Computer methods and programs in biomedicine最新文献

{"title":"Impact of carotid web orientation on long-term thrombus growth risk: A numerical study using fluid-solid interaction","authors":"Songling Fu ,&nbsp;Zhaokun Li ,&nbsp;Xinhui Liu ,&nbsp;Pan Song ,&nbsp;Mingyuan Liu ,&nbsp;Jun Wen","doi":"10.1016/j.cmpb.2025.109049","DOIUrl":"10.1016/j.cmpb.2025.109049","url":null,"abstract":"<div><h3>Background</h3><div>Carotid web (CaW) is a rare fibromuscular dysplasia lesion at the carotid bifurcation linked to thromboembolic events in young patients. CaW-induced hemodynamic disturbances contribute to thrombosis, but the impact of CaW morphology on long-term thrombotic risk remains unclear.</div></div><div><h3>Method</h3><div>This study developed three-dimensional numerical models based on patient-specific carotid artery anatomy with CaW angles of 30°, 60°, and 90° (models A, B, and C). Unlike prior studies assuming rigid walls, this work incorporated vessel wall elasticity and thrombus growth mechanisms. Fluid–structure interaction simulations analyzed wall deformation, stress, and hemodynamic parameters to assess the effects of CaW angles and wall elasticity on thrombus risk.</div></div><div><h3>Results</h3><div>Simulations showed that larger CaW angles increased maximum wall deformation at the bifurcation and web region by ∼5 % and maximum stress in the web area by &gt;10 %, indicating higher mechanical load at 90° Flow field analysis revealed that larger angles reduced internal carotid artery velocity and generated recirculation zones. Wall shear stress metrics indicated expanded low-shear and high-residence-time regions in model C, suggesting complex flow patterns. Thrombus growth models showed pronounced thrombus formation at 60° and 90°, with rigid wall simulations underestimating thrombus areas by 15.7 %, 44.5 %, and 85.7 % for models A, B, and C, respectively.</div></div><div><h3>Conclusion</h3><div>Larger CaW angles are associated with increased flow disturbances and elevated thrombus risk. Variations in the web angle influence thrombus distribution, and incorporating vessel wall elasticity into the analysis enhances risk assessment. Clinically, evaluating carotid geometry and wall elasticity can help optimize thrombus risk evaluation and guide intervention strategies.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109049"},"PeriodicalIF":4.8,"publicationDate":"2025-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145010935","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":"Artificial intelligence assisted ultrasound for selective screening of hip dysplasia at children’s health care centers in the Netherlands","authors":"Maartje M.R. Verhoeven ,&nbsp;Willemijn M. Klein ,&nbsp;Wouter de Monye ,&nbsp;Fleur Kersten ,&nbsp;Lucy J. Smit ,&nbsp;Emilie L.M. Ruiter ,&nbsp;Chris L. de Korte ,&nbsp;Thomas L.A. van den Heuvel","doi":"10.1016/j.cmpb.2025.109047","DOIUrl":"10.1016/j.cmpb.2025.109047","url":null,"abstract":"<div><h3>Objectives</h3><div>To investigate whether adding an artificial intelligence-assisted hip ultrasound (CHC-US) to the selective screening for developmental dysplasia of the hip (DDH) at Child Health Care (CHC) centers can reduce the number of referrals for hip ultrasound without an increase of missed cases.</div></div><div><h3>Methods</h3><div>We conducted a cross-sectional diagnostic study at four CHC centers in the Netherlands between May 2022 and December 2022. All participating infants received both a CHC-US and a Hospital-US. The results of both ultrasounds (US) were analysed to evaluate the diagnostic performance of the CHC-US. Descriptive statistics, frequencies, and linear weighted Cohen’s Kappa were used for statistical analysis.</div></div><div><h3>Results</h3><div>Data from 105 infants (210 hips) were included. In 59 infants (56.2 %) both hips were classified as not having DDH according to the CHC-US, of which two hips were diagnosed with DDH according to the Hospital-US. However, re-evaluation of the Hospital-US images showed no DDH in both cases.</div></div><div><h3>Conclusions</h3><div>This study demonstrates that CHC-US in selective screening of DDH could reduce the number of hip US referrals by approximately half. CHC-US can be performed by a CHC physician, making integration of US into routine infant welfare visits at CHC-centers feasible.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109047"},"PeriodicalIF":4.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145046607","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 weakly-supervised deep learning model for fast localisation and delineation of the skeleton, internal organs, and spinal canal on whole-body diffusion-weighted MRI (WB-DWI)","authors":"Antonio Candito ,&nbsp;Alina Dragan ,&nbsp;Richard Holbrey ,&nbsp;Ana Ribeiro ,&nbsp;Ricardo Donners ,&nbsp;Christina Messiou ,&nbsp;Nina Tunariu ,&nbsp;Dow-Mu Koh ,&nbsp;Matthew D Blackledge","doi":"10.1016/j.cmpb.2025.109043","DOIUrl":"10.1016/j.cmpb.2025.109043","url":null,"abstract":"<div><h3>Background and Objective</h3><div>Apparent Diffusion Coefficient (ADC) values and Total Diffusion Volume (TDV) from Whole-body diffusion-weighted MRI (WB-DWI) are recognised cancer imaging biomarkers. However, manual disease delineation for ADC and TDV measurements is unfeasible in clinical practice, demanding automation. As a first step, we propose an algorithm to generate fast and reproducible probability maps of the skeleton, adjacent internal organs (liver, spleen, urinary bladder, and kidneys), and spinal canal.</div></div><div><h3>Methods</h3><div>We developed an automated deep-learning pipeline based on a 3D patch-based Residual U-Net architecture that localises and delineates these anatomical structures on WB-DWI. The algorithm was trained using “soft-labels” (non-binary segmentations) derived from a computationally intensive atlas-based approach. For training and validation, we employed a multi-centre WB-DWI dataset comprising 532 scans from patients with Advanced Prostate Cancer (APC) or Multiple Myeloma (MM), with testing on 45 patients.</div></div><div><h3>Results</h3><div>Our weakly-supervised deep learning model achieved an average dice score of 0.67 for whole skeletal delineation, 0.76 when excluding ribcage, 0.83 for internal organs, and 0.86 for spinal canal, with average surface distances below 3 mm. Relative median ADC differences between automated and manual full-body delineations were below 10 %. The model was 12x faster than the atlas-based registration algorithm (25 s vs. 5 min). Two experienced radiologists rated the model’s outputs as either “good” or “excellent” on test scans, with inter-reader agreement from fair to substantial (Gwet’s AC1=0.27–0.72).</div></div><div><h3>Conclusion</h3><div>The model offers fast, reproducible probability maps for localising and delineating body regions on WB-DWI, potentially enabling non-invasive imaging biomarkers quantification to support disease staging and treatment response assessment.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109043"},"PeriodicalIF":4.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145019201","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":"Heart failure diagnosis and ejection fraction classification via feature fusion model using non-contact vital sign signals","authors":"Shen Feng ,&nbsp;Xianda Wu ,&nbsp;Huan Cen ,&nbsp;Sinan Chen ,&nbsp;Baoxian Yu ,&nbsp;Zhiqiang Pang ,&nbsp;Pengtao Sun ,&nbsp;Han Zhang","doi":"10.1016/j.cmpb.2025.109031","DOIUrl":"10.1016/j.cmpb.2025.109031","url":null,"abstract":"<div><h3>Background and objectives:</h3><div>Ballistocardiography (BCG) has emerged as a promising modality for home-based heart failure (HF) monitoring, yet existing single-dimensional manual feature analyses fail to adequately characterize left ventricular ejection fraction (LVEF <span><math><mi>&lt;</mi></math></span> 40%) dynamics. We address this limitation by developing a hybrid feature fusion framework that synergizes manual feature engineering with deep learning for improved HF diagnosis and LVEF classification.</div></div><div><h3>Methods:</h3><div>83 participants were recruited from a hospital, with their samples categorized into two (healthy and HF) and three classes (healthy, LVEF <span><math><mo>≥</mo></math></span> 40% HF, and LVEF <span><math><mi>&lt;</mi></math></span> 40% HF) based on clinical diagnosis. Non-contact vital signs were collected from supine participants using a piezoelectric sensor, and the BCG and respiratory signals were isolated using filters. We developed a model that integrates manual with deep features extracted from BCG and respiratory signals, to enhance the accuracy of HF diagnosis and LVEF classification. Additionally, we designed a multi-scale ResNet-BiLSTM network model to extract deep features from the signals, effectively capturing dynamic changes and intrinsic patterns across various time scales.</div></div><div><h3>Results:</h3><div>Ablation experiments show that the proposed method outperforms traditional manual methods, achieving classification accuracies of 98.20% and 98.76% for two and three-class HF classification under five-fold cross-validation, respectively.</div></div><div><h3>Conclusions:</h3><div>This study establishes a healthcare-oriented framework for at-home diagnosis of HF and LVEF classification, facilitating rapid preliminary screening and auxiliary diagnosis in non-clinical settings.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109031"},"PeriodicalIF":4.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934006","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":"Haemodynamic numerical simulation of hybrid surgical repairs for thoracoabdominal aortic aneurysms","authors":"Luca Bertoglio ,&nbsp;Stefano Bonardelli ,&nbsp;Giuseppe Dalla Vecchia ,&nbsp;Antonio Ghidoni ,&nbsp;Gianmaria Noventa ,&nbsp;Marco Ravanelli","doi":"10.1016/j.cmpb.2025.108998","DOIUrl":"10.1016/j.cmpb.2025.108998","url":null,"abstract":"<div><h3>Background and objective</h3><div>: The hybrid surgical repair is a feasible alternative to conventional open surgical or total endovascular repairs for thoracoabdominal aneurysms. However, a small number of patients are treated every year with this procedure, and for this reason, a negligible amount of numerical or measured data is available in the literature. Moreover, the complex and highly variable stent graft design in hybrid surgical repairs means that a priori prediction of haemodynamic flow parameters and clinical or surgical outcomes remain challenging. The goal of this work is to appraise the clinical relevance of computational fluid dynamics and numerical results in the setting of hybrid surgical repairs.</div></div><div><h3>Methods:</h3><div>Numerical simulations are carried out on three patients with a large range of elements of the meshes to assess the spatial convergence of the result. Flow rates and geometries are calculated and reconstructed in the post-operative conditions with phase-contrast magnetic resonance imaging.</div></div><div><h3>Results:</h3><div>Numerical results demonstrate higher accuracy with respect to measurements. In fact, the measured outflow rates are not able to match the measured inflow rate in any patient. From the point of view of the spatial convergence of the results, the acceptable mesh depends on the quantity of interest, e.g., (<span><math><mi>i</mi></math></span>) in terms of the time-averaged outflow rates, the mesh with 4 <span><math><mo>×</mo></math></span> 10⁶ elements is acceptable, (<span><math><mrow><mi>i</mi><mi>i</mi></mrow></math></span>) in terms of the maximum values in the distribution of the wall shear stresses, the mesh with 16 <span><math><mo>×</mo></math></span> 10⁶ elements is acceptable, while (<span><math><mrow><mi>i</mi><mi>i</mi><mi>i</mi></mrow></math></span>) in terms of the numerical dissipation, only the mesh with 64 <span><math><mo>×</mo></math></span> 10⁶ elements is acceptable for all patients.</div></div><div><h3>Conclusions</h3><div>: Numerical results demonstrate that computational fluid dynamics can be used, especially in hybrid surgical repairs, to generate potentially actionable predictive insights with implications for surveillance and enhanced postoperative management. Moreover, numerical results can also be used in preoperative surgical planning coupled with geometry optimization algorithms to identify the best designs for the patient.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"271 ","pages":"Article 108998"},"PeriodicalIF":4.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144916683","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":"DECODE: An open-source cloud-based platform for the noninvasive management of peripheral artery disease","authors":"Mohammed A. AboArab ,&nbsp;Miloš Anić ,&nbsp;Vassiliki T. Potsika ,&nbsp;Hassan Saeed ,&nbsp;Manahil Zulfiqar ,&nbsp;Andrzej Skalski ,&nbsp;Elisabetta Stretti ,&nbsp;Vassilis Kostopoulos ,&nbsp;Spyridon Psarras ,&nbsp;Giancarlo Pennati ,&nbsp;Francesca Berti ,&nbsp;Lemana Spahić ,&nbsp;Leo Benolić ,&nbsp;Nenad Filipović ,&nbsp;Dimitrios I. Fotiadis","doi":"10.1016/j.cmpb.2025.109037","DOIUrl":"10.1016/j.cmpb.2025.109037","url":null,"abstract":"<div><div><strong>Background and Objective:</strong> Peripheral artery disease (PAD) is a progressive vascular condition affecting &gt;237 million individuals worldwide. Accurate diagnosis and patient-specific treatment planning are critical but are often hindered by limited access to advanced imaging tools and real-time analytical support. This study presents DECODE, an open-source, cloud-based platform that integrates artificial intelligence, interactive 3D visualization, and computational modeling to improve the noninvasive management of PAD.</div><div><strong>Methods:</strong> The DECODE platform was designed as a modular backend (Django) and frontend (React) architecture that combines deep learning–based segmentation, real-time volume rendering, and finite element simulations. Peripheral artery and intima–media thickness segmentation were implemented via convolutional neural networks, including extended U-Net and nnU-Net architectures. Centreline extraction algorithms provide quantitative vascular geometry analysis. Balloon angioplasty simulations were conducted via nonlinear finite element models calibrated with experimental data. Usability was evaluated via the System Usability Scale (SUS), and user acceptance was assessed via the Technology Acceptance Model (TAM).</div><div><strong>Results:</strong> Peripheral artery segmentation achieved an average Dice coefficient of 0.91 and a 95th percentile Hausdorff distance 1.0 mm across 22 computed tomography dataset. Intima-media segmentation evaluated on 300 intravascular optical coherence tomography images demonstrated Dice scores 0.992 for the lumen boundaries and 0.980 for the intima boundaries, with corresponding Hausdorff distances of 0.056 mm and 0.101 mm, respectively. Finite element simulations successfully reproduced the mechanical interactions between balloon and artery models in both idealized and subject-specific geometries, identifying pressure and stress distributions relevant to treatment outcomes. The platform received an average SUS score 87.5, indicating excellent usability, and an overall TAM score 4.21 out of 5, reflecting high user acceptance.</div><div><strong>Conclusions:</strong> DECODE provides an automated, cloud-integrated solution for PAD diagnosis and intervention planning, combining deep learning, computational modeling, and high-fidelity visualization. The platform enables precise vascular analysis, real-time procedural simulation, and interactive clinical decision support. By streamlining image processing, enhancing segmentation accuracy, and enabling <em>in-silico</em> trials, DECODE offers a scalable infrastructure for personalized vascular care and sets a new benchmark in digital health technologies for PAD.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109037"},"PeriodicalIF":4.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144925996","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":"Exploring foundation models for multi-class muscle segmentation in MR images of neuromuscular disorders: A comparative analysis of accuracy and uncertainty","authors":"Nicola Casali ,&nbsp;Alessandro Brusaferri ,&nbsp;Giuseppe Baselli ,&nbsp;Marco Moscatelli ,&nbsp;Domenico Aquino ,&nbsp;Marina Grisoli ,&nbsp;Giovanna Rizzo ,&nbsp;Alfonso Mastropietro","doi":"10.1016/j.cmpb.2025.109035","DOIUrl":"10.1016/j.cmpb.2025.109035","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Deep learning (DL) models have shown promise for skeletal muscle (SM) segmentation in MR images, which is crucial for extracting biomarkers in neuromuscular disorders (NMDs). However, to ensure safe clinical use, models should provide uncertainty estimates, allowing radiologists to assess predictions and intervene when needed. Foundation Models (FMs) have the potential to play a significant role due to their strong generalization capabilities and well-calibrated predictions. However, their applicability in this context has not yet been explored. This study aims to develop an accurate and trustworthy technique by fine-tuning FMs to delineate fatty-infiltrated SM fascicles in NMD patients.</div></div><div><h3>Methods:</h3><div>We fine-tuned Segment Anything Model (SAM) and MedSAM using two configurations – encoder/decoder and decoder only – and compared their performance against state-of-the-art 2D and 3D nnU-Net using a dataset of thigh MR images from 76 NMD patients, categorized into Early, Moderate, and Severe fatty infiltration groups. Accuracy was evaluated using the Dice Similarity Coefficient (DSC), while Uncertainty Quantification (UQ) was evaluated using the Expected Calibration Error (ECE) and the Negative Log-Likelihood (NLL). Deep Ensembles were used to convey epistemic uncertainty in addition to the aleatoric counterpart.</div></div><div><h3>Results:</h3><div>SAM’s fine-tuned encoder/decoder outperformed nnU-Net 3D in Moderate and Severe cases (DSC: 0.886 vs 0.883 and 0.857 vs 0.850) and was comparable in Early (DSC: 0.925). MedSAM did not show an advantage over SAM. Regarding UQ, SAM exhibited superior calibration in Moderate and Severe groups (ECE: 3.6% vs. 5.1% and 3.3% vs. 7.1%),</div></div><div><h3>Conclusions:</h3><div>In conclusion, our findings demonstrate that fine-tuning SAM yields superior performance, considering both accuracy and UQ metrics, highlighting its enhanced reliability in challenging NMD imaging scenarios.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109035"},"PeriodicalIF":4.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144934005","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":"In silico comparison of two non-invasive pre-procedural virtual coronary revascularisation techniques for personalised cardiovascular medicine","authors":"Mengzhe Lyu ,&nbsp;Ryo Torii ,&nbsp;Ce Liang ,&nbsp;Qiaoqiao Li ,&nbsp;Xifu Wang ,&nbsp;Yiannis Ventikos ,&nbsp;Duanduan Chen","doi":"10.1016/j.cmpb.2025.109046","DOIUrl":"10.1016/j.cmpb.2025.109046","url":null,"abstract":"<div><h3>Background and objectives</h3><div>Non-invasive pre-procedural prediction of post-PCI vessel morphology and CT angiography–derived fractional flow reserve (CT-FFR) can inform coronary revascularisation planning. However, the capabilities of different CT-based virtual coronary revascularisation (VCR) techniques need further investigation.</div></div><div><h3>Methods</h3><div>This study compared two CT-based VCR techniques: a virtual coronary intervention (VCI) method and a radius correction (RC) method. The two techniques applied to 9 vessel cases were examined according to the accuracy of luminal cross-section area, luminal centreline curvature and predicted post-PCI CT-FFR. Post-PCI computed tomography angiography reference standard were used for further validation.</div></div><div><h3>Results</h3><div>The measured post-PCI cross-sectional area was 18.74 ± 4.30 mm². The VCI-predicted area was 17.29 ± 3.48 mm² (mean difference: −1.45 ± 1.96 mm²; limits of agreement: −5.29 to 2.38), whereas the RC-predicted area was 9.42 ± 1.30 mm² (mean difference: −9.32 ± 3.78 mm²; limits of agreement: −16.72 to −1.92). The measured post-PCI centreline curvature was 0.16 ± 0.02 mm^-1. VCI predicted 0.15 ± 0.04 mm⁻¹ (mean difference: −0.01 ± 0.05 mm⁻¹; limits of agreement: −0.12 to 0.09), whereas RC predicted 0.24 ± 0.07 mm⁻¹ (mean difference: 0.08 ± 0.07 mm⁻¹; limits of agreement: −0.05 to 0.21). The post-PCI CCTA-derived CT-FFR (functional reference) was 0.92 ± 0.09. VCI predicted 0.90 ± 0.08 (mean difference: −0.02 ± 0.03; limits of agreement: −0.08 to 0.04) and RC predicted 0.90 ± 0.06 (mean difference: −0.02 ± 0.05; limits of agreement: −0.12 to 0.09).</div></div><div><h3>Conclusions</h3><div>Both non-invasive, pre-procedural techniques showed good numerical agreement with computational post-PCI CT-FFR in this pilot cohort. However, the VCI method outperformed the RC method in predicting luminal cross-sectional area and luminal centreline curvature. The cross-sectional area of the stented vessel was underestimated, and the average curvature was overestimated in the RC method.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109046"},"PeriodicalIF":4.8,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144925997","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":"BPMambaMIL: A bio-inspired prototype-guided multiple instance learning for oncotype DX risk assessment in histopathology","authors":"Yongxin Guo ,&nbsp;Ziyu Su ,&nbsp;Onur C. Koyun ,&nbsp;Hao Lu ,&nbsp;Robert Wesolowski ,&nbsp;Gary Tozbikian ,&nbsp;M. Khalid Khan Niazi ,&nbsp;Metin N. Gurcan","doi":"10.1016/j.cmpb.2025.109039","DOIUrl":"10.1016/j.cmpb.2025.109039","url":null,"abstract":"<div><div>Breast cancer remains one of the most prevalent malignancies among women, with hormone receptor-positive (HR+)/human epidermal growth factor receptor 2-negative (HER2–) breast cancers constituting a majority, with treatment decisions often guided by genomic assays such as the 21-gene recurrence score assay, Oncotype DX. Although Oncotype DX provides critical prognostic and predictive insights, its high cost and limited accessibility create substantial barriers, especially for patients with constrained financial resources. To reduce the test cost, we aim to leverage H&amp;E-stained whole slide images (WSIs) to predict Oncotype DX risk. Since WSIs are extremely large and contain redundant information, directly processing them is both computationally expensive and prone to errors. To address these limitations, we introduce a bio-inspired prototype-guided model (BPMambaMIL), a novel weakly supervised learning framework that integrates the Mamba mechanism with prototypical guidance to predict Oncotype DX score intervals directly from pathology images. Our model was evaluated on an in-house dataset with clinical Oncotype DX scores, where it achieved an AUC of 0.839, a 5.61 % improvement over the baseline model (MambaMIL), and demonstrated robust predictive performance, particularly in identifying high-risk score ranges (accuracy: 0.714 vs 0.419). Further assessments on two public breast cancer pathology image datasets using six state-of-the-art models underscored BPMambaMIL’s generalizability on research-based ODX scores and binary tumor classification tasks. By evaluating various clinical scenarios, the proposed method not only enhances the accuracy of breast cancer recurrence risk predictions but also offers a cost-effective alternative to genomic assays, thus improving clinical outcomes.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"272 ","pages":"Article 109039"},"PeriodicalIF":4.8,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145039321","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 boundary condition remapping technique for computational fluid dynamics simulations of nasal exhalation","authors":"Matthew Cook,&nbsp;Sara Vahaji,&nbsp;Kiao Inthavong","doi":"10.1016/j.cmpb.2025.109028","DOIUrl":"10.1016/j.cmpb.2025.109028","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>During exhalation, complex geometry in the larynx generates the pharyngeal jet, where higher velocity air is directed to the rear of the airway, influencing airflow downstream in the nasal passage. This study investigates the impact of boundary condition settings on the accuracy of airflow simulations in truncated airway geometries during exhalation, focusing on the nasopharynx and nasal passage. In addition to traditional inlet profiles, we tested a new method of remapping a profile from a complete airway to a truncated airway.</div></div><div><h3>Methods:</h3><div>Using remapped velocity profiles extracted from simulations of complete airways, we compare their performance against traditional inlet profiles, including uniform, parabolic, and power-law profiles. The performance of inlet profiles was tested in a set of airway geometries varying the inlet extension length, angle, and cross section.</div></div><div><h3>Results:</h3><div>The results demonstrate that the remapped boundary condition provides the most accurate representation of downstream flow, particularly in replicating the pharyngeal jet and lateral asymmetry, with minimal error at bends and bifurcations in the nasal passage when applied to a truncated version of the airway it is sourced from. The study highlights the limitations of extended geometries, showing that shorter extensions (3×D) yield lower errors than longer ones (5×D). Angling the extensions further improves accuracy by smoothing abrupt transitions. The parabolic profile offers no significant advantage over uniform and power-law profiles, emphasising the importance of selecting an appropriate boundary condition in capturing the complex flow characteristics of the laryngeal jet.</div></div><div><h3>Conclusions:</h3><div>The proposed remapping technique generalises irregular velocity profiles to circular cross-sections, enabling their application across varied geometries without requiring full airway simulations. This method improves computational efficiency and accuracy, making it adaptable for diverse respiratory, cardiovascular, and industrial applications. Future work must examine a larger sample of larynx and nasal passage geometries to further validate the use of this technique in applying profiles across geometries.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"271 ","pages":"Article 109028"},"PeriodicalIF":4.8,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144916557","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}