Computer methods and programs in biomedicine最新文献

{"title":"Automated neuroradiological support systems for multiple cerebrovascular disease markers — A systematic review and meta-analysis","authors":"Jesse Phitidis ,&nbsp;Alison Q. O’Neil ,&nbsp;William N. Whiteley ,&nbsp;Beatrice Alex ,&nbsp;Joanna M. Wardlaw ,&nbsp;Miguel O. Bernabeu ,&nbsp;Maria Valdés Hernández","doi":"10.1016/j.cmpb.2025.108715","DOIUrl":"10.1016/j.cmpb.2025.108715","url":null,"abstract":"<div><div>Cerebrovascular diseases (CVD) can lead to stroke and dementia. Stroke is the second leading cause of death world wide and dementia incidence is increasing by the year. There are several markers of CVD that are visible on brain imaging, including: white matter hyperintensities (WMH), acute and chronic ischaemic stroke lesions (ISL), lacunes, enlarged perivascular spaces (PVS), acute and chronic haemorrhagic lesions, and cerebral microbleeds (CMB). Brain atrophy also occurs in CVD. These markers are important for patient management and intervention, since they indicate elevated risk of future stroke and dementia. We systematically reviewed automated systems designed to support radiologists reporting on these CVD imaging findings. We considered commercially available software and research publications which identify <em>at least two</em> CVD markers. In total, we included 29 commercial products and 13 research publications. Two distinct types of commercial support system were available: those which identify acute stroke lesions (haemorrhagic and ischaemic) from computed tomography (CT) scans, mainly for the purpose of patient triage; and those which measure WMH and atrophy regionally and longitudinally. In research, WMH and ISL were the markers most frequently analysed together, from magnetic resonance imaging (MRI) scans; lacunes and PVS were each targeted only twice and CMB only once. For stroke, commercially available systems largely support the emergency setting, whilst research systems consider also follow-up and routine scans. The systems to quantify WMH and atrophy are focused on neurodegenerative disease support, where these CVD markers are also of significance. There are currently no openly validated systems, commercially, or in research, performing a comprehensive joint analysis of all CVD markers (WMH, ISL, lacunes, PVS, haemorrhagic lesions, CMB, and atrophy).</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108715"},"PeriodicalIF":4.9,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143629202","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":"The impact of training image quality with a novel protocol on artificial intelligence-based LGE-MRI image segmentation for potential atrial fibrillation management","authors":"A.K. Berezhnoy ,&nbsp;A.S. Kalinin ,&nbsp;D.A. Parshin ,&nbsp;A.S. Selivanov ,&nbsp;A.G. Demin ,&nbsp;A.G. Zubov ,&nbsp;R.S. Shaidullina ,&nbsp;A.A. Aitova ,&nbsp;M.M. Slotvitsky ,&nbsp;A.A. Kalemberg ,&nbsp;V.S. Kirillova ,&nbsp;V.A. Syrovnev ,&nbsp;K.I. Agladze ,&nbsp;V.A. Tsvelaya","doi":"10.1016/j.cmpb.2025.108722","DOIUrl":"10.1016/j.cmpb.2025.108722","url":null,"abstract":"<div><h3>Background</h3><div>Atrial fibrillation (AF) is the most common cardiac arrhythmia, affecting up to 2 % of the population. Catheter ablation is a promising treatment for AF, particularly for paroxysmal AF patients, but it often has high recurrence rates. Developing in silico models of patients' atria during the ablation procedure using cardiac MRI data may help reduce these rates.</div></div><div><h3>Objective</h3><div>This study aims to develop an effective automated deep learning-based segmentation pipeline by compiling a specialized dataset and employing standardized labeling protocols to improve segmentation accuracy and efficiency. In doing so, we aim to achieve the highest possible accuracy and generalization ability while minimizing the burden on clinicians involved in manual data segmentation.</div></div><div><h3>Methods</h3><div>We collected LGE-MRI data from VMRC and the cDEMRIS database. Two specialists manually labeled the data using standardized protocols to reduce subjective errors. Neural network (nnU-Net and smpU-Net++) performance was evaluated using statistical tests, including sensitivity and specificity analysis. A new database of LGE-MRI images, based on manual segmentation, was created (VMRC).</div></div><div><h3>Results</h3><div>Our approach with consistent labeling protocols achieved a Dice coefficient of 92.4 % ± 0.8 % for the cavity and 64.5 % ± 1.9 % for LA walls. Using the pre-trained RIFE model, we attained a Dice score of approximately 89.1 % ± 1.6 % for atrial LGE-MRI imputation, outperforming classical methods. Sensitivity and specificity values demonstrated substantial enhancement in the performance of neural networks trained with the new protocol.</div></div><div><h3>Conclusion</h3><div>Standardized labeling and RIFE applications significantly improved machine learning tool efficiency for constructing 3D LA models. This novel approach supports integrating state-of-the-art machine learning methods into broader in silico pipelines for predicting ablation outcomes in AF patients.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108722"},"PeriodicalIF":4.9,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143644700","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":"Light scattering imaging modal expansion cytometry for label-free single-cell analysis with deep learning","authors":"Zhi Li ,&nbsp;Xiaoyu Zhang ,&nbsp;Guosheng Li ,&nbsp;Jun Peng ,&nbsp;Xuantao Su","doi":"10.1016/j.cmpb.2025.108726","DOIUrl":"10.1016/j.cmpb.2025.108726","url":null,"abstract":"<div><h3>Background and Objective</h3><div>Single-cell imaging plays a key role in various fields, including drug development, disease diagnosis, and personalized medicine. To obtain multi-modal information from a single-cell image, especially for label-free cells, this study develops modal expansion cytometry for label-free single-cell analysis.</div></div><div><h3>Methods</h3><div>The study utilizes a deep learning-based architecture to expand single-mode light scattering images into multi-modality images, including bright-field (non-fluorescent) and fluorescence images, for label-free single-cell analysis. By combining adversarial loss, L1 distance loss, and VGG perceptual loss, a new network optimization method is proposed. The effectiveness of this method is verified by experiments on simulated images, standard spheres of different sizes, and multiple cell types (such as cervical cancer and leukemia cells). Additionally, the capability of this method in single-cell analysis is assessed through multi-modal cell classification experiments, such as cervical cancer subtypes.</div></div><div><h3>Results</h3><div>This is demonstrated by using both cervical cancer cells and leukemia cells. The expanded bright-field and fluorescence images derived from the light scattering images align closely with those obtained through conventional microscopy, showing a contour ratio near 1 for both the whole cell and its nucleus. Using machine learning, the subtyping of cervical cancer cells achieved 92.85 % accuracy with the modal expansion images, which represents an improvement of nearly 20 % over single-mode light scattering images.</div></div><div><h3>Conclusions</h3><div>This study demonstrates the light scattering imaging modal expansion cytometry with deep learning has the capability to expand the single-mode light scattering image into the artificial multimodal images of label-free single cells, which not only provides the visualization of cells but also helps for the cell classification, showing great potential in the field of single-cell analysis such as cancer cell diagnosis.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108726"},"PeriodicalIF":4.9,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143669322","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":"Mechanical stimulation of cochlea for treatment of hearing loss: Comparison between forward stimulation and reverse stimulation with an active cochlear model","authors":"Zhaohai Liu ,&nbsp;Houguang Liu ,&nbsp;Weiwei Guo ,&nbsp;Wei Chen ,&nbsp;Wen Liu ,&nbsp;Shanguo Yang","doi":"10.1016/j.cmpb.2025.108721","DOIUrl":"10.1016/j.cmpb.2025.108721","url":null,"abstract":"<div><h3>Background and objective</h3><div>Reverse stimulation is a stimulation mode of the active middle-ear implants (AMEIs), targeted at moderate conductive hearing loss and mixed hearing loss. However, previous studies investigated reverse stimulation through passive cochlear models that simulate profound sensorineural hearing loss, which is beyond the AMEI's indications. Therefore, we investigated the cochlear responses to reverse stimulation under different hearing loss and compared them with those to forward stimulation.</div></div><div><h3>Methods</h3><div>The human ear model consists of a human ear macro dynamic model, a cochlear micro dynamic model, and a cochlear circuit model. The human ear macro dynamic model and cochlear micro dynamic model were developed by simplifying the human ear tissues into stiffness, damping, and mass. The cochlear active amplification was realized by coupling the cochlear circuit model. Based on the model, the cochlear responses to forward and reverse stimulation were calculated.</div></div><div><h3>Results</h3><div>The results show that the cochlear responses to reverse stimulation are higher than those to forward stimulation, and the difference in cochlear responses decreases and then increases with increasing stimulus magnitude. Conductive hearing loss significantly reduces cochlear response to forward stimulation but has less effect on reverse stimulation. Outer hair cell hearing loss significantly reduces cochlear response to both forward and reverse stimulation, but the effect diminishes to nothing as the stimulation amplitude increases.</div></div><div><h3>Conclusions</h3><div>This study compared the cochlear responses differences in normal hearing and hearing loss to forward and reverse stimulation, contributing to the optimization of the round window stimulating AMEIs.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108721"},"PeriodicalIF":4.9,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143679703","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":"Predicting strength of femora with metastatic lesions from single 2D radiographic projections using convolutional neural networks","authors":"Alexander Synek ,&nbsp;Emir Benca ,&nbsp;Roxane Licandro ,&nbsp;Lena Hirtler ,&nbsp;Dieter H. Pahr","doi":"10.1016/j.cmpb.2025.108724","DOIUrl":"10.1016/j.cmpb.2025.108724","url":null,"abstract":"<div><h3>Background and objective</h3><div>Patients with metastatic bone disease are at risk of pathological femoral fractures and may require prophylactic surgical fixation. Current clinical decision support tools often overestimate fracture risk, leading to overtreatment. While novel scores integrating femoral strength assessment via finite element (FE) models show promise, they require 3D imaging, extensive computation, and are difficult to automate. Predicting femoral strength directly from single 2D radiographic projections using convolutional neural networks (CNNs) could address these limitations, but this approach has not yet been explored for femora with metastatic lesions. This study aimed to test whether CNNs can accurately predict strength of femora with metastatic lesions from single 2D radiographic projections.</div></div><div><h3>Methods</h3><div>CNNs with various architectures were developed and trained using an FE model generated training dataset. This training dataset was based on 36,000 modified computed tomography (CT) scans, created by randomly inserting artificial lytic lesions into the CT scans of 36 intact anatomical femoral specimens. From each modified CT scan, an anterior-posterior 2D projection was generated and femoral strength in one-legged stance was determined using nonlinear FE models. Following training, the CNN performance was evaluated on an independent experimental test dataset consisting of 31 anatomical femoral specimens (16 intact, 15 with artificial lytic lesions). 2D projections of each specimen were created from corresponding CT scans and femoral strength was assessed in mechanical tests. The CNNs’ performance was evaluated using linear regression analysis and compared to 2D densitometric predictors (bone mineral density and content) and CT-based 3D FE models.</div></div><div><h3>Results</h3><div>All CNNs accurately predicted the experimentally measured strength in femora with and without metastatic lesions of the test dataset (<em>R</em>²≥0.80, <em>CCC</em>≥0.81). In femora with metastatic lesions, the performance of the CNNs (best: <em>R</em>²=0.84, <em>CCC</em>=0.86) was considerably superior to 2D densitometric predictors (<em>R</em>²≤0.07) and slightly inferior to 3D FE models (<em>R</em>²=0.90, <em>CCC</em>=0.94).</div></div><div><h3>Conclusions</h3><div>CNNs, trained on a large dataset generated via FE models, predicted experimentally measured strength of femora with artificial metastatic lesions with accuracy comparable to 3D FE models. By eliminating the need for 3D imaging and reducing computational demands, this novel approach demonstrates potential for application in a clinical setting.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"265 ","pages":"Article 108724"},"PeriodicalIF":4.9,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143740116","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":"Linguistic-grammar profile of Polish patients with anorexia nervosa","authors":"Stella Maćkowska ,&nbsp;Katarzyna Rojewska ,&nbsp;Dominik Spinczyk","doi":"10.1016/j.cmpb.2025.108717","DOIUrl":"10.1016/j.cmpb.2025.108717","url":null,"abstract":"<div><h3>Background and objective</h3><div>The process of diagnosing and treating anorexia is fraught with many challenges. Physiologically unstable patient status in the first period of treatment, the barrier between patient-therapist, and patient's resistance constitute an essential negative element in accurate diagnosis and appropriate therapy selection. For this reason, there was a need to create a tool using elements of natural language processing to support the psychologist's work in the diagnostic process to verify and validate the expert hypotheses.</div></div><div><h3>Methods</h3><div>The research proposed that linguistic-grammatical profiles be created among the research and control groups using elements of natural language processing. After the general part of speech tagging, the rules for detailed analysis were developed for adjectives, verbs (including the verb “to be”), pronoun “I” and the possessive pronoun “my”, cognitive words and characteristic terms related to body image. The choice of rules was dictated by the state of art and literature review. The obtained results were subjected to statistical analysis.</div></div><div><h3>Results</h3><div>A detailed analysis showed a strong negative sentiment associated with body image among patients with anorexia. In the control group, the same analysis revealed opposite results. In this group, people are aware of their physical imperfections, but it does not distort their body image. Statistically significant differences were observed in all concept categories except for the noun group. Statistical analysis was not conducted for the following concept classes: personal pronoun “I”, verb “to be” in the past form, verb “to be” in the future form, and general verbs in past form due to the insufficient number of occurrences of these concepts in the written notes.</div></div><div><h3>Conclusion</h3><div>The adopted NLP methods and the tools used in the designed projective method may be helpful in the psychological diagnosis of anorexia, due to the demonstrated differentiation between healthy and people with anorexia, providing detailed information about the patient and its required minimally invasive character.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108717"},"PeriodicalIF":4.9,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611159","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":"Advancing hierarchical neural networks with scale-aware pyramidal feature learning for medical image dense prediction","authors":"Xiang Liu ,&nbsp;James Liang ,&nbsp;Jianwei Zhang ,&nbsp;Zihan Qian ,&nbsp;Phoebe Xing ,&nbsp;Taige Chen ,&nbsp;Shanchieh Yang ,&nbsp;Chijioke Chukwudi ,&nbsp;Liang Qiu ,&nbsp;Dongfang Liu ,&nbsp;Junhan Zhao","doi":"10.1016/j.cmpb.2025.108705","DOIUrl":"10.1016/j.cmpb.2025.108705","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Hierarchical neural networks are pivotal in medical imaging for multi-scale representation, aiding in tasks such as object detection and segmentation. However, their effectiveness is often limited by the loss of intra-scale information and misalignment of inter-scale features. Our study introduces the Integrated-Scale Pyramidal Interactive Reconfiguration to Enhance feature learning (INSPIRE).</div></div><div><h3>Methods:</h3><div>INSPIRE focuses on intra-scale semantic enhancement and precise inter-scale spatial alignment, integrated with a novel spatial-semantic back augmentation technique. We evaluated INSPIRE’s efficacy using standard hierarchical neural networks, such as UNet and FPN, across multiple medical segmentation challenges including brain tumors and polyps. Additionally, we extended our evaluation to object detection and semantic segmentation in natural images to assess generalizability.</div></div><div><h3>Results:</h3><div>INSPIRE demonstrated superior performance over standard baselines in medical segmentation tasks, showing significant improvements in feature learning and alignment. In identifying brain tumors and polyps, INSPIRE achieved enhanced precision, sensitivity, and specificity compared to traditional models. Further testing in natural images confirmed the adaptability and robustness of our approach.</div></div><div><h3>Conclusions:</h3><div>INSPIRE effectively enriches semantic clarity and aligns multi-scale features, achieving integrated spatial-semantic coherence. This method seamlessly integrates with existing frameworks used in medical image analysis, thereby promising to significantly enhance the efficacy of computer-aided diagnostics and clinical interventions. Its application could lead to more accurate and efficient imaging processes, essential for improved patient outcomes.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"265 ","pages":"Article 108705"},"PeriodicalIF":4.9,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143759641","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":"Physiological modeling of autonomic regulation of cardiac system under graded exercise","authors":"Tao Wang ,&nbsp;JianKang Wu ,&nbsp;Fei Qin ,&nbsp;Hong Jiang ,&nbsp;Xiang Xiao ,&nbsp;YongGang Tong ,&nbsp;ChuChu Liao ,&nbsp;ZhiPei Huang","doi":"10.1016/j.cmpb.2025.108704","DOIUrl":"10.1016/j.cmpb.2025.108704","url":null,"abstract":"<div><h3>Background and Objective</h3><div>: Dysfunction of the autonomic nervous system (ANS) plays a critical role in the progression and assessment of cardiovascular diseases, neurological disorders, and various other pathologies. Therefore, a quantitative assessment of ANS function is vital for personalized medicine in these diseases. However, direct measurements of ANS activity can be costly and invasive, prompting researchers to adopt indirect methods for quantitative evaluation. These methods typically involve mathematical techniques, such as statistical analysis and mathematical modeling, to interpret cardiovascular fluctuations in response to external stimuli.The purpose of this study is to develop a non-invasive mathematical method that quantitatively assesses ANS function during graded exercise.</div></div><div><h3>Methods:</h3><div>In this study, we present a physiological mathematical model for autonomic regulation of the cardiac system under graded exercise, which recognizes the crucial role of the ANS in controlling heart rate during physical activity. The model utilizes the metabolic equivalent of walking as the input and heart rate as the output, with model parameters serving as quantitative measures of personalized ANS function. Experimental data were collected from groups with different health statuses and genders. Mann–Whitney U non-parametric tests were conducted on the model parameters to assess performance between individuals who frequently engage in aerobic exercise (15 participants, aerobic exercise frequency of more than 4 times/week) and those who barely exercise (15 participants, aerobic exercise frequency of 1 time per week or less), as well as between male and female participants.</div></div><div><h3>Results:</h3><div>The experimental results indicate that our model effectively quantitatively assesses ANS function across groups with different health statuses and genders (P <span><math><mo>&lt;</mo></math></span> 0.05). Additionally, the model provides precise estimations of heart rate, yielding a Root Mean Square Error of 2.79 beats per minute, a Mean Absolute Error of 2.18 beats per minute, and an R-squared value of 0.93.</div></div><div><h3>Conclusion:</h3><div>Our findings suggest that the proposed physiological mathematical model offers a non-invasive and user-friendly tool for measuring ANS function and monitoring cardiovascular health. This approach is feasible for home application, thereby reducing the need for professional supervision, and supports the early detection and personalized management of cardiovascular diseases. As a result, it enhances clinical decision-making and improves patient outcomes.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108704"},"PeriodicalIF":4.9,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143636342","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 study on brain tumor dynamics in two-dimensional irregular domain with variable-order time-fractional derivative","authors":"Harshad Sakariya,&nbsp;Ravi Shankar Prasad,&nbsp;Sushil Kumar","doi":"10.1016/j.cmpb.2025.108700","DOIUrl":"10.1016/j.cmpb.2025.108700","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Understanding tumor growth in the brain is a crucial and complex challenge. This study aims to develop and analyze a brain tumor growth model that incorporates variable-order time-fractional derivatives within a two-dimensional irregular domain. The purpose is to explore the effects of time-fractional orders, mutation rates, and growth parameters on tumor dynamics.</div></div><div><h3>Methods:</h3><div>The model employs the finite difference method for temporal discretization and Gaussian radial basis functions based on Kansa’s method for spatial variables. Ulam–Hyers stability analysis is performed to ensure the model’s stability and the existence and uniqueness of the solution are established. Additionally, the stability and convergence of the scheme are analyzed. Code verification is conducted to confirm the accuracy and reliability of the computational approach. Key parameters, such as the mutation rate <span><math><msub><mrow><mi>K</mi></mrow><mrow><mi>c</mi></mrow></msub></math></span> and growth parameters <span><math><msub><mrow><mi>σ</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span> and <span><math><msub><mrow><mi>σ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span>, are investigated under various time-fractional derivative orders, including variable orders.</div></div><div><h3>Results:</h3><div>The numerical simulations provide a detailed analysis of tumor cell dynamics, accounting for heterogeneity and fractional effects. Graphical representations reveal novel behaviors induced by variable-order time-fractional derivatives, including their impact on tumor cell population growth. Changes in the mutation rate and growth parameters significantly influence the results, demonstrating sensitivity to parameter variations.</div></div><div><h3>Conclusions:</h3><div>This study demonstrates that the integration of variable-order time-fractional derivatives into brain tumor models introduces memory effects, revealing new insights into tumor behavior. The findings highlight the importance of fractional-order parameters in accurately modeling brain tumor growth, which could have potential implications for predicting tumor progression and developing targeted treatments.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108700"},"PeriodicalIF":4.9,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143629115","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":"Optimization of primary screw stability in Trabecular bone using neural network-based models","authors":"Yijun Zhou ,&nbsp;Benedikt Helgason ,&nbsp;Stephen J. Ferguson ,&nbsp;Cecilia Persson","doi":"10.1016/j.cmpb.2025.108720","DOIUrl":"10.1016/j.cmpb.2025.108720","url":null,"abstract":"<div><h3>Background and Objective</h3><div>Screw implant stability in bone is crucial to the success of many orthopaedic procedures, yet the relationship between screw design parameters and specific bone characteristics remains underexplored. This study aims to optimize screw designs to enhance primary stability by leveraging subject-specific bone data and advanced surrogate modelling techniques.</div></div><div><h3>Methods</h3><div>In this study, 2880 screw pull-out simulations were conducted to assess primary screw stability by analysing pull-out stiffness and strength. The resulting dataset was used to develop surrogate models using multiple linear regression, random forest, and neural networks (NN). An optimization process was then applied to find optimal screw designs for 80 distinct trabecular bone specimens, in terms of inner diameter, pitch, and thread angle.</div></div><div><h3>Results</h3><div>The models, trained with various input parameters, including bone morphological parameters and computed tomography images, promisingly predicted the results of the simulations. The prediction errors varied by model type, with multiple linear regression yielding approximately 12 % error, while non-linear machine learning models achieved lower errors, ranging between 2–6 %. The series of subsequent optimization tasks provided optimized screw designs showing statistically significant improvements in pull-out stiffness and strength compared to the average screw designs (approximately 16 and 14 %, respectively). This even though our study focused only on screw design parameters that generally have a smaller impact on stability compared to factors such as screw outer diameter and insertion depth.</div></div><div><h3>Conclusions</h3><div>Multiple linear regression models were found to be insufficient for generating optimized screw configurations, and more complex surrogate models, such as NN, are needed. It could be concluded that different trabecular bone morphologies can benefit from distinct optimal screw designs. The insights gained from this study could have implications for the development of patient-specific orthopaedic treatments.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"264 ","pages":"Article 108720"},"PeriodicalIF":4.9,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143641154","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}