Biomedical Physics & Engineering Express最新文献

{"title":"Tomographic quality evaluation of flexible x-ray scintillating imaging skins.","authors":"Rozenn Raffaut, Robert Moss, Agostino Stilli, Danail Stoyanov","doi":"10.1088/2057-1976/aea910","DOIUrl":"10.1088/2057-1976/aea910","url":null,"abstract":"<p><p><i>Objective.</i>Detecting abnormal tissue in minimally invasive surgery is challenging due to the lack of direct palpation for the surgeon, with visualisation often being limited to white light endoscope images of the tissue surface, providing no information about inner structures. Imaging skins are being explored as a novel method of surgical image guidance: they are flexible and stretchable scintillating films that can act as indirect x-ray detectors when placed directly on the organ surface.<i>Approach.</i>Tomographic imaging of a custom quality evaluation phantom was performed using the imaging skin system. Image quality metrics, including spatial resolution, contrast, and uniformity, were obtained and compared to a conventional flat-panel micro-CT system.<i>Main results.</i>The imaging skin scan achieved a spatial resolution of 3.11 lp mmat 10% contrast. While these results were worse than that of the micro-CT system, the imaging skin system still met the requirements for clinical CT, demonstrating its clinical suitability.<i>Significance.</i>The feasibility of 3D x-ray imaging using imaging skins directly placed onto the imaged object has been demonstrated, achieving high resolution, comparable to clinical CT. Future work will involve artefact corrections and more advanced reconstruction techniques for the intended application of imaging skins for real-time surgical guidance.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148987698","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":"Inter-species model selection for translational analysis of [¹⁷⁷Lu]Lu-OPS201 biokinetic parameters.","authors":"Nur Rahmah Hidayati, Annisa Rachman, Deni Hardiansyah, Bisma Barron Patrianesha, Nunung Nuraeni, Supriyanto Ardjo Pawiro","doi":"10.1088/2057-1976/aea7a3","DOIUrl":"10.1088/2057-1976/aea7a3","url":null,"abstract":"<p><p>Robust estimation of time-integrated activity coefficients (TIACs) is essential for internal dosimetry in radiopharmaceutical therapy. This study developed a structured translational framework for predicting the kidney biokinetics and TIAC of [¹⁷⁷Lu]Lu-OPS201 in humans from preclinical biodistribution data in mice and pigs. Candidate sum-of-exponential functions (SoEFs) were fitted to the mean kidney time-activity data from mice, pigs, and humans. The most suitable SoEF was selected using sequential goodness-of-fit criteria, including visual assessment of the fitted curves and parameter coefficients of variation (CV) below 50%, followed by comparison of the weighted corrected Akaike information criterion (%wAICc) among the evaluable models. The reference human TIAC (rTIAC) was calculated from the selected model fitted to the human data. Human TIACs were then predicted using 5 previously reported interspecies scaling methods and a proposed approach based on mass scaling of the SoEF parameters derived from the mouse and pig data. Predictive performance was evaluated using the relative deviation (RD) between each predicted TIAC (pTIAC) and the rTIAC. A 2-parameter SoEF provided the most suitable description of the kidney time-activity data, with all parameters estimated precisely (CV < 50%) and %wAICc ∼100%. The resulting rTIAC for human kidneys was 5.41 ± 0.36 h. The proposed SoEF-parameter mass-scaling method showed the closest agreement with the human reference, yielding a pTIAC of 5.93 ± 2.08 h, corresponding to a RD of 9.67%, the smallest deviation among all 6 translational approaches evaluated. Combining systematic biokinetic model selection with SoEF-parameter mass scaling may improve preclinical-to-clinical TIAC prediction for [¹⁷⁷Lu]Lu-OPS201 kidney dosimetry. However, the optimal model and scaling method are likely specific to the radiopharmaceutical, target organ, and species evaluated. This reproducible framework provides a hypothesis-generating approach for early-stage human dosimetry estimation and supports further validation of translational methods in radiopharmaceutical therapy.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148975669","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 pneumatic McKibben muscle-based limb volume phantom model for wearable strain sensor validation.","authors":"Nick Levinski, Lavan Sumanan, Derry Tong, Cameron Sjaarda, Jacob Kowalski, Carolyn L Ren","doi":"10.1088/2057-1976/aea7a4","DOIUrl":"10.1088/2057-1976/aea7a4","url":null,"abstract":"<p><p>Wearable strain sensors are a promising technology for continuous health monitoring; however, existing validation approaches present a fundamental trade-off. Bench-top methods offer high resolution and repeatability but lack fidelity to real-life use conditions, whereas human-subject testing provides physiological relevance with limited resolution, repeatability and coverage across use cases. We present a novel validation platform-a limb volume phantom model (LVPM)-that bridges this gap by integrating a soft pneumatic artificial (McKibben) muscle with a compliant limb phantom and a closed-loop, camera-based control system for precise circumferential regulation. The LVPM provides a 24 cm-long region of uniform circumference along its longitudinal axis and enables programmable control of the circumference from 30-37 cm with an optical diameter resolution of approximately 1 mm, < 0.5% steady-state error, and settling times below 40 s. To demonstrate its utility, a soft, highly stretchable capacitive strain sensor with an interdigitated electrode architecture was evaluated using the LVPM and compared against conventional uniaxial tensile testing and tape-based measurements on a human limb. The LVPM achieved a measurement range and resolution comparable to uniaxial testing while producing strain-response trends that more closely resemble<i>in vivo</i>behavior, without the resolution and repeatability limitations inherent to manual measurements. These results establish the LVPM as an effective and practical validation platform for wearable strain sensors, offering a balanced combination of biomechanical compliance, controllability, repeatability, measurement range, and resolution.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148975696","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":"Efficient multi-temporal 3D learning for breast DCE-MRI classification with temporal-as-channel encoding and early spatial attention.","authors":"Thuy Nguyen Thi, Hung Le Minh","doi":"10.1088/2057-1976/aea7a5","DOIUrl":"10.1088/2057-1976/aea7a5","url":null,"abstract":"<p><p>Accurate breast cancer classification in dynamic contrast-enhanced (DCE) magnetic resonance imaging requires effective integration of volumetric morphology and temporal contrast-enhancement kinetics. In this study, we evaluate a multi-temporal 3D framework for three-class breast classification (normal, benign, malignant) using the multi-center ODELIA dataset. The framework uses a temporal-as-channel (TaC) input representation that maps three DCE subtraction phases onto the native three-channel input of a pretrained R3D-18 backbone, enabling early cross-phase interaction without a dedicated temporal module. A 3D Spatial attention module (SAM), adapted from the spatial branch of CBAM, is inserted at the first residual stage (SAM-L1) to refine early volumetric features. Experiments identify second-phase subtraction (Sub) as the best-performing single-volume input among those evaluated and show higher observed performance for the 3D full-volume configuration than for the 2D maximum-intensity projection configuration. TaC yields a larger gain in the 3D setting than in the 2D setting, while the attention ablations identify SAM-L1 as the best-performing attention configuration among those tested. The TaC+SAM-L1 configuration achieves an internal-test average score of 81.46%, with a micro-averaged area under the receiver operating characteristic curve of 91.50%, sensitivity at 90% specificity of 76.74%, and specificity at 90% sensitivity of 76.16%.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148975708","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":"SSCA-DIPNet: structured sparse and channel-attentive deep image prior network for highly-accelerated MRI.","authors":"Yue Yin, Da Cao, Can Hu, Yue Jiang, Run Zhang, Jinsong Xia, Wei Wang, Xiuhan Li, Lei Duan","doi":"10.1088/2057-1976/aea7a1","DOIUrl":"10.1088/2057-1976/aea7a1","url":null,"abstract":"<p><p>magnetic resonance imaging (MRI) aids diagnosis yet suffers slow scanning. k-space undersampling causes ill-posed reconstruction, with existing methods flawed. This study builds a data-efficient network to tackle high-acceleration, limited-data MRI reconstruction. We proposed the structured sparse and channel-attentive deep image prior network (SSCA-DIPNet), which integrated three innovative modules based on ISTA-Net+: (1) a channel-adaptive structured sparsity module (SSM) for differential feature preservation, (2) a channel-attention deep image prior (DIP) module for artifact repair, (3) a globally learnable fusion mechanism combined with symmetric loss to enhance model generalization. Two branch models were derived from ISTA-Net+ baseline: SS-Net with only the SSM, and CA-DIPNet with only the channel-attention DIP module. Experiments on two clinical datasets (179 training samples each) and supplementary fast MRI raw k-space validation are performed under 5×/10×/20× acceleration and evaluated via peak signal-to-noise ratio and structural similarity index. The proposed model consistently outperforms ISTA-Net+, SS-Net and CA-DIPNet across all scenarios. Using scarce data, SSCA-DIPNet yields robust reconstruction with refined anatomical details and reduced artifacts. It balances reconstruction performance and data efficiency, offering a fresh paradigm for iterative-unfolding MRI reconstruction to support clinical fast scanning under limited data.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148975251","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":"Evaluation of U-Net based architectures for synthetic CT generation from dual-contrast MRI.","authors":"Seyede Sajede Mousavi Eshkiki, Alireza Sadremomtaz","doi":"10.1088/2057-1976/aea580","DOIUrl":"10.1088/2057-1976/aea580","url":null,"abstract":"<p><p>Deep learning-based MRI-to-CT synthesis supports MR-only radiotherapy and PET/MR attenuation correction, but comparisons of U-Net variants are confounded by differences in datasets, preprocessing, training protocols, and evaluation metrics. This study performed a controlled, tissue-aware, and complexity-aware comparison of five U-Net-based architectures, including U-Net, ResU-Net, Attention U-Net, U-Net++, and attention deep residual U-Net (ADR-U-Net), for brain sCT generation from T1-weighted and FLAIR MRI in 37 subjects from the CERMEP-IDB-MRXFDG database. Subject-level five-fold cross-validation reduced split-dependent bias and prevented slice-level information leakage. Evaluation included mean absolute error (MAE), root mean square error (RMSE), peak signal-to-noise ratio, structural similarity index measure, tissue-specific errors in Hounsfield units (HU), paired Wilcoxon signed-rank tests with Bonferroni correction, and computational complexity analysis. ADR-U-Net achieved the lowest whole-image MAE of 37.38 ± 6.01 HU and RMSE of 109.72 ± 17.35 HU. Its MAE improvement was 1.60 HU relative to ResU-Net, indicating a modest gain over the closest comparator. A focused ADR-U-Net modality-ablation analysis showed lower whole-image MAE with dual-contrast input (37.38 ± 6.01 HU) than with T1-only (40.79 ± 5.00 HU) or FLAIR-only input (38.41 ± 6.30 HU). Bone remained the dominant source of HU error, and U-Net++ had the highest number of floating-point operations without achieving the best accuracy. These findings suggest that model selection should consider tissue-specific HU errors, subject-level robustness, and computational cost. Because no external validation or downstream clinical endpoint was included, the results should be interpreted as controlled within-dataset technical evidence rather than evidence of clinical readiness.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148928739","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":"Task-agnostic explainable contrastive learning model for malignant cell detection in urine cytology.","authors":"Arathy Menon N P, Ram S Iyer, Pournami P N, Jayaraj P B, Pranab Dey","doi":"10.1088/2057-1976/aea429","DOIUrl":"10.1088/2057-1976/aea429","url":null,"abstract":"<p><p>Bladder cancer remains one of the most prevalent urological malignancies, where early detection is critical for improving patient outcomes. Urine cytology provides a non-invasive and cost-effective screening modality; however, the development of automated diagnostic systems is challenged by the scarcity of expert-annotated data and the limited interpretability of deep learning models. To address these challenges, we propose EnBCDet, an explainable self-supervised framework for malignant cell detection in urine cytology images. The proposed approach leverages contrastive self-supervised learning to learn discriminative cytological representations from limited annotations and incorporates a specialized backbone within a single-stage object detection architecture for efficient malignant cell localization. In addition, we introduce a novel entropy-based explanation framework that quantitatively evaluates the information content of activation maps, enabling gradient-free, class-agnostic, and task-independent interpretation of learned feature representations. Experimental evaluation on cytological samples collected from 150 individuals demonstrates that EnBCDet outperforms existing approaches, achieving a mean precision of 0.991 and a mean recall of 0.926 while maintaining computational efficiency. Clinical validation further shows over 90% agreement with expert annotations, highlighting the practical utility of the proposed system. To the best of our knowledge, this is among the first studies to systematically interpret encoder-level representations in self-supervised cytological analysis. The proposed entropy-based framework provides robust insights into backbone feature learning, advancing both diagnostic performance and explainability for computer-aided urine cytology. The analysis code supporting this study is publicly available on GitHub at.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":"12 5","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149011820","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":"Optical activation of surface-functionalized BNNTs for dual-mode photothermal and photodynamic effects under low-intensity irradiation.","authors":"Sagar Kandel, Malsha Nanayakkara, David Diercks, Santaneel Ghosh, Somesree Mitra","doi":"10.1088/2057-1976/aea4ba","DOIUrl":"10.1088/2057-1976/aea4ba","url":null,"abstract":"<p><p>Boron nitride nanotubes (BNNTs) offer a unique combination of wide bandgap, high thermal conductivity, and chemical stability that positions them as promising candidates for emerging photothermal and photodynamic cancer therapies; however, their biomedical use has been limited by poor aqueous dispersibility and weak optical response under physiologically compatible excitation. In this work, highly water-dispersible BNNTs were synthesized through sodium dodecyl sulfate (SDS) surface functionalization, enabling, for the first time, a dual-mode optical response under low-intensity visible and ultraviolet (UV) irradiation. Comprehensive structural and compositional analyses confirmed uniform surfactant adsorption and the emergence of defect-mediated surface states that broaden optical absorption into the visible range. These modified BNNTs exhibited measurable photothermal heating under mild 520 nm excitation and, uniquely, generated potent ROS under UV exposure, leading to strong cytotoxicity even without direct nanotube-cell contact. Biocompatibility assays with PC12 neural cells demonstrated low intrinsic toxicity up to 25<i>µ</i>g ml^-1, while UV -irradiated BNNT supernatants induced marked reductions in viability and pronounced nuclear abnormalities, consistent with apoptosis driven by photochemically generated oxidative species. The discovery that SDS-functionalized BNNTs can be activated through two orthogonal optical pathways at low energy thresholds establishes a novel multifunctional nanoplatform that circumvents limitations of traditional carbon-based nanostructures and high-power actuation methods. These findings highlight the potential of BNNT systems for minimally invasive, synergistic photothermal-photodynamic therapy and underscore the need for future<i>in vivo</i>studies to evaluate therapeutic efficacy, biodistribution, and long-term biocompatibility.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148918332","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":"Cell-laden 3D bioprinted hydrogels for wound healing: cellular mechanism, bioprinting determinants and translational perspectives.","authors":"Navjeet Singh, Dimple Sethi Chopra, Ankush Kumar, Abhishek Gupta, Dhandeep Singh, Nirmal Singh","doi":"10.1088/2057-1976/aea426","DOIUrl":"10.1088/2057-1976/aea426","url":null,"abstract":"<p><p>Cell-laden three-dimensional (3D) bioprinted hydrogels have emerged as a promising technique for enhanced wound healing by integrating biomaterials, living cells, and precise fabrication technology. Chronic wounds that persist for an extended period of time, especially when pathologically affected by diseases like diabetes, can pose a serious threat to patients' health. Conventional dressing materials typically fail to provide adequate mechanical strength, drug release control, and complete tissue regeneration. In contrast, 3D bioprinting can allow for the selective deposition of cells into the extracellular matrix-mimetic hydrogels, making it possible to construct biomimetic scaffolds for tissue regeneration. This review provides a clear overview of how key cellular components such as mesenchymal stem cells, adipose-derived stem cells (ADSCs), fibroblasts, and multicellular systems contribute to important wound healing processes like immunomodulation, angiogenesis, cell proliferation, and re-epithelialisation. It also explores how bioprinting factors, including bioink rheology, cross-linking methods, and processing conditions, influence printability, the structural stability of scaffolds, and cell survival. The close relationship between scaffold design and how cells respond is emphasised as a key factor in determining how effective the treatment will be. Although pre-clinical studies have shown promising results, bringing this research into real-world clinical use remains challenging due to issues like lack of standardisation, difficulties in scaling up, regulatory hurdles, and high costs.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148910510","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":"Limitations of simple machine-learning surrogates for cross-tissue prediction of radial177Ludose point kernels.","authors":"Houcine Ennassiri","doi":"10.1088/2057-1976/aea427","DOIUrl":"10.1088/2057-1976/aea427","url":null,"abstract":"<p><p><i>Objective.</i>To determine whether simple regularized linear regression models, relying solely on radial distance, mass density and elemental tissue composition as input features, can outperform the conventional water-kernel interpolation baseline for predicting radial dose point kernels (DPKs) ofacross a diverse set of homogeneous tissues and one additional low-density homogeneous medium.<i>Approach.</i>Radial DPKs forwere generated using OpenGATE 10 Monte Carlo simulations in 12 homogeneous media covering soft tissues, bone-like structures, and low-density lung-like tissues, together with one additional low-density homogeneous medium (g cm) evaluated separately as an out-of-distribution extrapolation test. Three scikit-learn models (RidgeCV, LassoCV, and ElasticNetCV) were trained with robust scaling and evaluated through leave-one-tissue-out (LOTO) and leave-one-family-out cross-validation, as well as on the independent low-density case. Performance was assessed using mean absolute percentage error on physical dose (), logarithmic root mean square error (), relative integral energy error, and relative errors onand, systematically compared against a water-kernel baseline obtained by linear resampling of a single simulated water DPK.<i>Main results.</i>ElasticNetCV was the best-performing linear surrogate overall, achieving a meanof 224% in LOTO validation (median 117%), followed by LassoCV and RidgeCV. However, none of the linear surrogates outperformed the water-kernel baseline on the main physical performance metrics. Mean gains relative to the baseline remained negative across tissue families, ranging from aboutpercentage points in bone-like tissues to aboutpercentage points in low-density lung-like tissues. Relative integral energy errors reached several thousand percent in lung-like media, and the separate low-density homogeneous medium also showed severe degradation (best surrogateversusfor the water-kernel baseline). Although some surrogate predictions showed lower log-domain errors than the baseline in selected cases, these improvements did not translate into better dosimetric accuracy in physical dose units.<i>Significance.</i>Simple linear machine-learning surrogates based on radial distance, mass density and macroscopic elemental composition did not improve, and often substantially worsened, the accuracy of radialDPK prediction compared with standard water-kernel interpolation. This negative benchmark delineates the limitations of basic tabular linear regression for capturing density- and composition-dependent kernel changes across a range of homogeneous media relevant to nuclear medicine dosimetry, and provides a reproducible reference framework for future development of more advanced surrogate models.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":"12 5","pages":""},"PeriodicalIF":2.0,"publicationDate":"2026-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149011841","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}