Medical physics最新文献

{"title":"CT texture phantom dataset with paired image quality assessments for quantitative imaging.","authors":"Morgan A Daly, John M Hoffman, Andrew M Hernandez, Ali Uneri, Bino A Varghese, Joshua Levy, Michael F McNitt-Gray","doi":"10.1002/mp.70633","DOIUrl":"10.1002/mp.70633","url":null,"abstract":"<p><p>To establish a public repository of computed tomography (CT) texture phantom images paired with objective 3D image quality measurements from multiple scanner models using a diverse range of imaging protocols, facilitating investigations into relationships between image quality and quantitative imaging features. Three specialized CT phantoms were scanned: (1) the Corgi^® phantom for image quality assessment, (2) a radiomics liver phantom, and (3) an open-source 3D-printed texture phantom. Image quality assessment included measurement of contrast-to-noise ratio, 3D modulation transfer function, and 3D noise power spectrum. Data were acquired on four CT scanner models from two manufacturers at five <math><msub><mi>CTDI</mi> <mrow><mi>v</mi> <mi>o</mi> <mi>l</mi></mrow> </msub> </math> levels (2.05-17.11 mGy) and reconstructed with eight different kernels, yielding 160 total conditions (combinations of scanner, dose, kernel). Data were validated for integrity and completeness, resulting in the exclusion of six conditions. The paired texture phantom and image quality (PTP-IQ) dataset includes: (1) DICOM image series of two texture phantoms acquired across the 154 conditions, as well as (2) image quality metrics derived from each corresponding set of scanner, acquisition and reconstruction settings provided in HDF5 format. This dataset enables the development and validation of harmonization methods for multi-center quantitative imaging studies, investigation of protocol-dependent QIF variability, and optimization of acquisition protocols for radiomics applications. The controlled and systematic study design facilitates isolation of individual protocol effects on quantitative measurements.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 9","pages":"e70633"},"PeriodicalIF":3.2,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13527279/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148868815","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 \"Knowledge-distilled diffusion models for improving cone-beam CT image quality with meta-learning under imbalanced data\".","authors":"","doi":"10.1002/mp.70608","DOIUrl":"10.1002/mp.70608","url":null,"abstract":"","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70608"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13433738/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148671740","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-world assessment of a deep learning neural network algorithm for prostate cancer detection in MRI using true de novo data.","authors":"Steven M Shea, Michael Wesolowski, Abdulrahman Hashem, Ari Goldberg, Cara Joyce, Gopal Gupta, Robert Grimm, Heinrich von Busch, Bin Lou, Ali Kamen","doi":"10.1002/mp.70614","DOIUrl":"10.1002/mp.70614","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Deep-learning neural network algorithms for detecting prostate cancer in MRI have proliferated in the literature. However, out of 30+ studies published since the PROSTATEx challenge, no studies tested the performance of their algorithm against using true external image data sets (studies came from an outside institution that did not supply any training data to the algorithm) while validating against MR-US fusion biopsy or whole-mount prostatectomy. Using true external data sets paints a much clearer picture of real-world clinical performance of an algorithm.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;This work will assess the performance of a published deep learning (DL) neural network algorithm to detect prostate cancer using external studies. The main difference from other studies is the combination of using only MR-US fusion biopsy results as a gold standard; using test data from an institution that did not supply any training data for this version of the algorithm (including studies acquired with an endorectal coil, which were not in the original training set); and comparing the performance of algorithm-generated regions-of-interest (ROIs) versus algorithm heat maps.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;Patients were included in the study if they had a prostate MRI with at least one radiologist-drawn target on MRI and underwent MR-US fusion biopsy where the target was sampled for pathological analysis. Patients were excluded if they had any history of prostate cancer treatment, had previously undergone MR-US fusion biopsy at our institution, were missing MRI acquisitions, had artifacts in image sets, or if the study had been shared for future algorithm development. MR image data was assessed using a DL research prototype (XProstate) from Siemens Healthineers that produced (a) ROIs in suspected cancer areas with a level of suspicion (LoS) score and (b) heat maps with LoS scores across the entire gland. The XProstate prototype had been trained with 2170 studies from eight different academic institutions. Clinical radiologist, XProstate ROI, and XProstate Heat Map scores were assessed with ROC analysis using pathology results from biopsy as a gold standard.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;202 unique patients were included for assessment of the XProstate research prototype. The ROC curve for the XProstate Heat Map LoS score generated the highest AUC (0.76, 95% CI: 0.70, 0.82) followed by clinical radiologist PI-RADS score (0.73, 95% CI: 0.68, 0.79) and by XProstate ROI LoS score (0.71, 95% CI: 0.65, 0.77). Neither the XProstate Heat Map (AUC difference = 0.03, 95% CI: -0.04, 0.10, p = 0.38) nor the XProstate ROI (AUC difference = -0.02, 95% CI: -0.09, 0.04, p = 0.43) was significantly different from the radiologist PI-RADS score.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;The XProstate prototype demonstrated equivalent performance as clinical radiologists when presented with de novo cases that would mirror a real-world clinical deploymen","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70614"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13431142/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148665369","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":"Patient-specific automated multi-class anatomical motion tracking on real-time cine MR images using deep learning techniques.","authors":"Kaishen Li, Sanjiv S Samant","doi":"10.1002/mp.70622","DOIUrl":"10.1002/mp.70622","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Real-time multi-class motion tracking on cine images during magnetic resonance-guided radiation therapy (MRgRT) would enable instant monitoring of anatomical structures, facilitating dynamic beam control to minimize overdose to normal tissues and maximize tumor dose. However, conventional segmentation strategies incur significant time delays, making their clinical application impractical. Emerging deep learning (DL) technologies can provide a path to address this clinical challenge.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;To propose a patient-specific framework for real-time automated simultaneous tracking of multiple anatomical structures on 2D magnetic resonance cine (cine MR) images acquired in two orthogonal planes, using DL techniques without the need for extensive training data.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;In this study, we evaluated SegResNet, a DL-based auto-segmentation model, and TransMorph, a DL-based registration model. Additionally, conventional rigid and B-spline registration algorithms served as baseline models. Our datasets consist of 10 subjects, with both coronal and sagittal image sequences acquired for each subject using Elekta Unity MR-Linac protocols. With ground truth contours manually delineated for every frame, each model was trained and tested on a subject and plane-specific basis. Dice similarity coefficient (DSC), 95th-percentile Hausdorff distance (HD95), average surface distance (ASD), and center-of-mass errors (COME) were chosen to evaluate the output segmentations of each model. Furthermore, to analyze segmentation motion consistency across two orthogonal planes, the temporal rhythm and physical magnitude of the liver centroid trajectory were evaluated using the Pearson correlation coefficient (PCC) and the mean-centered Concordance correlation coefficient (CCC), respectively. Finally, a phase-resolved analysis was conducted to assess the robustness of each model across the motion cycle.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;SegResNet yielded significantly better results than registration models (p &lt; 0.05). Averaged across all organs, planes, and subjects, SegResNet achieved an overall DSC of 0.97 ± 0.01, HD95 of 3.0 ± 0.9 mm, ASD of 1.1 ± 0.3 mm, and COME of 0.9 ± 0.4 mm. The PCC and mean-centered CCC of SegResNet are 0.77 ± 0.07 and 0.75 ± 0.08, respectively. Furthermore, phase-resolved analysis demonstrated SegResNet's phase-agnostic spatial stability, with no statistically significant differences (p &gt; 0.05) observed across the three kinematic states for any organ across all evaluated segmentation metrics. The average inference time for SegResNet is approximately 0.3 ms per frame.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusion: &lt;/strong&gt;SegResNet outperformed TransMorph, as well as conventional B-spline and rigid registration models in terms of segmentation accuracy, cross-plane consistency, and robustness across different motion phases, and computation efficiency, potentially allowing for real-time simultaneous","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70622"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13433807/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148671818","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 multidisciplinary experimental and methodological investigation of electron and proton minibeams in the framework of the INFN MIRO project.","authors":"Francesco Romano, Francesca Cella Zanacchi, Esther Ciarrocchi, Gaia Franciosini, Giuliana Milluzzo, Emanuele Scifoni, Anna Vignati, Said Ahmad, Silvia Arezzini, Andrea Attili, Marco Battestini, Jacopo Bernardini, Matteo Bettanin, Alessandra Bisio, Maria Giuseppina Bisogni, Igor Bodrenko, Giulio Bordieri, Valentina Bravatà, Alberto Burattini, Massimo Camarda, Francesco Paolo Cammarata, Simone Capaccioli, Lorenzo Castelli, Andrea Cavalieri, Mariagrazia Celentano, Enrica Chiadroni, Francesco Giuseppe Cordoni, Elena Corvaia, Mario Costa, Alessandro Curcio, Eleonora Da Pozzo, Martina De Felice, Francesca Del Debbio, Francesco D'Errico, Maria Cristina D'Oca, Beatrice D'Orsi, Umberto Deut, Elisabetta Durisi, Maria Evelina Fantacci, Stefano Farina, Giuseppe Felici, Arianna Ferro, Luca Ficcadenti, Antonino Formuso, Giusi Irma Forte, Simona Giordanengo, Alessia Giuliano, Lucia Giuliano, Luca Lanzanò, Stefania Linsalata, Simone Lossano, Margherita Maffei, Michela Marafini, Maurizio Marrale, Maurizio Massa, Luigi Masturzo, Enrico Mazzoni, Mauro Migliorati, Felix Mas Milian, Luigi Minafra, Andrea Moggi, Diango Montalvan Olivares, Marco Montefiori, Matteo Morrocchi, Andrea Mostacci, Franco Mostardi, Chinonso Okpuwe, Fabiola Paiar, Luigi Palumbo, Vincenzo Patera, Jake Harold Pensavalle, Gaia Pucci, Flaminia Quattrini, Alessandra Retico, Valeria Rosso, Giorgio Russo, Roberto Sacchi, Alessio Sarti, Silvia Scalisi, Camilla Scapicchio, Angelo Schiavi, Paolo Stochino, Enrica Strettoi, Francesco Tommasino, Valentina Tozzini, Giacomo Traini, Gabriele Trovato, Alice Usai, Francesco Urso, Fabio Di Martino","doi":"10.1002/mp.70537","DOIUrl":"10.1002/mp.70537","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;The clinical translation of Minibeam RT (MBRT) has recently started thanks to the first human treatments recently performed. However, despite experimental evidence, the impact of the dose distribution parameters involved on the magnitude of the effect itself and the underlying radiobiological mechanisms are still only partially understood. To address this issue, systematic investigations are needed through the implementation of advanced quantitative experiments with a multidisciplinary approach, which is the one proposed in the framework of the INFN funded MIRO (MInibeam RadiOtherapy) project.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;The aim of this work is to report on the on-going main activities recently carried out in the framework of the project, showing some of the main results achieved during the first 2 years of the project, in terms of: (i) facilities development and characterization; (ii) new dosimetric approaches; (iii) biological investigation of the effect; and (iv) development of the first tools for dose planning. The multidisciplinary approach adopted by this national Collaboration to tackle the main challenges of minibeam radiotherapy for a reliable and solid clinical translation is discussed.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;The facilities involved in the project are: (i) two facilities dedicated to low (up to 9 MeV) and medium (up to 18 MeV) energy electron minibeam studies, one of them also equipped with an electron FLASH LINAC to study possible synergistic effects with UHDR beams; (ii) one facility dedicated to proton minibeam studies, with energies from 70 to 140 MeV. Novel dosimetric approaches, are presented, mainly based on scintillators, silicon and silicon carbide detectors. Novel techniques for the analysis of biological samples are described, leveraging both integrated averaged data and spatially resolved analysis. Furthermore, a framework for biological modeling using a multiscale approach is presented. A dedicated dose-planning tool is currently under development to compare \"virtual\" minibeam plans with conventional ones and, consequently, to quantitatively investigate the potential for clinical translation.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;All the facilities were dosimetrically characterized, demonstrating the capability of producing controlled and reproducible electron and proton minibeams. These beams exhibited diverse physical parameters, including peak-to-valley ratios between 3 and 30 at the entrance and center-to-center distances between 2 and 3 mm. The developed detectors, specifically scintillators and silicon detectors, successfully measured the minibeam patterns with sub-millimeter resolution, while large-area silicon carbide detectors were used for average dose measurements. First in-vitro biological investigations performed with low energy minibeams clearly showed an enhanced survival fraction in the 16HBE healthy lung cells, while maintaining iso-effective cell killing in A549 cance","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70537"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13433822/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148671688","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":"Improving dual-panel in-beam PET imaging for proton therapy monitoring using 3D U-Net.","authors":"Dengyun Mu, Pengyuan Qi, Ao Qiu, Qi Liu, Qiuhui Ma, Peng Xiao, Lin Wan, Bingxuan Li, Jianchun Deng, Bensheng Qiu, Zhiyong Yang, Qingguo Xie","doi":"10.1002/mp.70609","DOIUrl":"10.1002/mp.70609","url":null,"abstract":"<p><strong>Background: </strong>In-beam positron emission tomography (PET) integrates dedicated detectors into proton therapy systems, enabling real-time acquisition of proton-induced positron-emitting activity. By pre-delivering a subset of single-energy proton spots as probes, in-beam PET has the potential to support online proton range verification and may inform subsequent spot delivery. However, the limited acquisition time and low yields of positron emitters result in noisy reconstructed images, and the open geometry of dual-panel PET further introduces stretching artifacts.</p><p><strong>Purpose: </strong>To address these issues, this study proposed a 3D U-Net-based post-processing method to enhance in-beam PET image quality and evaluated its feasibility through Monte Carlo simulations and preliminary experimental validation.</p><p><strong>Methods: </strong>Thirteen head computed tomography (CT) phantoms were irradiated with horizontally and vertically incident single-energy probes, representing two scenarios in which beams passed through relatively homogeneous and heterogeneous tissues. A phantom-level split was used, with seven CT phantoms for training, two for validation, and four held-out CT phantoms for testing. Proton probes in 3×3, 5×5, 7×7, and 9×9 spot patterns were simulated, with each spot delivering 2×10⁷ protons. In-beam PET acquisition was performed for 30 s. Unfiltered reconstructed PET images were used as network inputs, and ground-truth activity distributions served as labels. Gaussian-filtered PET images served as the baseline for performance evaluation. Image quality was quantified by peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM), and proton range verification accuracy was evaluated by absolute range error (ARE). Additional simulations with CT density perturbations were conducted to evaluate sensitivity to proton range shifts. Preliminary experimental validation was performed using an all-digital PET prototype and polymethyl methacrylate (PMMA) phantoms.</p><p><strong>Results: </strong>For horizontal beams, median PSNR values increased from 21.4-23.9 to 35.2-37.4, median SSIM values increased from 0.450-0.621 to 0.906-0.961, and median ARE values decreased from 1.05-1.18 mm to 0.52-0.73 mm. For vertical beams, median PSNR values improved from 21.9-23.2 to 29.2-34.1, median SSIM values improved from 0.480-0.560 to 0.745-0.895, and median ARE values decreased from 1.80-2.40 mm to 1.50-1.73 mm. The proposed method also demonstrated improved sensitivity to simulated range shifts and consistent improvements on experimental PET data.</p><p><strong>Conclusion: </strong>These findings demonstrate the preliminary feasibility of using 3D U-Net-based post-processing to enhance dual-panel in-beam PET image quality and improve PET-based proton range verification.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70609"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13430982/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148665407","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":"Evaluation of AI-derived LAD artery dose metrics for survival stratification in stage III NSCLC: A secondary analysis of RTOG 0617 trial.","authors":"Libing Zhu, Yi Rong, Randa Tao, Yun Bai, Nathan Y Yu, Quan Chen","doi":"10.1002/mp.70621","DOIUrl":"10.1002/mp.70621","url":null,"abstract":"<p><strong>Background: </strong>While the radiation dose to left anterior descending (LAD) artery is associated with overall survival (OS) in NSCLC patients, manual segmentation is labor-intensive. Several auto-segmentation models were developed on average 4DCT with promising geometric agreement, yet their feasibility in clinical outcome analysis on lung CT remains unclear.</p><p><strong>Purpose: </strong>This study evaluates whether automated LAD segmentation can reproduce, at the group level, the established dose-survival association in a multi-institutional clinical trial dataset.</p><p><strong>Methods: </strong>Three deep learning-based auto-segmentation (DLAS) models for automatic segmentation of the LAD coronary artery were applied to 460 patients from the NRG Oncology/RTOG 0617 dataset. Kaplan-Meier curves and Cox proportional hazard ratio (HR) were calculated using each model's segmentation. The LAD V15 Gy was used to stratify patients into high-and low-risk groups, and the cut-off values were determined by minimizing mean absolute error (MAE) between DLAS-derived and manual OS curves.</p><p><strong>Results: </strong>All three models demonstrated significantly different OS curves between the high risk and low risk groups (p < 0.05). The HR was 1.40 (95% CI: 1.02-1.93; p = 0.0381; MAE = 7.1%), 1.32 (95% CI: 1.03-1.68; p = 0.0258; MAE = 9.6%), and 1.37 (95% CI: 1.03-1.84; p = 0.0331; MAE = 7.9%) for the three models using a common V15 Gy ≥ 10% cut-off. After optimizing the cut-off, one model aligned well with what was derived from manual segmentation (MAE = 4.3%).</p><p><strong>Conclusion: </strong>Among the evaluated models, one DL model showed the strongest performance of survival groups separation and achieved group-level overall survival stratification comparable to that previously reported using manual contours in the multi-institutional cohort. Further validation with paired manual contours and pre-specified cutoffs is warranted.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70621"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13431141/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148665352","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":"Photon minibeam-based LATTICE radiotherapy for small and medium-sized tumors: A dosimetric planning study.","authors":"Wei Wu, Nimita Shinde, Jiaxin Li, Isabelle Vanhaezebrouck, Kai Jiang, Yuting Lin, Qiang Li, Hao Gao","doi":"10.1002/mp.70596","DOIUrl":"10.1002/mp.70596","url":null,"abstract":"<p><strong>Background: </strong>Lattice radiotherapy (LRT) is a spatially fractionated technique that delivers three-dimensional high-dose vertices within tumors. However, conventional LRT is constrained by geometric limitations when applied to small and medium-sized tumor volumes, primarily due to the large beam sizes that restrict optimal lattice pattern formation.</p><p><strong>Purpose: </strong>This study aims to investigate a photon minibeam-based LATTICE radiotherapy (mini-LRT) approach designed to overcome the geometric limitations of conventional LRT and enable effective spatially fractionated treatment for small and medium-sized tumors adjacent to critical organs.</p><p><strong>Methods: </strong>Three brain cases and three lung cases, with clinical target volumes ranging from 12.34 cc to 40.88 cc, were presented for treatment planning comparison between conventional LRT and mini-LRT. A multi-collimator delivery strategy employing shifted slit patterns was implemented to achieve complementary spatial coverage. Dose calculation and optimization were performed using Monte Carlo simulations and solved via an iterative convex relaxation algorithm. Reference stereotactic body radiation therapy (SBRT) plans were generated for all six cases, and setup-error robustness analysis was performed for two representative cases using nominal and shifted dose-influence matrix scenarios.</p><p><strong>Results: </strong>Mini-LRT produced denser lattice structures in all cases, generating 5-11 vertices compared to only 1-3 vertices achievable with conventional LRT. Vertex diameters were reduced from 15.0 mm to 3.0-4.0 mm, and vertex-to-vertex distances decreased from 25.0 mm to 12.5 mm. The lattice volume ratio ranged from 0.68% to 1.07% with mini-LRT. Organs at risk (OARs) dose reductions were observed in the evaluated cases, including reduced mean dose to brainstem, heart, and esophageal. Furthermore, mini-LRT achieved a higher peak-to-valley dose ratio (PVDR) ranging from 2.19 to 2.94, compared to 1.75-2.34 for conventional LRT. Reference SBRT plans achieved clinical target volume (CTV) D₉₅ = 50 Gy in all cases but showed elevated D_0.03cc to adjacent OARs in selected anatomically constrained cases. In the representative robustness analysis, worst-case setup-error scenarios showed modest PVDR degradation.</p><p><strong>Conclusions: </strong>This proof-of-concept dosimetric planning study suggests that photon minibeam-based LATTICE radiotherapy can generate compact spatially fractionated dose distributions in selected small and medium-sized tumors. Compared with conventional LRT, mini-LRT increased vertex density, reduced lattice volume ratio, and improved OAR sparing in the evaluated cases. Additional experimental validation, motion assessment, and larger patient studies are required before clinical translation.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70596"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13431144/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148665327","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":"Ultrasound-based thyroid nodule segmentation with deep hybrid convolutional network.","authors":"Fan Lu, Hao Sun, Binbin Jiang, Zhongyi Zhang, Ge Ren, Jing Cai, Yidong Gu, Peng Ren, Tao Peng","doi":"10.1002/mp.70548","DOIUrl":"10.1002/mp.70548","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Automatic segmentation of ultrasound-based thyroid nodules can assist physicians in more efficiently and accurately assessing thyroid diseases. However, thyroid ultrasound imaging presents certain unique challenges, such as blurred boundaries of nodules, strong heterogeneity in tissue echoes, and limited contrast between the lesions and adjacent normal thyroid tissue, which hinder the precise identification of nodules in clinical assessment.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;To overcome these obstacles, we propose a deep hybrid convolutional network, named UTNseg, for precisely automatic segmentation of ultrasound-based thyroid nodules.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;UTNseg contains four newly designed modules upon TransUNet: (1) the spatial-channel feature strong capture (STransformer) module is designed to reinforce the ability of encoder on capturing spatial and channel features, (2) the redundant feature map utilization (RMU) module is devised to mitigate redundant information in the skip connections, (3) the composite convolution (CConv) module is newly designed to further preserve the key multi-scale features of the target thyroid nodules, (4) the lightweight edge refinement (LER) module is newly proposed to refine the initial segmentation outputs. UTNseg was evaluated on two publicly available ultrasound datasets of thyroid nodules (TN3K with 3493 images and DDTI with 637 images). Datasets were divided at the patient level, with 4000 training images (2000 images from TN3K and 2000 images from DDTI, augmented by random rotation, scaling, and horizontal flipping on 445 images), 761 validation images (698 images from TN3K and 63 images from DDTI), and 924 test images (795 images from TN3K and 129 images from DDTI). Additionally, separate experiments were performed on these two datasets in comparative experiments. And the performance was compared with six state-of-the-art segmentation methods (U-Net, TransUNet, TRFENet, UNeXt, TRFE+ and BPAT-UNet). Evaluation metrics mainly included Dice similarity coefficient (DSC) and 95th percentile Hausdorff distance (HD95). Statistical significance level was analyzed through the Wilcoxon signed-rank test, and the effect size was quantified by Cohen's &lt;math&gt;&lt;mi&gt;d&lt;/mi&gt;&lt;/math&gt; . Holm-Bonferroni correction was employed to correct for multiple comparisons.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;On the TN3K dataset, UTNseg achieved 79.52% DSC and 16.32 HD95. On the DDTI dataset, UTNseg achieved 86.21% DSC and 12.76 HD95. On the hybrid dataset, UTNseg achieved 80.77% DSC and 14.56 HD95. These results showed that UTNseg achieved superior performance. Compared with state-of-the-art approaches, UTNseg was statistically significant ( &lt;math&gt;&lt;mrow&gt;&lt;mi&gt;p&lt;/mi&gt; &lt;mo&gt;&lt;&lt;/mo&gt; &lt;mn&gt;0.05&lt;/mn&gt;&lt;/mrow&gt; &lt;/math&gt; ), and the corresponding effect sizes were mostly medium or large, further quantifying the practical importance of the improvements.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;UTNseg demonstrated excellent s","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70548"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13431149/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148665379","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 hybrid deep learning framework for real-time speckle reduction and image enhancement on portable ultrasound systems.","authors":"Hyunwoo Cho, Jaeseok Lee, Jongsoo Lee, Yangmo Yoo, Jinbum Kang","doi":"10.1002/mp.70620","DOIUrl":"10.1002/mp.70620","url":null,"abstract":"<p><strong>Background: </strong>Speckle patterns in ultrasound images often obscure anatomical details, leading to diagnostic uncertainty. Recently, various deep learning-based techniques have been introduced to effectively suppress speckle; however, their high computational costs pose challenges for low-resource devices, such as portable ultrasound systems.</p><p><strong>Purpose: </strong>To address this issue, we introduce Edge Speckle Reduction and Image Enhancement (EdgeSRIE), a lightweight hybrid deep learning framework for real-time speckle reduction and image enhancement in portable ultrasound imaging.</p><p><strong>Methods: </strong>The proposed framework consists of an unsupervised despeckling branch and a self-supervised deblurring branch trained separately in sequence with AdamW (learning rate 1 × 10^- ⁴) and an L2 loss. The integrated model was converted to an 8-bit deployment model using post-training quantization on a low-resource system-on-chip. Training used 1779 B-mode images (BUSI and HC18), validation used 77 raw acquisitions (PICMUS and CUBDL), and external evaluation used 94 EdgeFlow UH-10 bladder images. EdgeSRIE was compared with OSRAD, OBNLM, DIAE, DUNet, BRUNet, and USNet using contrast-to-noise ratio (CNR), speckle signal-to-noise ratio (SSNR), average gradient magnitude (AGM), and structural similarity index measure (SSIM); paired inference used two-sided Wilcoxon signed-rank tests with Holm correction and rank-biserial and Hedges' g effect sizes.</p><p><strong>Results: </strong>Across the four representative cases, post-training-quantized EdgeSRIE achieved the largest mean improvement in CNR (64.9 ± 21.0%) and the smallest mean decrease in AGM (-16.6 ± 33.2%). SSNR (109.8 ± 45.8%) was in the OBNLM-led range, whereas SSIM remained slightly lower than OSRAD and OBNLM on its native [0, 1] scale. The deployed model contained 17.67K parameters and reached 64.10 frames/s on the target hardware. In the case-matched analysis (n = 8), EdgeSRIE showed Holm-corrected significant advantages over all six baselines in CNR, over five of six in SSNR, and over four of six in AGM. For SSIM, EdgeSRIE was significantly higher than the four deep learning baselines but lower than OSRAD and OBNLM. Across the statistically significant comparisons, the magnitude-based effect sizes were large (Hedges' g ≥ 0.8).</p><p><strong>Conclusions: </strong>These results support the feasibility of EdgeSRIE as a compact, deployment-oriented framework that balances speckle suppression, structural preservation, and real-time execution for portable ultrasound imaging.</p>","PeriodicalId":94136,"journal":{"name":"Medical physics","volume":"53 8","pages":"e70620"},"PeriodicalIF":3.2,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13430983/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148665341","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}