Medical physics最新文献

{"title":"Development of a high-fidelity phantom for training ultrasound-guided radiofrequency ablation of thyroid nodules","authors":"Tsung Han Yang, Nguyen-Ngan-Ha Lam, Natalie Tanjaya, Tsu-Chi Hsu, Tsung-Wei Lin, Wei-Siang Ciou, Wei-Che Lin, Yi-Chun Du","doi":"10.1002/mp.70035","DOIUrl":"10.1002/mp.70035","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <h3> Background</h3>\u0000 \u0000 <p>Thyroid nodules (TNs) are common solid or fluid-filled lumps in the thyroid gland, often benign but requiring treatment when they grow or cause symptoms. Ultrasound-guided radiofrequency ablation (RFA) has emerged as a minimally invasive alternative to surgery, particularly for benign TNs. However, precise execution is crucial, as the thyroid gland is surrounded by critical structures known as the “dangerous triangle,” including the recurrent laryngeal nerve and blood vessels. Inadequate targeting or excessive heat application during RFA can lead to complications. Currently, alternative training using phantoms can help inexperienced surgeons enhance surgical techniques and procedural safety. However, current phantom models often lack realistic tissue responses, particularly in mimicking protein coagulation, carbonization, and hydrodissection.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Purpose</h3>\u0000 \u0000 <p>This study aimed to develop a high-fidelity anthropomorphic neck and thyroid phantom that has similar ultrasound imaging characteristics and RFA response to that of human tissue. The phantom was designed to simulate key procedural steps, including ultrasound-guided hydrodissection and ablation-induced tissue changes, to support training in RFA of TNs.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Methods</h3>\u0000 \u0000 <p>The thyroid and neck phantom's anatomical structure was reconstructed using Computed Tomography (CT) imaging to create a 3D-printed mold. It was fabricated using biomimetic dual-network artificial materials (BDAM) through a multi-step molding process. The material characteristics, including the acoustic properties, ultrasound imaging, impedance, electrical conductivity, and thermal ablation response, were systematically evaluated. The phantom underwent ultrasound-guided hydrodissection before RFA, and the resulting ablation zones were compared with those observed in animal tissues.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Results</h3>\u0000 \u0000 <p>The phantom's material properties were validated and compared to human muscle and thyroid tissue characteristics from the literature. Additionally, the phantom produced clear ultrasound images during hydrodissection, effectively demonstrating the separation of tissue from the nodule. It also exhibited localized bubbling and coagulative carbonization in response to thermal ablation under ultrasound imaging. The ablation zone closely resembled that observed in pig liver tissues, with a standard deviation (SD) of ≤0.2 cm.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Conclusion</h3>\u0000 \u0000 <p>A high-fidelity phantom for training ultrasound","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145139943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mamba-Convolutional UNet for multi-modal medical image synthesis","authors":"WenLong Lin,&nbsp;Yu Luo,&nbsp;Jie Ling,&nbsp;FengHuan Li,&nbsp;Jing Qin,&nbsp;ZhiChao Yin,&nbsp;Shun Yao","doi":"10.1002/mp.70029","DOIUrl":"10.1002/mp.70029","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <h3> Background</h3>\u0000 \u0000 <p>Preoperative evaluation frequently relies on multi-modal medical imaging to provide comprehensive anatomical and functional insights. However, acquiring such multi-modal data often involves high scanning costs and logistical challenges. Additionally, in practical applications, it is inconvenient to collect sufficient matched multi-modal data to train different models for different cross-modality synthesis tasks.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Purpose</h3>\u0000 \u0000 <p>To address these issues, we propose a novel dual-branch architecture, named Mamba-Convolutional UNet, for multi-modal medical image synthesis. Furthermore, to enable cross-modal synthesis capabilities even under data scarcity, we address the practical challenge of limited paired multi-modal training data by introducing a simple reprogramming layer.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Methods</h3>\u0000 \u0000 <p>The proposed Mamba-Convolutional UNet adopts a U-shaped architecture featuring parallel SSM and convolutional branches. The SSM branch leverages Mamba to capture long-range dependencies and global context, while the convolutional branch extracts fine-grained local features through spatial operations. Then, an attention mechanism is utilized to integrate global and local features. To enhance adaptability across modalities with limited data, a lightweight reprogramming layer is incorporated into the Mamba module, allowing knowledge transfer from one cross-modal synthesis task to another without requiring extensive retraining.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Results</h3>\u0000 \u0000 <p>We conducted five multi-modal medical image synthesis tasks on three datasets to validate the performance of our model. The results demonstrate that the performance of Mamba-Convolutional UNet significantly outperforms that of six baseline models. Moreover, Mamba-Convolutional UNet can attain comparable performance to the current state-of-the-art methods by fine-tuning the model for other synthesis tasks with only 25% of the data.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Conclusions</h3>\u0000 \u0000 <p>The proposed Mamba-Convolutional UNet features a dual-branch structure that effectively combines global and local features for enhanced medical image understanding. And the Mamba block's reprogramming layer addresses challenges in target modality transformation during insufficient training.</p>\u0000 </section>\u0000 </div>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145139954","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparison of organ volumes and standardized uptake values in [18F]FDG-PET/CT images using MOOSE and TotalSegmentator to segment CT images","authors":"Julie Auriac,&nbsp;Christophe Nioche,&nbsp;Narinée Hovhannisyan-Baghdasarian,&nbsp;Charlotte Loisel,&nbsp;Romain-David Seban,&nbsp;Nina Jehanno,&nbsp;Lalith Kumar Shiyam Sundar,&nbsp;Thomas Beyer,&nbsp;Irène Buvat,&nbsp;Fanny Orlhac","doi":"10.1002/mp.70025","DOIUrl":"10.1002/mp.70025","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <h3> Background</h3>\u0000 \u0000 <p>Manual segmentation of organs from PET/CT images is a time-consuming and highly operator-dependent task. Open software solutions are now available to automatically segment all major anatomical structures in CT images.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Purpose</h3>\u0000 \u0000 <p>We compared the volumes and standardized uptake values (SUV) extracted from [18F]FDG-PET/CT patient scans for 33 anatomical structures segmented using two deep learning (DL) algorithms to determine if they are interchangeable.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Methods</h3>\u0000 \u0000 <p>Baseline [18F]FDG-PET/CT images were collected retrospectively for 315 women with metastatic breast cancer. A total of 33 anatomical volumes of interest (VOI) were segmented from the whole-body CT scans using both MOOSE v.3.0.14 and TotalSegmentator v.2.0.5 and copied onto the corresponding PET images. For each VOI, the volume from the CT image and SUVmax, SUVpeak and SUVmean from the PET image were extracted. The resulting values were compared using the relative difference for each feature.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Results</h3>\u0000 \u0000 <p>Following DL segmentation, resulting organ volumes differed by less than 10% for 19/33 organs in more than 80% (252/315) of patients. Four organs were segmented with volume differences greater than 20% in 1/5th of patients: bladder (48%, <i>p</i> &lt; 0.0001), portal and splenic veins (34%, <i>p</i> &lt; 0.0001), thyroid (16%, <i>p</i> &lt; 0.0001), adrenal glands (15%, <i>p</i> &lt; 0.0001). SUVmax and SUVpeak were affected by the choice of DL algorithms, with values differing by less than 10% in more than 80% of patients for only 16 and 19 out of 33 organs, respectively. In contrast, SUVmean was less affected with differences of less than 10% in more than 80% of patients for all anatomical structures, except the bladder, lungs and skull.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Conclusions</h3>\u0000 \u0000 <p>The two software tools produce similar results in volume estimates for most anatomical structures. SUVmean is less dependent on the segmentation algorithm than SUVmax and SUVpeak and shows excellent reproducibility for all anatomical structures studied except for the bladder, the lungs and the skull.</p>\u0000 </section>\u0000 </div>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://aapm.onlinelibrary.wiley.com/doi/epdf/10.1002/mp.70025","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145139914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multi-modal diffusion model for noise reduction of particle number limited Monte Carlo dose calculation for carbon ion radiotherapy","authors":"Jueye Zhang,&nbsp;Youfang Lai,&nbsp;Haonan Feng,&nbsp;Xiangde Luo,&nbsp;Kai-Wen Li,&nbsp;Tenghui Wang,&nbsp;Cheng Chang,&nbsp;Gen Yang,&nbsp;Chen Lin,&nbsp;Tian Li,&nbsp;Chao Yang,&nbsp;Yibao Zhang","doi":"10.1002/mp.70021","DOIUrl":"10.1002/mp.70021","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;The low computation efficiency impeded the broad application of Monte Carlo (MC) simulation to particle therapy. The existing deep learning (DL) methods for fast dose calculation lacked physics-based interpretability, hence, may introduce additional risks, especially for the more complex carbon ion radiotherapy.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;To develop and validate a multi-modal diffusion model, Diff-MC, for noise reduction of particle number limited MC dose calculation, potentially supporting better optimization and faster online adaptation for carbon ion radiotherapy.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;By using multi-modal data such as CT images, dose maps using a low number of primary particles and beam parameters, and so forth, Diff-MC was developed to generate a dose map adaptively based on the beam state. To enable effective inter-modal interactions, a hybrid-fusion strategy was applied to integrate the data-level, feature-level, and decision-level fusion. The model was evaluated on a highly heterogeneous dataset, including 15 000 paired beamlet data cropped from 20 CTs for training and validating, 500 paired beamlet data cropped from other 5 CTs for testing, and 500 paired beamlet data cropped from another 100 CTs for generalizability test. All datasets encompassed various geometry and beamlet physics parameters such as energy distribution and number of primary particles, and so forth.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Using the MC simulation based on high number of primary particles as ground-truth, the Diff-MC achieved nearly linear acceleration and high accuracy of gamma passing rate up to 99.25% under the criteria of 3 mm, 3%, 10% cutoff. The performance was significantly higher (all &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;p&lt;/mi&gt;\u0000 &lt;mo&gt;&lt;&lt;/mo&gt;\u0000 &lt;mn&gt;0.01&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$p&lt;0.01$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;) than the UNet-based models (96.17%) and transformer-based models (97.81%). The accuracy achieved by Diff-MC in the generalizability test was 99.22%. The lateral dose, integral depth dose (IDD), and percentage depth dose (PDD) of Diff-MC were also more consistent with the ground-truth than that of conventional AI models.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Conclusions&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;The proposed Diff-MC method displayed high efficiency and robustness in carbon ion dose calculation. By mai","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145139963","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparison of optimized magnetic resonance sequences for patient-specific treatment planning in surface brachytherapy","authors":"Michael J. Lavelle,&nbsp;Evangelia Kaza,&nbsp;Phillip M. Devlin,&nbsp;Ivan M. Buzurovic","doi":"10.1002/mp.70031","DOIUrl":"10.1002/mp.70031","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;The clinical standard practice of surface brachytherapy (SB) planning has long been to use computed tomography (CT) imaging to visualize applicators for catheter reconstruction in the treatment planning process. Recent work in SB has suggested that magnetic resonance (MR)-guidance can be used in place of CT-guidance in SB planning to utilize the increased soft tissue contrast for visualization of diseased tissue. This soft tissue visualization can be used to verify the target depth for enhanced coverage of the clinical target volume. Two optimized MR sequences (pointwise encoded time reduction with radial acquisition (PETRA) and volumetric interpolated breath-hold examination (VIBE) obtaining Dixon in-phase (DIP) and Dixon opposed-phase (DOP)) have been shown to detect sufficient signal from the silicone-based applicators to perform accurate catheter reconstruction and produce SB treatment plans.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;This study compares three in-house MR series optimized for applicator visualization to determine which is best-suited for SB planning based on tissue contrast and applicator visibility. This study then applies this series to produce MR-only SB treatment plans geometrically and dosimetrically comparable to those produced by CT-only for a phantom and eight patients.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;An anthropomorphic phantom (True Phantom Solutions, Canada) with applicators (Elekta, Netherlands) on the foot and hand and eight patients undergoing SB for Dupuytren's Contracture/Palmar fascial fibromatosis were imaged by two optimized MR sequences: 1) PETRA and 2) VIBE obtaining DIP and DOP images. CT scans were acquired for verification. SB planning was performed in Oncentra Brachy (Elekta, Netherlands) treatment planning software using three MR series and CT. MR-based and CT-based plans were compared for geometric and dosimetric accuracy. Geometric accuracy was determined by registering CT-based to MR-based catheter digitizations and calculating distances between corresponding dwell positions. Patient MR images were compared using signal-to-noise ratios (SNR's) and contrast-to-noise ratios (CNR's) for various regions of interest (ROIs) including bone, fat, muscle, and applicator. The series with the greatest tissue contrast and applicator visualization was used to produce treatment plans. MR-based plans were compared to CT-based plans by point-based dose differences (DD's). The MR-based plan was rigidly registered to the CT-based plans, and the isodose volumes were segmented to V150, V125, V100, V95, V90, V80, and V65 and compared using the Dice similarity coefficient (DSC) and volumetric similarity (VS) metric.&lt;/p&gt;\u0000 ","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Triple-energy photon-counting x-ray imaging for bone-strontium estimation: A simulation study","authors":"Jesse Tanguay,&nbsp;Bobby Tang,&nbsp;Eric Da Silva","doi":"10.1002/mp.18125","DOIUrl":"10.1002/mp.18125","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Strontium quantification in bone is clinically relevant but typically requires specialized stand-alone systems. Photon-counting detectors offer energy-resolved imaging that may enable low-dose estimation of both strontium concentration and bone mineral density in a single acquisition.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;To evaluate the feasibility of triple-energy photon-counting x-ray imaging for low-dose quantification of strontium in bone, using a simulation framework that accounts for energy bin sensitivity, detector noise, and anatomical geometry.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;A forward model of a photon-counting detector was used to simulate energy-resolved x-ray measurements through a simplified model of the human finger, incorporating cortical bone, trabecular bone, and soft tissue. Strontium uptake was modeled as a mass concentration relative to bone. A generalized least-squares estimator was used to compute the strontium-to-bone concentration from energy-resolved measurements. We optimized tube voltage and energy thresholds for three clinically relevant anode/filter combinations and three levels of electronic noise (5, 10, and 15 keV), with the goal of minimizing the limit of quantification (LOQ) and absorbed dose. A Fisher information analysis was conducted to assess the relative contribution of each energy bin to estimation precision.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Optimal tube voltages and thresholds depended strongly on electronic noise but only modestly on anode/filter choice. At a 5 keV noise floor, an LOQ of 100 ppm could be achieved with an absorbed dose of &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mo&gt;∼&lt;/mo&gt;\u0000 &lt;annotation&gt;$sim$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;13 &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;μ&lt;/mi&gt;\u0000 &lt;mi&gt;Gy&lt;/mi&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$mu{rm Gy}$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;, whereas 10 and 15 keV noise levels required &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mo&gt;∼&lt;/mo&gt;\u0000 &lt;annotation&gt;$sim$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;100 &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;μ&lt;/mi&gt;\u0000 &lt;mi&gt;Gy&lt;/mi&gt;\u0000 &lt;/mrow&gt;\u0000 ","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://aapm.onlinelibrary.wiley.com/doi/epdf/10.1002/mp.18125","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126904","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Analytical expression of the \u0000 \u0000 β\u0000 $beta$\u0000 coefficient of cell survival curves predicted by the NanOx model in the low-energy range","authors":"Mario Alcocer-Ávila,&nbsp;Étienne Testa,&nbsp;Michaël Beuve","doi":"10.1002/mp.70008","DOIUrl":"10.1002/mp.70008","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;In cancer research, clonogenic assays are often performed as a means to determine the response of a given cell line to radiation exposure. The resulting cell survival fractions as a function of absorbed dose are usually fitted to a linear-quadratic (LQ) expression involving two coefficients, &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mi&gt;α&lt;/mi&gt;\u0000 &lt;annotation&gt;$alpha$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt; and &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mi&gt;β&lt;/mi&gt;\u0000 &lt;annotation&gt;$beta$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;, describing the cell's radiosensitivity. However, &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mi&gt;β&lt;/mi&gt;\u0000 &lt;annotation&gt;$beta$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt; is particularly hard to compute with accuracy. On the other hand, biophysical models are developed for predicting the enhanced biological efficiency of heavy ions compared to photons. These models provide a more mechanistic description of the biological effects induced by ionizing radiation, while allowing the estimation of the &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mi&gt;α&lt;/mi&gt;\u0000 &lt;annotation&gt;$alpha$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt; and &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mi&gt;β&lt;/mi&gt;\u0000 &lt;annotation&gt;$beta$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt; coefficients.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;In this work, we propose an analytical expression for the fast computation of the &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mi&gt;β&lt;/mi&gt;\u0000 &lt;annotation&gt;$beta$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt; coefficient for ions with energies ranging from &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mo&gt;∼&lt;/mo&gt;\u0000 &lt;annotation&gt;$sim$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;1 to &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mo&gt;∼&lt;/mo&gt;\u0000 &lt;annotation&gt;$sim$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;25 MeV/n.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;The analytical expression for &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mi&gt;β&lt;/mi&gt;\u0000 &lt;annotation&gt;$beta$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt; was derived starting from the formalism of the NanOx biophysical model and introducing a set of approximations. The latter consider that the irradiation is carried out under track-segment conditions (as","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transforming [177Lu]Lu-PSMA-617 treatment planning: Machine learning-based radiodosiomics and swin UNETR using pretherapy PSMA positron emission tomography/computed tomography (PET/CT)","authors":"Elmira Yazdani,&nbsp;Aryan Neizehbaz,&nbsp;Najme Karamzade-Ziarati,&nbsp;Farshad Emami,&nbsp;Habibeh Vosoughi,&nbsp;Mahboobeh Asadi,&nbsp;Atefeh Mahmoudi,&nbsp;Mahdi Sadeghi,&nbsp;Saeed Reza Kheradpisheh,&nbsp;Parham Geramifar","doi":"10.1002/mp.70030","DOIUrl":"10.1002/mp.70030","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Personalized pretreatment dosimetry planning is crucial for optimizing [&lt;sup&gt;177&lt;/sup&gt;Lu]Lu–prostate-specific membrane antigen-617 (Lu-PSMA) radioligand therapy (RLT) in patients with metastatic castration-resistant prostate cancer (mCRPC).&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;This study addresses two goals. First, we develop a machine learning (ML)-based pretreatment planning model to predict post-therapy absorbed doses (ADs) in metastatic lesions by integrating clinical biomarkers (CBs) with radiomic features (RFs) and dosiomic features (DFs) extracted from [⁶⁸Ga]Ga-PSMA-11 (Ga-PSMA) positron emission tomography/computed tomography (PET/CT), thereby improving predictive accuracy. Second, we develop a transformer-based deep learning (DL) architecture to predict Monte Carlo (MC)-derived dose rate maps (DRMs), minimizing reliance on computationally intensive MC simulations.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;For the ML objective, retrospective posttreatment dosimetry data from 20 patients with mCRPC treated with Lu-PSMA RLT were used as ground truth labels. Patient-specific MC dosimetry was employed on Ga-PSMA PET/CT images using the GATE v9.1 toolkit to generate DRMs. After image preprocessing, RFs and DFs were extracted from Ga-PSMA CT images and DRMs using LIFEx v7.4.0. Multiple feature selection techniques, including recursive feature elimination (RFE), mutual information, Boruta, LASSO, and Elastic Net, were applied and evaluated. The Benjamini-Hochberg correction (&lt;i&gt;q&lt;/i&gt; &lt; 0.05) was used to control for false discovery rate following each method. Multiple nonlinear regression models were trained using leave-one-out cross-validation (LOOCV), and model interpretability was assessed using SHAP and LIME radar plots. A shifted windows UNET Transformers (Swin UNETR) architecture with self-supervised learning (SSL) pretraining was employed to predict voxel-wise PET-based DRMs for the DL objective. The model was fine-tuned on MC-labelled DRM data from 30 patients (including 10 additional cases) using 5-fold cross-validation.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Among multiple feature selection strategies, RFE was ultimately selected for final modelling based on its superior predictive performance. The ensemble tree regressor (ETR) using selected CT RFs, PET DFs, and significant CBs achieved an R&lt;sup&gt;2&lt;/sup&gt; = 0.82 and RMSE = 0.67 Gy/GBq. For DRM prediction, the SSL-pretrained Swin UNETR achieved an R&lt;sup&gt;2&lt;/sup&gt; of 0.97, NRMSE of 0.003 Gy/GBq, and a Gamma pass rate of 99.08%, closely matching MC-derived DRMs.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;sectio","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing semi-supervised learning for fine-grained 3D cerebrovascular segmentation with cross-consistency and uncertainty estimation","authors":"Yousuf Babiker M. Osman,&nbsp;Cheng Li,&nbsp;Nazik Elsayed,&nbsp;Alou Diakite,&nbsp;Shuqiang Wang,&nbsp;Shanshan Wang","doi":"10.1002/mp.70017","DOIUrl":"10.1002/mp.70017","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Background&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Accurate delineation of the cerebral blood vessel from time-of-flight magnetic resonance angiography (TOF-MRA) data is essential to the analysis, diagnosis, and treatment of pathologies related to the cerebral blood supply. The limitations of supervised deep learning approaches in terms of annotation cost and applicability necessitate the exploration of alternative approaches that can effectively address these challenges and facilitate the real-world clinical deployment of automatic 3D cerebrovascular segmentation.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Purpose&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;To address the challenges of limited labeled data by exploiting the intricate structures of vessels and developing a method to assess the reliability of generated pseudo-labels, with the ultimate goal of enhancing the efficiency of unlabeled data utilization and improving segmentation accuracy.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Methods&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;We introduce a cross-consistency dual uncertainty quantification mean teacher method for semi-supervised learning fine-grained 3D cerebrovascular segmentation from TOF-MRA images. To effectively incorporate knowledge from unlabeled samples, we present a dual-consistency learning approach that jointly pertains to pixel-image transformation consistent equivariant and feature perturbation invariance. Following that, in an attempt to guarantee more confidence in unsupervised learning, we evaluate the segmentation uncertainty using the predictions from both the student and teacher models and employ them in collaboration for guiding consistency regularization. Additionally, we boost the pixel-level prediction performance by employing a region-specific supervised loss only for the annotated input samples.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Results&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;Quantitative and qualitative results on two publicly available datasets show that the proposed method yielded better results than state-of-the-art semi-supervised learning methods for cerebrovascular segmentation. Specifically, our method achieved a dice similarity coefficient of 83.3% and intersection-over-union of 71.5% on the IXI dataset, surpassing the baseline uncertainty-aware mean teacher method by 1.7% and 2.8%, respectively.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 \u0000 &lt;section&gt;\u0000 \u0000 &lt;h3&gt; Conclusion&lt;/h3&gt;\u0000 \u0000 &lt;p&gt;The framework's ability to achieve competitive performance across various metrics showcases its potential for relieving human annotation efforts for accurate cerebrovascular extraction tasks, where its effectiveness in handling unlabeled data can offer significant advantages.&lt;/p&gt;\u0000 &lt;/section&gt;\u0000 ","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126822","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A deep learning framework for accurate mammographic mass classification using local context attention module","authors":"Ibrahim Abdelhalim,&nbsp;Yassir Almalki,&nbsp;Abdelrahman Abdallah,&nbsp;Rasha Karam,&nbsp;Sharifa Alduraibi,&nbsp;Mohammad Basha,&nbsp;Hassan Mohamed,&nbsp;Mohammed Ghazal,&nbsp;Ali Mahmoud,&nbsp;Norah Saleh Alghamdi,&nbsp;Sohail Contractor,&nbsp;Ayman El-Baz","doi":"10.1002/mp.18119","DOIUrl":"10.1002/mp.18119","url":null,"abstract":"<div>\u0000 \u0000 \u0000 <section>\u0000 \u0000 <h3> Background</h3>\u0000 \u0000 <p>Dense breast tissue significantly increases breast cancer (BC) risk. However, current mammographic methods for classifying BC are often subjective and unreliable, which complicates the task of accurate evaluation.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Purpose</h3>\u0000 \u0000 <p>This study introduces a deep learning method with a local context attention module (LCAM), using dual mammogram views aligned with BI-RADS to enhance grading consistency and accuracy in BC classification across four groups by leveraging local context around masses.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Methods</h3>\u0000 \u0000 <p>Specific regions of interest (ROIs) containing dense tissue around breast masses are identified from dual mammogram views, providing additional insights for predicting BC BI-RADS categories. These ROIs are then input into a convolutional neural network (CNN)-based model, which is crucial for selecting and differentiating radiomic features associated with BI-RADS. To enhance our model's ability to distinguish salient radiomic features associated with mass malignancy, the LCAM sequentially infers attention maps along two separate dimensions: channel and spatial. These attention maps are subsequently multiplied with the input feature map for adaptive feature refinement.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Results</h3>\u0000 \u0000 <p>Examining 3020 patients across four BI-RADS categories while leveraging dual mammogram views demonstrates the robust performance of the proposed framework, achieving a sensitivity of 82.46% and a specificity of 91.42% in identifying BI-RADS grading relevant to breast masses.</p>\u0000 </section>\u0000 \u0000 <section>\u0000 \u0000 <h3> Conclusions</h3>\u0000 \u0000 <p>We introduced a novel CNN-based framework that utilizes dual mammogram views for the BC classification. It utilizes LCAM, which further understands the local characteristics surrounding breast masses, aiming to enhance the accuracy and consistency of classification outcomes.</p>\u0000 </section>\u0000 </div>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 10","pages":""},"PeriodicalIF":3.2,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145126790","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}