for expansions across different structure boundaries, e.g., cerebellum and brainstem connecting at the cerebellar peduncles.</div></div><div><h3>Results</h3><div>Mean (±std) Dice Similarity Coefficient (DSC) for the hemispheres, brainstem, cerebel-lum, chiasm, optic nerves, midline and ventricles were (98.5 ± 0.8)%, (92.5 ± 2.8)%, (96.7 ± 2.2)% (63.9 ± 12.2)%, (83.8 ± 9.0)%, (81.2 ± 7.0) and (91.5 ± 3.9)%. Mean (±std) 95 % Hausdorff distance (HD95) were, in mm, 1.9 ± 2.5, 7.0 ± 5.4, 1.8 ± 1.2, 7.2 ± 3.2, 2.3 ± 1.0, 9.5 ± 10.5, and 3.8 ± 3.1, respectively. Auto-generated CTVs are compared against reference CTVs (15-mm expansion constrained by manually-contoured brain structures). The automatic CTVs showed excellent similarity to the reference CTVs with mean (±std) Surface DSC with 2 mm tolerance and HD95 scores of (95.6 ± 3.4)% and (1.4 ± 1.2) mm, respectively. A physician’s quality assessment reported that the automated method would result in a substantial amount of time saved in 85 % of CTV delineations.</div></div><div><h3>Conclusion</h3><div>We have successfully incorporated expert knowledge to improve the neuroanatom-ical consistency of automatically-generated CTVs for glioma.</div></div>","PeriodicalId":36850,"journal":{"name":"Physics and Imaging in Radiation Oncology","volume":"36 ","pages":"Article 100865"},"PeriodicalIF":3.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145576303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Workflow to detect exceeded dose constraints in pancreatic stereotactic body irradiation after intrafraction motion during magnetic resonance-guided adaptive radiotherapy using a deep learning-refined contour propagation tool 使用深度学习-精细轮廓传播工具检测磁共振引导自适应放射治疗中胰腺立体定向照射中超过剂量限制的工作流程

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-11-13 DOI: 10.1016/j.phro.2025.100866

Christina Sarosiek, Asma Amjad, Renae Conlin, Beth Erickson, William A. Hall, Eric S. Paulson

Introduction

The adapt-to-shape (ATS) process on the MR-Linac involves manual contour edits followed by treatment plan re-optimization on daily pre-beam MRIs. A verification image is acquired after plan optimization to assess the dose distribution with respect to intrafraction motion using the pre-beam contours. We introduce here a workflow to automatically detect organ motion of gastrointestinal structures that results in exceeded planned dose constraints.

Materials and methods

The workflow first transferred the contours and dose distribution created on the daily pre-beam MRI to the verification MRI. A deep learning-refined contour propagation (DL-RCP) tool, trained on 79 images, improved the transferred contours and the dose to 0.03 cm³ (D_{0.03 cm³}) is updated. The workflow notified the physician if D_{0.03 cm³} exceeds the constraint. We tested the workflow on 48 daily ATS fractions of 11 patients treated for pancreatic cancer (33–40 Gy in 5 fractions). We added manually drawn contours to the verification images for reference.

Results

The Dice similarity coefficient and mean distance to agreement for the clinical/DL-refined contours were 0.56/0.71 and 9.08/5.08 mm, respectively. The workflow detected exceeded constraints with specificity 0.90, sensitivity 0.75, and accuracy 0.85. In one case, the duodenal D_{0.03 cm³} was 29.9 Gy for the clinical contour, and 36.0 Gy and 35.6 Gy with the reference and DL-refined contours.

Conclusion

The proposed method detected exceeded dose constraints in gastrointestinal structures due to intrafraction motion during ATS planning for pancreatic treatments and can aid in the clinical decision to re-optimize the plan on the verification MR.

MR-Linac上的形状适应（ATS）过程包括手动轮廓编辑，然后对每日预束mri进行治疗计划重新优化。在计划优化后获得验证图像，以利用预束轮廓评估相对于屈光度运动的剂量分布。我们在这里介绍一个工作流程来自动检测胃肠道结构的器官运动，导致超过计划的剂量限制。材料和方法该工作流程首先将每日预束MRI上生成的轮廓和剂量分布转移到验证MRI上。使用深度学习-精细轮廓传播（DL-RCP）工具对79张图像进行训练，改进了转移轮廓，并更新了剂量至0.03 cm3 （D0.03 cm3）。如果D0.03 cm3超过限制，工作流通知医生。我们对11例胰腺癌患者的48个每日ATS分数（5个分数33-40 Gy）进行了工作流程测试。我们将手工绘制的轮廓添加到验证图像中以供参考。结果临床轮廓与dl轮廓的Dice相似系数和平均吻合距离分别为0.56/0.71和9.08/5.08 mm。检测到的工作流程超出约束，特异性0.90，灵敏度0.75，准确性0.85。1例十二指肠D0.03 cm3临床轮廓为29.9 Gy，参考轮廓和dl -精轮廓分别为36.0 Gy和35.6 Gy。结论本方法在胰腺治疗ATS计划过程中可检测到胃肠道结构中由于牵引力运动导致的剂量超出限制，可在验证MR上帮助临床决策重新优化计划。

{"title":"Workflow to detect exceeded dose constraints in pancreatic stereotactic body irradiation after intrafraction motion during magnetic resonance-guided adaptive radiotherapy using a deep learning-refined contour propagation tool","authors":"Christina Sarosiek, Asma Amjad, Renae Conlin, Beth Erickson, William A. Hall, Eric S. Paulson","doi":"10.1016/j.phro.2025.100866","DOIUrl":"10.1016/j.phro.2025.100866","url":null,"abstract":"<div><h3>Introduction</h3><div>The adapt-to-shape (ATS) process on the MR-Linac involves manual contour edits followed by treatment plan re-optimization on daily pre-beam MRIs. A verification image is acquired after plan optimization to assess the dose distribution with respect to intrafraction motion using the pre-beam contours. We introduce here a workflow to automatically detect organ motion of gastrointestinal structures that results in exceeded planned dose constraints.</div></div><div><h3>Materials and methods</h3><div>The workflow first transferred the contours and dose distribution created on the daily pre-beam MRI to the verification MRI. A deep learning-refined contour propagation (DL-RCP) tool, trained on 79 images, improved the transferred contours and the dose to 0.03 cm<sup>3</sup> (D<sub>0.03 cm<sup>3</sup></sub>) is updated. The workflow notified the physician if D<sub>0.03 cm<sup>3</sup></sub> exceeds the constraint. We tested the workflow on 48 daily ATS fractions of 11 patients treated for pancreatic cancer (33–40 Gy in 5 fractions). We added manually drawn contours to the verification images for reference.</div></div><div><h3>Results</h3><div>The Dice similarity coefficient and mean distance to agreement for the clinical/DL-refined contours were 0.56/0.71 and 9.08/5.08 mm, respectively. The workflow detected exceeded constraints with specificity 0.90, sensitivity 0.75, and accuracy 0.85. In one case, the duodenal D<sub>0.03 cm<sup>3</sup></sub> was 29.9 Gy for the clinical contour, and 36.0 Gy and 35.6 Gy with the reference and DL-refined contours.</div></div><div><h3>Conclusion</h3><div>The proposed method detected exceeded dose constraints in gastrointestinal structures due to intrafraction motion during ATS planning for pancreatic treatments and can aid in the clinical decision to re-optimize the plan on the verification MR.</div></div>","PeriodicalId":36850,"journal":{"name":"Physics and Imaging in Radiation Oncology","volume":"36 ","pages":"Article 100866"},"PeriodicalIF":3.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145576372","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Automated target misalignment correction for cone beam computed tomography-based online adaptive radiotherapy of locally advanced lung cancer patients 锥形束ct在线自适应放疗对局部晚期肺癌患者靶位失调的自动校正

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-12-03 DOI: 10.1016/j.phro.2025.100885

Carmen Seller-Oria, Suzanne van Beek, Suzan Gerrets, Sanne van Weerdenburg, Paula Bos, Simon van Kranen, Marloes Frantzen-Steneker, Zeno Gouw, Jan-Jakob Sonke, Peter Remeijer

Background and purpose

Locally advanced lung cancer patients are commonly treated with daily cone beam CT (CBCT) guided radiotherapy using one treatment isocenter. Due to differential motion between primary tumor (GTV_prim) and affected lymph nodes, a compromise needs to be made during daily patient alignment, requiring enlarged treatment margins. In this work, an online adaptive (OART) strategy was proposed to correct for residual target misalignments and enable treatment margin reduction.

Material and methods

We developed in-house an application that produced a synthetic CT (sCT) and delineations to correct for residual target misalignments. A deformation vector field (DVF) was created by using conventional CBCT-to-CT rigid target registrations. The DVF was applied to the planning CT (pCT) and delineations to generate a sCT where GTV_prim was loco-rigidly shifted into the correct position. Twenty CBCTs of eight patients were selected to assess sCTs in terms of GTV_prim position (via CBCT-to-sCT and CBCT-to-pCT registration vector lengths), pixel-wise sCT-pCT Hounsfield unit (HU) errors inside GTV_prim, and sCT-pCT GTV_prim volume differences.

Results

Median vector lengths were 5.1 mm relative to pCTs, and 0.7 mm relative to sCTs, demonstrating the ability of the proposed tool to correct residual misalignments. Median HU errors across all scans were within 1 HU, and the median GTV_prim volume difference was −3.7 %.

Conclusions

A correction method for residual target misalignments in locally advanced lung cancer patients was proposed. It automatically produces sCTs and delineations, enabling OART implementation without the need for manual delineation corrections, and with potentially smaller treatment margins.

背景与目的局部晚期肺癌患者通常采用每日锥形束CT （CBCT）引导的单中心放射治疗。由于原发肿瘤（GTVprim）和受影响淋巴结之间的差异运动，需要在日常患者对齐时做出妥协，需要扩大治疗范围。在这项工作中，提出了一种在线自适应（OART）策略来纠正残留的靶标错位并使治疗余量减小。材料和方法我们在内部开发了一个应用程序，该应用程序产生合成CT （sCT）和圈定来纠正残留的目标错位。采用常规的cbct - ct刚性目标配准方法建立了变形向量场（DVF）。DVF应用于规划CT （pCT）和描绘生成sCT，其中GTVprim局部刚性移位到正确位置。选择8例患者的20个cbct来评估GTVprim位置（通过CBCT-to-sCT和CBCT-to-pCT注册向量长度）、GTVprim内像素级sCT-pCT Hounsfield单位（HU）误差以及sCT-pCT GTVprim体积差异。结果：相对于pct，中位矢量长度为5.1 mm，相对于sct，中位矢量长度为0.7 mm，表明所提出的工具有能力纠正剩余的不对中。所有扫描的HU误差中位数在1 HU以内，GTVprim体积差中位数为- 3.7%。结论提出了局部晚期肺癌患者残留靶位失调的校正方法。它可以自动生成sct和圈定，无需手动圈定校正就可以实现OART，并且潜在的处理余量更小。

{"title":"Automated target misalignment correction for cone beam computed tomography-based online adaptive radiotherapy of locally advanced lung cancer patients","authors":"Carmen Seller-Oria, Suzanne van Beek, Suzan Gerrets, Sanne van Weerdenburg, Paula Bos, Simon van Kranen, Marloes Frantzen-Steneker, Zeno Gouw, Jan-Jakob Sonke, Peter Remeijer","doi":"10.1016/j.phro.2025.100885","DOIUrl":"10.1016/j.phro.2025.100885","url":null,"abstract":"<div><h3>Background and purpose</h3><div>Locally advanced lung cancer patients are commonly treated with daily cone beam CT (CBCT) guided radiotherapy using one treatment isocenter. Due to differential motion between primary tumor (GTV<sub>prim</sub>) and affected lymph nodes, a compromise needs to be made during daily patient alignment, requiring enlarged treatment margins. In this work, an online adaptive (OART) strategy was proposed to correct for residual target misalignments and enable treatment margin reduction.</div></div><div><h3>Material and methods</h3><div>We developed in-house an application that produced a synthetic CT (sCT) and delineations to correct for residual target misalignments. A deformation vector field (DVF) was created by using conventional CBCT-to-CT rigid target registrations. The DVF was applied to the planning CT (pCT) and delineations to generate a sCT where GTV<sub>prim</sub> was loco-rigidly shifted into the correct position. Twenty CBCTs of eight patients were selected to assess sCTs in terms of GTV<sub>prim</sub> position (via CBCT-to-sCT and CBCT-to-pCT registration vector lengths), pixel-wise sCT-pCT Hounsfield unit (HU) errors inside GTV<sub>prim</sub>, and sCT-pCT GTV<sub>prim</sub> volume differences.</div></div><div><h3>Results</h3><div>Median vector lengths were 5.1 mm relative to pCTs, and 0.7 mm relative to sCTs, demonstrating the ability of the proposed tool to correct residual misalignments. Median HU errors across all scans were within 1 HU, and the median GTV<sub>prim</sub> volume difference was −3.7 %.</div></div><div><h3>Conclusions</h3><div>A correction method for residual target misalignments in locally advanced lung cancer patients was proposed. It automatically produces sCTs and delineations, enabling OART implementation without the need for manual delineation corrections, and with potentially smaller treatment margins.</div></div>","PeriodicalId":36850,"journal":{"name":"Physics and Imaging in Radiation Oncology","volume":"36 ","pages":"Article 100885"},"PeriodicalIF":3.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145680903","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Imaging, automation and workflow challenges for online adaptive proton therapy 在线自适应质子治疗的成像、自动化和工作流程挑战

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-11-24 DOI: 10.1016/j.phro.2025.100875

Dirk Wagenaar , Francesca Albertini , Ludvig P. Muren , Barbara Knäusl

引用次数: 0

Using three-dimensional equieffective dose mapping to audit a methodology for calculating permitted doses for head and neck reirradiation 使用三维等有效剂量图审核头颈部再照射允许剂量的计算方法

IF 3.3 Q2 ONCOLOGY

Physics and Imaging in Radiation Oncology

Pub Date : 2025-10-01 Epub Date: 2025-11-17 DOI: 10.1016/j.phro.2025.100867

David Nash , Sarah Muscat , Chelmis M. Thiong’o , Eliana Vasquez Osorio , Antony L. Palmer

Background and purpose

Reirradiation is increasingly common, but accounting for prior dose to the spinal cord and brainstem is challenging. This study compares a simple isodose-based method of previous dose compensation with voxel-wise equivalent dose in 2 Gy fractions (EQD_2/2) dose mapping to assess accuracy and safety.

Materials and methods

Ten head and neck reirradiation cases were retrospectively reviewed. During planning, original dose distributions were mapped onto the reirradiation computed tomography scan. Isodoses within the cord and brainstem were used to segment dose-level substructures, with cumulative doses kept within defined tolerances. Following 3D EQD_2/2 mapping, the same cases were audited by recalculating cumulative maximum doses. Cumulative doses from both methods were compared.

Results

Retrospective EQD_2/2 analysis confirmed all cumulative doses were within tolerance, with up to 9.7 Gy difference between the two methods. In two cases, cord and brainstem doses approached tolerance.

Conclusions

A simple, isodose-based approach to spinal cord and brainstem reirradiation tolerance calculation has been shown to be safe when retrospectively compared with voxel-wise EQD_2/2 mapping. This method can be implemented in any planning system capable of generating contours from isodose lines, offering a practical alternative where advanced EQD_2/2 dose accumulation software is unavailable.

背景和目的放射治疗越来越普遍，但对脊髓和脑干的先前剂量的解释是具有挑战性的。本研究比较了一种简单的基于异剂量的先前剂量补偿方法与2 Gy分数（EQD2/2）剂量测绘的体素等效剂量，以评估准确性和安全性。材料与方法对10例头颈部再照射病例进行回顾性分析。在计划期间，将原始剂量分布映射到再照射计算机断层扫描上。脊髓和脑干内的等剂量用于分割剂量水平的亚结构，累积剂量保持在规定的耐受范围内。在3D EQD2/2制图后，通过重新计算累积最大剂量对相同病例进行审计。比较了两种方法的累积剂量。结果回顾性EQD2/2分析证实所有累积剂量均在耐受范围内，两种方法之间的差异高达9.7 Gy。在两个病例中，脊髓和脑干剂量接近耐受。结论与EQD2/2体素制图相比，一种简单的、基于异剂量的脊髓和脑干再照射耐受计算方法是安全的。该方法可以在任何能够从等剂量线生成等高线的规划系统中实施，在没有先进的EQD2/2剂量累积软件的情况下提供实用的替代方案。

{"title":"Using three-dimensional equieffective dose mapping to audit a methodology for calculating permitted doses for head and neck reirradiation","authors":"David Nash , Sarah Muscat , Chelmis M. Thiong’o , Eliana Vasquez Osorio , Antony L. Palmer","doi":"10.1016/j.phro.2025.100867","DOIUrl":"10.1016/j.phro.2025.100867","url":null,"abstract":"<div><h3>Background and purpose</h3><div>Reirradiation is increasingly common, but accounting for prior dose to the spinal cord and brainstem is challenging. This study compares a simple isodose-based method of previous dose compensation with voxel-wise equivalent dose in 2 Gy fractions (EQD<sub>2/2</sub>) dose mapping to assess accuracy and safety.</div></div><div><h3>Materials and methods</h3><div>Ten head and neck reirradiation cases were retrospectively reviewed. During planning, original dose distributions were mapped onto the reirradiation computed tomography scan. Isodoses within the cord and brainstem were used to segment dose-level substructures, with cumulative doses kept within defined tolerances. Following 3D EQD<sub>2/2</sub> mapping, the same cases were audited by recalculating cumulative maximum doses. Cumulative doses from both methods were compared.</div></div><div><h3>Results</h3><div>Retrospective EQD<sub>2/2</sub> analysis confirmed all cumulative doses were within tolerance, with up to 9.7 Gy difference between the two methods. In two cases, cord and brainstem doses approached tolerance.</div></div><div><h3>Conclusions</h3><div>A simple, isodose-based approach to spinal cord and brainstem reirradiation tolerance calculation has been shown to be safe when retrospectively compared with voxel-wise EQD<sub>2/2</sub> mapping. This method can be implemented in any planning system capable of generating contours from isodose lines, offering a practical alternative where advanced EQD<sub>2/2</sub> dose accumulation software is unavailable.</div></div>","PeriodicalId":36850,"journal":{"name":"Physics and Imaging in Radiation Oncology","volume":"36 ","pages":"Article 100867"},"PeriodicalIF":3.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0