Medical physics最新文献

{"title":"AAPM Spring Clinical Meeting - Program.","authors":"","doi":"10.1002/mp.70459","DOIUrl":"https://doi.org/10.1002/mp.70459","url":null,"abstract":"","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70459"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147825219","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}

{"title":"Prediction of VMAT gamma passing rates using 3D CNNs based on leaf position analysis and gradient class activation mapping for plan complexity evaluation.","authors":"Johannes Berchtold, Sara Vockner, Ivan Messner, Markus Stana, Falk Röder, Frank Wolf, Christoph Gaisberger","doi":"10.1002/mp.70468","DOIUrl":"https://doi.org/10.1002/mp.70468","url":null,"abstract":"<p><strong>Background: </strong>Volumetric Modulated Arc Therapy (VMAT) is a highly conformal radiotherapy technique that enables precise tumor irradiation while sparing surrounding healthy tissue. However, the high technical demands this technique places on Linear Accelerators (LinAc) necessitate reliable quality assurance (QA) tools. The Gamma Passing Rate (GPR), commonly used to compare planned and delivered dose distributions, requires extensive measurement resources. Many existing predictive metrics, such as the popular Modulation Complexity Score (MCS), are independent of the beam model, limiting their accuracy. Consequently, identifying appropriate metrics and their individual thresholds can be challenging.</p><p><strong>Purpose: </strong>This study aims to predict the GPR of VMAT arcs using a three-dimensional convolutional neural network (3D CNN). Gradient-weighted Class Activation Mapping (Grad-CAM) is applied to improve interpretability, identify weak segments, and potentially reveal beam model limitations.</p><p><strong>Methods: </strong>A 3D CNN was trained on 140 6 MV VMAT arcs, with 30 arcs each used for validation and testing. All plans were delivered by an Elekta Harmony Pro LinAc with 4° control point (CP) spacing. Input data included discretized beam's eye view (BEV) representations and segment-specific monitor unit (MU) values. GPR evaluation was performed using a Delta4+ phantom with a 1%/ 2 mm criterion. Data augmentation enhanced training diversity. Grad-CAM was used to visualize influential plan regions.</p><p><strong>Results: </strong>After 36 epochs, the model achieved a mean absolute error (MAE) of 2.0%(test set) and 1.3%(training set). With cropped input, the best MAEs were 2.1%(test) and 1.5%(training). Grad-CAM analysis indicated that dynamic delivery aspects had more influence on prediction accuracy than static features like field shape.</p><p><strong>Conclusions: </strong>This study highlights the potential of deep learning for automated GPR prediction, offering a more efficient QA workflow. Especially in time-critical settings like online adaptive radiotherapy, where traditional measurement-based QA is often impractical, this model provides a scalable solution to ensure treatment safety. The use of Grad-CAM enables insight into beam model and LinAc performance, allowing refinement of treatment planning and improved QA precision in clinical practice.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70468"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13143553/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147848223","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-time, high-precision detection of aortic dissection key components for emergency triage using an enhanced deep learning model.","authors":"Houde Wu, Fei Chen, Ruiqing Xia, Longshuang Wang, Zelong Zhang, Jinghui Zhang, Qingzhao Cheng, Li Guo","doi":"10.1002/mp.70461","DOIUrl":"https://doi.org/10.1002/mp.70461","url":null,"abstract":"<p><strong>Background: </strong>Aortic dissection (AD) is a life-threatening cardiovascular emergency. For Type B AD (TBAD), rapid and accurate identification of the true lumen (TL), false lumen (FL), and false lumen thrombus (FLT) from CTA is critical for risk stratification and treatment planning. However, existing deep learning solutions often lack real-time capability and fail to address the detection of FLT.</p><p><strong>Purpose: </strong>To develop a real-time, high-precision deep learning framework for the simultaneous detection of all three key AD components to support emergency triage.</p><p><strong>Methods: </strong>We propose AD-YOLO11, an enhanced YOLOv11 model integrating three key innovations: (1) a Recursive Information Distillation Network (RIDNet) for CTA noise suppression, (2) a Triplet Attention Mechanism for spatial and channel feature enhancement, and (3) the MPDIoU loss function for optimized bounding box regression. The model was trained and internally validated on a dataset of 25 176 slices from 106 TBAD patients and externally validated on 18 238 slices from 71 independent patients.</p><p><strong>Results: </strong>On internal validation, AD-YOLO11 achieved a precision of 0.991 ± 0.004, recall of 0.936 ± 0.006, mAP@0.5 of 0.951 ± 0.007, and mAP@0.5:0.95 of 0.883 ± 0.008. It maintained high performance on the external test set, demonstrating strong generalizability. The inference speed was 3.18 ± 0.23 ms per slice on GPU, and it remained clinically feasible on CPU (53.15 ± 2.76 ms per slice).</p><p><strong>Conclusions: </strong>AD-YOLO11 achieves millisecond-level, high-accuracy detection of all three critical Type B aortic dissection components from CTA images. Its efficient inference on both GPU and CPU makes it a promising frontline tool for rapid triage in emergency and resource-limited settings, effectively complementing time-consuming 3D segmentation for aortic dissection assessment.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70461"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147871062","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}

{"title":"Correction to \"Lightweight deep training network for lymph nodes segmentation from head and neck CT images\".","authors":"","doi":"10.1002/mp.70453","DOIUrl":"https://doi.org/10.1002/mp.70453","url":null,"abstract":"","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70453"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147793236","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}

{"title":"A Hybrid-Mamba network with dual level attention fusion for multimodal COPD diagnosis.","authors":"Feifan Zhang, Dinghui Wu, Shuguang Han, Hao Wang, Wenlong Zhang, Chenhui Ma","doi":"10.1002/mp.70393","DOIUrl":"https://doi.org/10.1002/mp.70393","url":null,"abstract":"<p><strong>Background: </strong>Chronic obstructive pulmonary disease (COPD) remains difficult to diagnose reliably due to limitations of conventional spirometry and CT interpretation. Although deep learning has shown promise, CNNs are constrained by local receptive fields, and ViTs are constrained by high computational cost, highlighting the relevance of multimodal integration of CT and clinical data for improving COPD diagnostic accuracy.</p><p><strong>Purpose: </strong>In this study, we propose a Hybrid-Mamba Network with Dual Level Attention Fusion for multimodal COPD diagnosis.</p><p><strong>Methods: </strong>We retrospectively enrolled 381 participants (184 with COPD and 197 healthy controls). Clinical data encompassed basic information, respiratory symptoms, blood gas analysis, pulmonary function tests, and blood routine tests. The framework employs a Hybrid-Mamba architecture for efficient CT feature representation, leverages a tailored Hybrid-DWConv-AAS Block for enhanced feature integration, and incorporates a Dual Level Attention Fusion Block to adaptively integrate CT and clinical data.</p><p><strong>Results: </strong>On the test set, our proposed network achieved an AUC of 0.985 and an accuracy of 0.947, with an average per-patient inference time of 97.54 ms, while maintaining robust diagnostic performance under simulated perturbations.</p><p><strong>Conclusions: </strong>These findings indicate that the framework provides an efficient and robust approach for COPD diagnosis.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70393"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147793519","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}

{"title":"Spectral deep learning-based patient and bowtie scatter correction for clinical photon-counting CT.","authors":"Lukas Hennemann, Julien Erath, Andreas Heinkele, Eric Fournié, Martin Petersilka, Karl Stierstorfer, Marc Kachelrieß","doi":"10.1002/mp.70442","DOIUrl":"10.1002/mp.70442","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;The presence of scatter in computed tomography degrades image quality, and can be caused by the patient and by other components in the beam path, such as the bowtie filter. While conventional energy-integrating detectors do not provide spectral distinction, photon-counting (PC) detectors are energy-selective and provide spectral information about the incoming X-ray photons. Since each energy threshold is affected differently by scatter, this spectral information implicitly encodes the scatter content of a projection.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;The purpose of this work is to investigate how the spectral information can be exploited to improve deep learning (DL)-based scatter correction. Furthermore, the performance of joint and separate patient and bowtie scatter correction will be investigated, addressing that bowtie scatter has not been considered in current DL-based approaches.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;We present a DL-based approach that can estimate bowtie and patient scatter jointly and compare it against a separate correction. We also introduce neural network-based methods that incorporate the spectral information inherent in PCCT for scatter correction. We present networks that estimate scatter for up to four energy thresholds simultaneously. Training and validation was performed with Monte Carlo data as well as with real data measured by a clinical PCCT system.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;When comparing joint and separate patient and bowtie scatter estimation, both methods reduce the mean absolute error (MAE) from 8 HU to 1 HU. All proposed DSE methods effectively reduce scatter artifacts and perform better than the convolution-based reference approach. Incorporating the spectral information further improves the performance, with the DSE variant with four energy thresholds achieving the best overall results for all thresholds. For all energy thresholds tested, the spectral DSE methods reduced scatter errors originating from the patient and the bowtie in PCCT from up to 8 HU to below 1 HU. In addition to the global MAE, we report a critical MAE (MAE&lt;sub&gt;10&lt;/sub&gt;) restricted to voxels with uncorrected errors &lt;math&gt;&lt;semantics&gt;&lt;mo&gt;&gt;&lt;/mo&gt; &lt;annotation&gt;$&gt;$&lt;/annotation&gt;&lt;/semantics&gt; &lt;/math&gt; 10 HU, as such deviations are visually perceptible in soft tissue and exceed the noise level of modern CT systems. In all test cases, the proposed spectral methods reduced the MAE&lt;sub&gt;10&lt;/sub&gt; from 23.8 HU in the uncorrected images to 1.6 HU after spectral correction. The affected voxels comprised on average 25 % of the image volume, indicating a significant reduction in artifact intensity in the most affected areas. In virtual monoenergetic images (VMI), the application of spectral neural networks resulted in a significant reduction in MAE from &lt;math&gt;&lt;semantics&gt;&lt;mo&gt;∼&lt;/mo&gt; &lt;annotation&gt;$sim$&lt;/annotation&gt;&lt;/semantics&gt; &lt;/math&gt; 16 HU to &lt;math&gt;&lt;semantics&gt;&lt;mo&gt;∼&lt;/mo&gt; &lt;annotation&gt;$sim$&lt;/annotation&gt;&lt;/semantics&gt; &lt;/math&gt; 2 HU ","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70442"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13125421/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147793336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A modified lethal-potentially lethal model for oxygen-mediated FLASH sparing in stem cell niches.","authors":"Sumin Zhou","doi":"10.1002/mp.70469","DOIUrl":"https://doi.org/10.1002/mp.70469","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Ultra-high dose rate (FLASH) irradiation can reduce normal-tissue toxicity while preserving tumor control, but a mechanistic explanation consistent with classical radiobiology remains incomplete. In particular, oxygen-depletion arguments based solely on bulk tissue oxygenation can appear inconsistent with clinically relevant fraction sizes, motivating a DNA-target-level oxygen formulation.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;To develop a theory-based mechanistic extension of the Lethal and Potentially Lethal (LPL) model that explains oxygen-mediated FLASH trends without prescribing dose rate-dependent radiosensitivity, and to identify the baseline nuclear oxygen window in which sparing is expected to be largest.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;We introduce an explicit Precursor Lesion population whose fate is governed by competing chemical restitution/repair versus oxygen-dependent fixation into potentially lethal and lethal lesion channels. Fixation kinetics are coupled to a time-varying nuclear oxygen tension, &lt;math&gt; &lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;p&lt;/mi&gt; &lt;msub&gt;&lt;mi&gt;O&lt;/mi&gt; &lt;mn&gt;2&lt;/mn&gt;&lt;/msub&gt; &lt;mrow&gt;&lt;mo&gt;(&lt;/mo&gt; &lt;mi&gt;t&lt;/mi&gt; &lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt; &lt;/mrow&gt; &lt;annotation&gt;$p{{O}_2}( t )$&lt;/annotation&gt;&lt;/semantics&gt; &lt;/math&gt; , which decreases via radiolytic depletion during irradiation and recovers toward a baseline via reduced-order reoxygenation kinetics. To address the oxygen paradox, we distinguish bulk vascular oxygenation from a lower effective DNA target-level oxygenation that may arise in regulated stem-cell niches because of niche hypoxia and intracellular oxygen consumption. Oxygen modulation is implemented through a mechanistic exponential OER formulation parameterized by an oxygen-fixation rate constant, while retaining classical LPL behavior in the conventional low-dose-rate limit.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;The model predicts that oxygen-mediated FLASH sparing is largest when baseline nuclear oxygenation lies in an intermediate physiologic-hypoxia regime, corresponding to the steep oxygen-responsive portion of the OER curve. In the reference parameter set, a quiescent normal-tissue niche with baseline &lt;math&gt; &lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;p&lt;/mi&gt; &lt;msub&gt;&lt;mi&gt;O&lt;/mi&gt; &lt;mrow&gt;&lt;mn&gt;2&lt;/mn&gt; &lt;mo&gt;,&lt;/mo&gt; &lt;mn&gt;0&lt;/mn&gt;&lt;/mrow&gt; &lt;/msub&gt; &lt;/mrow&gt; &lt;annotation&gt;$p{{O}_{2,0}}$&lt;/annotation&gt;&lt;/semantics&gt; &lt;/math&gt;  = 3 mmHg shows appreciable sparing under FLASH delivery, whereas sparing is minimal when the baseline lies near either the OER floor ( &lt;math&gt; &lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;p&lt;/mi&gt; &lt;msub&gt;&lt;mi&gt;O&lt;/mi&gt; &lt;mrow&gt;&lt;mn&gt;2&lt;/mn&gt; &lt;mo&gt;,&lt;/mo&gt; &lt;mn&gt;0&lt;/mn&gt;&lt;/mrow&gt; &lt;/msub&gt; &lt;/mrow&gt; &lt;annotation&gt;$p{{O}_{2,0}}$&lt;/annotation&gt;&lt;/semantics&gt; &lt;/math&gt;  = 0.2 mmHg) or the OER saturation plateau ( &lt;math&gt; &lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;p&lt;/mi&gt; &lt;msub&gt;&lt;mi&gt;O&lt;/mi&gt; &lt;mrow&gt;&lt;mn&gt;2&lt;/mn&gt; &lt;mo&gt;,&lt;/mo&gt; &lt;mn&gt;0&lt;/mn&gt;&lt;/mrow&gt; &lt;/msub&gt; &lt;/mrow&gt; &lt;annotation&gt;$p{{O}_{2,0}}$&lt;/annotation&gt;&lt;/semantics&gt; &lt;/math&gt;  = 30 mmHg). Sensitivity analyses preserve this intermediate oxygen window while shifting the magnitude and threshold of the effect.&lt;/p&gt;&lt;p&gt;&lt;str","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70469"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147871067","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}

{"title":"Ion recombination correction in reference dosimetry for pencil beam scanned proton beams.","authors":"Jun Ken Gan, Kah Seng Lew, Clifford Ghee Ann Chua, Calvin Wei Yang Koh, Kang Hao Lee, Masashi Yagi, Wen Siang Lew, James Cheow Lei Lee, Sung Yong Park, Hong Qi Tan","doi":"10.1002/mp.70466","DOIUrl":"https://doi.org/10.1002/mp.70466","url":null,"abstract":"<p><strong>Background: </strong>The 2024 IAEA TRS-398 revision updated recommendations for reference dosimetry and ion recombination corrections in pencil beam scanned (PBS) proton beams.</p><p><strong>Purpose: </strong>This study evaluates the revised ion recombination methods for monoenergetic synchrotron-based PBS proton system across different energies, monitor units (MU), and ionization chamber types.</p><p><strong>Methods: </strong>Reference-field measurements were performed using a synchrotron system at 70.2, 150.2, and 228.7 MeV and at 6, 50, and 200 MU. Charge-collection data were acquired using PTW Farmer and Advanced Markus chambers across 20-400 V. Ion recombination correction factors ( <math> <semantics><msub><mi>k</mi> <mi>s</mi></msub> <annotation>${{{bm{k}}}_{bm{s}}}$</annotation></semantics> </math> ) were determined using the Jaffé plot extrapolation method and the TRS-398 two-voltage method (TVM) under different time structure assumptions. Charge multiplication in the chamber was addressed using both low voltage linear fitting and a semiempirical exponential model.</p><p><strong>Results: </strong>For low energy, low MU fields, and TVM yielded <math> <semantics><msub><mi>k</mi> <mi>s</mi></msub> <annotation>${{{bm{k}}}_{bm{s}}}$</annotation></semantics> </math> values within ∼1% of the Jaffé extrapolation. For high-energy, high-MU fields, maximum differences of 6.25% (Farmer) and 1.62% (Advanced Markus) were observed. The synchrotron beam exhibited energy, MU, and chamber dependent time structure behavior, producing pulsed-like or continuous-like characteristics. Misclassification of the time structure resulted in additional <math> <semantics><msub><mi>k</mi> <mi>s</mi></msub> <annotation>${{{bm{k}}}_{bm{s}}}$</annotation></semantics> </math> deviations of up to 2.49% (Farmer) and 0.59% (Advanced Markus). Charge multiplication was observed in the Advanced Markus chamber at voltages > 150 V. The exponential fitting successfully modeled this response and produced <math> <semantics><msub><mi>k</mi> <mi>s</mi></msub> <annotation>${{{bm{k}}}_{bm{s}}}$</annotation></semantics> </math> values agreeing with low voltage fits within 1.5%, while avoiding subjective voltage cutoff selection.</p><p><strong>Conclusion: </strong>The revised TRS-398 provides accurate ion recombination corrections for monoenergetic PBS fields at low energies and low MU. However, accuracy of ion recombination correction decreases at higher energies and MU, particularly when the time structure was ambiguous or chamber dependent. Charge multiplication in small volume chambers presents an additional source of uncertainty not fully addressed by TRS-398. Incorporating charge multiplication fitting methods may improve the robustness of reference dosimetry in synchrotron-based PBS proton therapy.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70466"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147871070","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}

{"title":"Automated extraction of the plane of minimal hiatal dimensions and mid-sagittal plane from 3D transperineal ultrasound.","authors":"Zachary Szentimrey, Golafsoun Ameri, Christopher X Hong, R Y K Cheung, Ahmed Eltahawi, Eranga Ukwatta","doi":"10.1002/mp.70473","DOIUrl":"https://doi.org/10.1002/mp.70473","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Transperineal ultrasound (TPUS) is a valuable imaging tool for evaluating patients with a variety of pelvic floor disorders, including pelvic organ prolapse (POP). Currently, calculating measurements of anatomical structures and relationships as well as extracting the mid-sagittal (MS) plane of 2D and 3D ultrasound images are obtained manually, which is a time-consuming process and requires a reviewer with prior training in pelvic floor US interpretation. The need for manual analysis of ultrasound images has limited the broader adoption of TPUS for evaluating pelvic floor disorders in both research and clinical practice. An automated segmentation and plane extraction method would improve the ability to easily quantify pelvic anatomy relevant to pelvic floor disorders and improve the efficiency and reproducibility of POP diagnosis and treatment.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;To develop a fast, reproducible, and automated method of acquiring the MS plane, plane of minimal hiatal dimensions (PMHD), and segmentations of the pelvic floor organs from 3D TPUS images.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;Our method used a nnU-Net segmentation model to segment structures of interest in the 3D TPUS images. The model segmented the pubis symphysis (PS), urethra, bladder, rectum, rectal ampulla, and anorectal angle (ANA). The segmented output was then fed into a heuristics-based method to determine the PS and ANA to extract the MS plane and PMHD automatically. We used a dataset consisting of 161 3D TPUS images from 104 patients. 89 of the volumes were acquired in a resting state and 72 during the Valsalva maneuver. The segmentation and plane extraction algorithms were evaluated by comparing the results with manual segmentations and manual plane extraction methods using the dice similarity coefficients (DSC), mean absolute surface distance (MAD), and absolute angle difference (AAD), respectively. The Wilcoxon-signed rank statistical test was used with Bonferroni-correction to p &lt; 0.01. Cohen effect size was used for comparing model results.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;The nnU-Net segmentation model reported an average DSC(%) of 70.4%, 58.5%, 57.1%, 48.9%, 39.0%, and 19.8% for bladder, rectum, PS, urethra, ANA, and rectal ampulla respectively. The nnU-Net segmentation model achieved significantly higher DSC (p &lt; 0.01) for the urethra and rectum than all other tested models. Across all metrics, the nnU-Net segmentation model achieved an average effect size of 0.3, 0.5, 0.7, and 0.8 compared to a 3D ResNet34 + U-Net, 3D U-Net, 2D U-Net, and Attention 3D U-Net model, respectively. The average AADs between the automatically calculated plane slices and manually estimated planes dataset for the MS plane and PMHD were 3.8° and 2.4°, respectively. The PS and ANA segmentation centroids were used to calculate the MS plane and PMHD and they had distance errors of 3.6 mm and 4.4 mm.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;We developed an automa","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70473"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147871077","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}

{"title":"Performance evaluation of a clinical SPECT system with multifocal collimator for brain studies.","authors":"Tamaki Otani, Yamato Kunikane, Satoru Takashi, Shota Azane, Akihiko Fujita, Noritake Matsuda, Masafumi Amano","doi":"10.1002/mp.70451","DOIUrl":"https://doi.org/10.1002/mp.70451","url":null,"abstract":"<p><strong>Background: </strong>Single-photon emission computed tomography (SPECT) is an indispensable examination for evaluating brain function to diagnose dementia. During head examination, the collimator makes contact with the shoulders, which often causes the patient discomfort. A new multifocal high-resolution collimator (SMARTZOOM high resolution and extended [SZHRX]) can maintain image quality, even when there is distance between the patient and the collimator. Therefore, it may be a useful tool to prevent patient discomfort during head imaging.</p><p><strong>Purpose: </strong>In this study, we evaluated spatial resolution, sensitivity, uniformity, and %contrast in several basic experiments using different phantoms to clarify the image quality of the multifocal collimator.</p><p><strong>Methods: </strong>We used ^99mTc and ¹²³I nuclides, which were sealed in several phantoms and imaged using SZHRX. The rotation radius for SZHRX was varied from 24-34.5 cm. For comparison, low-energy high-resolution (LEHR) and low-medium energy general purpose (LMEGP) images were acquired with a rotation radius of 14 cm. Spatial resolution, sensitivity, uniformity, and %contrast were calculated from the images obtained using each phantom and collimator.</p><p><strong>Results: </strong>For both ^99mTc and ¹²³I, the wider the radius of rotation, the larger the full-width half maximum (FWHM) of SZHRX. The FWHM of SZHRX was larger than that of LEHR, and the FWHM of SZHRX was smaller than that of LMEGP. The sensitivity of SZHRX increased as the distance increased, regardless of nuclides. For ^99mTc, SZHRX had higher sensitivity than LEHR. The sensitivity of SZHRX for ¹²³I was lower than that of LMEGP at short distances, with the same sensitivity at a radius of rotation of 26-28 cm. The coefficient of variation (CV) of ^99mTc for SZHRX was lower than that for LEHR, and the further the distance, the lower the CV. For ¹²³I, the CVs of SZHRX and LMEGP were comparable. The %contrast of SZHRX worsened as the rotation radius increased. For ^99mTc, the %contrast of SZHRX was lower than that of LEHR. For ¹²³I, the %contrast of SZHRX and LMEGP were comparable.</p><p><strong>Conclusions: </strong>It is possible to acquire images with high spatial resolution and uniformity, even if the rotation radius is widened by the use of SZHRX. The image quality of SZHRX obtained in this study will be helpful for future clinical applications.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70451"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13107236/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147793349","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}