Medical physics最新文献

{"title":"Equivariant conditional diffusion model for head and neck CT image synthesis from CBCT.","authors":"Alzahra Altalib, Chunhui Li, Alessandro Perelli","doi":"10.1002/mp.70617","DOIUrl":"10.1002/mp.70617","url":null,"abstract":"<p><strong>Background: </strong>Cone-beam computed tomography (CBCT) is a commonly used modality for image guided radiotherapy (IGRT). It offers real time anatomical visualization with low acquisition cost and dose. Nevertheless, photon scattering and beam hindrance lead CBCT images to suffer from several artifacts. These involve inaccurate Hounsfield Unit (HU) values, which render a lower reliability towards the dose calculations and adaptive planning.</p><p><strong>Purpose: </strong>Computed tomography (CT), on the contrary, offers better image quality and accurate HU calibration, yet is typically acquired using offline mode and fails to capture the intra-treatment anatomical changes. This renders a need for developing an accurate CBCT to CT synthesis to mitigate the gap in imaging quality in the adaptive radiotherapy workflow.</p><p><strong>Methods: </strong>To cater to this, we propose a novel diffusion based conditional generative model, coined EqDiff-CT, to synthesize high quality CT images from CBCT. EqDiff-CT employs a denoising diffusion probabilistic model (DDPM) to iteratively inject noise and learn latent representations that enable reconstruction of anatomically consistent CT images. A group equivariant conditional U-Net backbone, implemented with e2cnn steerable layers, enforces rotational equivariance (cyclic C4 symmetry), helping preserve fine structural details while minimizing noise and artifacts.</p><p><strong>Results: </strong>The system was trained and validated on the SynthRAD2025 dataset, comprising CBCT and CT scans across multiple head and neck anatomical sites, and we compared it with advanced methods such as CycleGAN and DDPM. EqDiff-CT provided substantial gains in structural fidelity, HU accuracy and quantitative metrics. Visual findings further confirm the improved recovery, sharper soft tissue boundaries, and realistic bone reconstructions.</p><p><strong>Conclusions: </strong>The findings suggest that the diffusion model has offered a robust and generalizable framework for CBCT improvements. The proposed solution helps in improving the image quality as well as the clinical confidence in the CBCT guided treatment planning and dose calculations.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70617"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13433820/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148671743","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":"AAPM Spring Clinical Meeting - Program.","authors":"","doi":"10.1002/mp.70459","DOIUrl":"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":"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":"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":"Direct electrical measurement of tube-voltage waveforms in mammography systems using ALDEN-type high-voltage connectors.","authors":"Toru Negishi, Kiyomitsu Shinsho, Shinji Abe, Izumi Ogura","doi":"10.1002/mp.70477","DOIUrl":"10.1002/mp.70477","url":null,"abstract":"<p><strong>Background: </strong>Accurate tube-voltage characterization is essential for mammography quality assurance and estimation of mean glandular dose (MGD). In mammography systems equipped with ALDEN-type high-voltage connectors, direct electrical access to the tube-voltage circuit has historically been restricted, limiting evaluation primarily to non-invasive measurement methods.</p><p><strong>Purpose: </strong>To develop and evaluate a direct-connection adapter enabling invasive acquisition of tube-voltage waveforms in mammography systems using ALDEN-type high-voltage connectors and to compare waveform characteristics with those obtained using a non-invasive multimeter.</p><p><strong>Methods: </strong>A direct-connection adapter-based measurement system was developed to provide electrical access to the tube-voltage circuit without modifying the primary conduction path. Tube-voltage measurements were performed using a commercial tube-voltage/tube-current meter incorporating a fixed high-voltage divider (1:20 000). Tube-voltage waveforms were measured in three clinical mammography systems under multiple target/filter configurations using both the invasive method and a commercially available non-invasive multimeter (RaySafe X2). Tube-voltage ripple was quantified from time-resolved voltage data obtained under identical exposure conditions.</p><p><strong>Results: </strong>Mean tube-voltage values obtained using invasive and non-invasive approaches showed approximately linear agreement with preset voltage. However, ripple magnitudes differed substantially under certain beam conditions. The invasive method yielded ripple values ranging from 2.5% to 6.2%, whereas non-invasive measurements reached up to 48.5% under specific configurations.</p><p><strong>Conclusions: </strong>Direct electrical measurement of tube-voltage waveforms in mammography systems using ALDEN-type high-voltage connectors was demonstrated using the proposed adapter. Because the invasive method directly measures the electrical high-voltage waveform without signal reconstruction, it provides an electrical reference for evaluating potential variability in non-invasive waveform measurements and may support generator characterization and waveform-based investigations in research and equipment evaluation settings.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70477"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13175943/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147944384","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":"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":"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":"Design of a next-generation conventional scintillator x-ray detector: Improved spatial resolution and fill factor.","authors":"Scott S Hsieh","doi":"10.1002/mp.70486","DOIUrl":"10.1002/mp.70486","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;The most compelling advantage that photon counting detectors (PCDs) currently have over energy integrating detectors (EIDs) for CT applications may be their improved spatial resolution. However, PCDs are difficult to manufacture. Next-generation EIDs using conventional scintillator sensors could be easier to mass produce, but their spatial resolution is thought to be limited by fill factor considerations. Specifically, each EID pixel comprises a block of scintillator surrounded by inactive, optically reflective septa. The minimum width of optically reflective septa is ∼0.1 mm, and unless this is shrunk, decreases in pixel pitch inevitably cause an increasing fraction of dead space.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;To propose a modified system geometry in which the detector pixels are tilted by a few degrees, such that rays that are incident on reflective septa would later encounter scintillator. Because reflective septa are nearly transparent to x-rays compared to scintillator, this geometry restores much of the lost fill factor.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;Tilting can be achieved in several different ways. We discuss three implementation options: individual detector modules may be tilted, with gaps between modules that are seamless from the perspective of the x-ray source; the whole detector can be arranged to point towards a \"false\" focal spot that is offset slightly from the true focal spot; the scintillator wafer can structured differently, e.g. by sawing it at an angle. The pros and cons of each option are discussed. Tilting introduces blurring, but a smaller pixel size improves resolution, and the combination of both improves both resolution and fill factor. We used Monte Carlo simulations to compare a conventional detector with 1.0 mm, 0.75, or 0.5 mm pixel pitch to a tilted detector with 0.5 mm pixel pitch. We assumed negligible optical crosstalk and electronic noise, that the scintillator was 1.5 mm thick with attenuation similar to gadolinium oxysulfide, that the reflective septa was 0.1 mm wide at standard resolution or 0.08 mm at higher resolutions and composed of a low-attenuating polymer similar to acrylic infused with 20% titanium content. We also compared this to a simple model of a CdTe detector with 1.6 mm thickness, 0.3 mm pixel pitch, and Gaussian charge sharing.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Tilt increases the aperture of the pixel in an energy-dependent fashion. With a 1D ASG (85% fill factor) and at 120 kVp, the detection efficiencies (DQE(0)) of a conventional detector at 1.0 mm, conventional detector at 0.5 mm pixel pitch, CdTe detector, and tilted detector at 0.5 mm pixel pitch are 65%, 55%, 66%, and 64%, respectively. With a sparse 2D ASG (1.5 × 3.0 mm openings, 86% fill factor), the DQE(0) of a tilted detector with 0.75 mm is 69%. The DQE(0) of all detectors decreases at higher energy because of punch-through.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;By tilting detector modules a few deg","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 5","pages":"e70486"},"PeriodicalIF":3.2,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13231487/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147960372","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":"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":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13158172/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147871067","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":"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":"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":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13158169/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147871077","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":"Correction to \"Lightweight deep training network for lymph nodes segmentation from head and neck CT images\".","authors":"","doi":"10.1002/mp.70453","DOIUrl":"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":"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}