Biomedical Signal Processing and Control最新文献_第9页

Multi-Scale Hierarchical Context-Aware Survival Prediction Network based on whole slide images 基于全幻灯片图像的多尺度分层上下文感知生存预测网络

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-12 DOI: 10.1016/j.bspc.2025.108159

Jinmiao Song , Yatong Hao , Hui Zhai , Shuang Zhao , Tao Ning , Shan Huang , Xiaodong Duan

{"title":"Multi-Scale Hierarchical Context-Aware Survival Prediction Network based on whole slide images","authors":"Jinmiao Song , Yatong Hao , Hui Zhai , Shuang Zhao , Tao Ning , Shan Huang , Xiaodong Duan","doi":"10.1016/j.bspc.2025.108159","DOIUrl":"10.1016/j.bspc.2025.108159","url":null,"abstract":"<div><div>In high-resolution whole slide images (WSIs), multi-scale information is crucial for survival prediction. However, due to the ultra-large sizes of WSIs, existing methods have not fully utilized the multi-scale information at the gigapixel scale. Additionally, WSI-based survival prediction, as a patient-level multiple instance learning (MIL) task, is far more complex than WSI-level MIL which presents a significant challenge. To address these challenges, we propose a Multi-Scale Hierarchical Context-Aware Survival Prediction Network (MSASurv). This network progressively explores the tumor microenvironment, tumor-associated tissue structures, and patient-level tumor heterogeneity in WSIs. We validated our approach using five types of cancer from The Cancer Genome Atlas (TCGA), including 3,068 H&E-stained WSIs. Experimental results demonstrate that our proposed MSASurv algorithm outperforms previous weakly supervised methods by 3.3% to 16.8%. The code and models are publicly available at <span><span>https://github.com/yatonghao/MSASurv</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108159"},"PeriodicalIF":4.9,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144262309","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}

引用次数: 0

T-CADiff: Conditional guidance based diffusion model for medical image segmentation t - cadff：基于条件引导的医学图像分割扩散模型

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-12 DOI: 10.1016/j.bspc.2025.108139

Han Yang , Jiao Zhang , Lijiang Shao , Chao Jin , Jian Yang , Chen Qiao

{"title":"T-CADiff: Conditional guidance based diffusion model for medical image segmentation","authors":"Han Yang , Jiao Zhang , Lijiang Shao , Chao Jin , Jian Yang , Chen Qiao","doi":"10.1016/j.bspc.2025.108139","DOIUrl":"10.1016/j.bspc.2025.108139","url":null,"abstract":"<div><div>Diffusion models receive particular attention in the field of medical image segmentation due to their excellent quality and good image diversity. However, there are still some urgent issues remaining. Diffusion models used for image segmentation require the original image to guide generation, and the semantic features of the original images are often misaligned with the features of noisy segmentation masks, affecting the realism of the generated images. Meanwhile, most existing studies based on diffusion models still rely on the classic UNet structure, which does not fully utilize global information. To address the above issues, the model includes a condition guidance transformer (CGT) module to fuse features of the original images and noisy segmentation masks and learn global information. In addition to noise prediction, the real mask is also predicted and subsequently fed into a discriminator to increase the accuracy of predicted noise and the realism of the segmentation images. On the ISIC 2016 skin lesion segmentation dataset, our model achieved 11.04, 85.35%, 91.51%, 90.93%, and 95.63% on the HD95, IoU, Dice, sensitivity, and accuracy, respectively. It can be demonstrated that the proposed model significantly enhances segmentation accuracy, outperforming the state-of-the-art models.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108139"},"PeriodicalIF":4.9,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144262928","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}

引用次数: 0

Constrained optimal control drug scheduling models with different toxicity metabolism in cancer chemotherapy 肿瘤化疗中不同毒性代谢的约束最优控制药物调度模型

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-11 DOI: 10.1016/j.bspc.2025.108176

Emad Abdullah Musleh , Jeevan Kanesan , Joon Huang Chuah , Anand Ramanathan

{"title":"Constrained optimal control drug scheduling models with different toxicity metabolism in cancer chemotherapy","authors":"Emad Abdullah Musleh , Jeevan Kanesan , Joon Huang Chuah , Anand Ramanathan","doi":"10.1016/j.bspc.2025.108176","DOIUrl":"10.1016/j.bspc.2025.108176","url":null,"abstract":"<div><div>In cancer therapy optimization, various methods exist for modelling and managing the mathematical dynamics of cancer, each tailored to specific objectives. Among these, optimal control theory remains a powerful method for minimizing drug delivery in cancer therapy protocols, despite recent advancements in alternative methodologies. This method aligns with pharmacokinetic and pharmacodynamic principles, ensuring efficacy and allowing for the exploration of emerging techniques.</div><div>This study utilized bang-bang optimal control through discretization and nonlinear programming techniques facilitated by the Applied Modelling Programming Language (AMPL), linked with the Interior-Point optimization solver (IPOPT). This approach enabled the determination of extremal solutions that satisfied the system’s constraints. A key modification introduced was replacing the Heaviside function with a sigmoid function for smoother drug effect transitions while incorporating an additional equation for healthy cell dynamics. Interestingly, numerical results showed no significant difference between the Heaviside and sigmoid-based models, suggesting that the optimal control strategy remains unchanged and inherently favours bang-bang solutions.</div><div>The optimal control solutions demonstrated adaptability across various physiological conditions during cancer treatment, quantified through performance indices and residual cancer cell counts. Our results support the superiority of optimal control, achieving the highest performance index (31.1132) and the lowest residual cancer cell count (0.0307).</div><div>This study underscores the utility of optimal control in improving cancer treatment efficiency, reducing medication use, and lowering overall costs while enhancing the real-world applications of mathematical modelling in healthcare.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108176"},"PeriodicalIF":4.9,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144262308","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}

引用次数: 0

Adaptive Multi-Scale Convolution and Local Attention for single-channel EEG sleep staging 单通道脑电睡眠分期的自适应多尺度卷积与局部注意

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-10 DOI: 10.1016/j.bspc.2025.108084

Jingxin Fan , Mingfu Zhao , Lurui Wang , Bin Tang , Jingchuan Lu , Zhong He , Li Huang

{"title":"Adaptive Multi-Scale Convolution and Local Attention for single-channel EEG sleep staging","authors":"Jingxin Fan , Mingfu Zhao , Lurui Wang , Bin Tang , Jingchuan Lu , Zhong He , Li Huang","doi":"10.1016/j.bspc.2025.108084","DOIUrl":"10.1016/j.bspc.2025.108084","url":null,"abstract":"<div><div>Sleep staging is a critical component of sleep medicine, providing indispensable insights into sleep disorders and broader health outcomes. Single-channel EEG sleep staging offers a less invasive and more practical alternative to multi-channel methods. However, achieving high classification accuracy remains challenging, given the complex, non-stationary nature of EEG signals. In this study, we propose a new model that combines Adaptive Multi-Scale Convolution (AMSC) and Local Attention mechanisms. The proposed AMSC module captures temporal features across multiple scales, enabling the model to adapt to signal dynamics and focus on relevant frequency bands. This adaptability improves handling of the heterogeneity inherent in EEG data. The Local Attention module refines feature extraction by concentrating on the most informative segments of the signal, thereby improving the model’s capability to distinguish among various sleep stages. Our approach integrates normalization layers, a Multi-Layer Perceptron (MLP), and a Softmax classifier to ensure robust learning and accurate stage classification. Experimental validation on publicly available EEG datasets and self-collected datasets shows that the proposed model surpasses current state-of-the-art methods. Specifically, our model achieves an overall accuracy of 86.2% on the Sleep-EDF dataset, with a notable 59.1% accuracy for the challenging N1 stage. On our self-collected CQSH-OSSD dataset, the model achieves an overall accuracy of 79.4%, demonstrating strong generalization performance across different data sources. The proposed framework confirms the potential of combining adaptive convolution architectures with attention-based feature enhancement. It improves the effectiveness of sleep staging based on single-channel EEG.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108084"},"PeriodicalIF":4.9,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144239592","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}

引用次数: 0

Incremental modeling and its application for driver fatigue estimation 增量建模及其在驾驶员疲劳估计中的应用

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-09 DOI: 10.1016/j.bspc.2025.108046

Szilárd Kovács, János Botzheim

{"title":"Incremental modeling and its application for driver fatigue estimation","authors":"Szilárd Kovács, János Botzheim","doi":"10.1016/j.bspc.2025.108046","DOIUrl":"10.1016/j.bspc.2025.108046","url":null,"abstract":"<div><div>This article focuses on an Incremental Learning (IL)-based approach for examining a driver fatigue dataset and compares its performance with classical Machine Learning (ML) and deep learning techniques with respect to generalization capability. Driver fatigue is of significant concern in traditional driving conditions, contributing to many serious and fatal accidents worldwide. While self-driving cars may eventually alleviate this issue, until they are fully developed, “self-driving carsickness” introduces new problems. Various classification methods have been proposed in recent years to distinguish between drowsy and alert states, including both binary and multi-class classifications. However, the generalization capabilities of these methods are underexplored. IL, often seen as a data-hungry approach, optimizes both model parameters and structure. The primary goal is early recognition of fatigue, enabling timely intervention, and ensuring explainability for safety-critical systems. To address these issues, we propose a novel regression approach. We propose a deterministic regression model, guided by binary labels and using an Incremental Modeling (IM) framework, to address the challenges in fatigue recognition. Our model demonstrates superior performance on a challenging dataset, with improvements in accuracy (<span><math><mrow><mo>+</mo><mn>2</mn><mo>…</mo><mo>+</mo><mn>7</mn><mtext>%</mtext></mrow></math></span>), parameter count (<span><math><mrow><mo>−</mo><mn>99</mn><mo>…</mo><mo>−</mo><mn>100</mn><mtext>%</mtext></mrow></math></span>), speed (<span><math><mrow><mo>+</mo><mn>0</mn><mo>.</mo><mn>5</mn><mo>…</mo><mo>−</mo><mn>0</mn><mo>.</mo><mn>1</mn></mrow></math></span> s), and latency (<span><math><mrow><mo>−</mo><mn>70</mn><mi>μ</mi><mi>s</mi></mrow></math></span>). It also offers flexibility with optional personalization, illustrating the strength and adaptability of IM for fatigue detection in various conditions.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108046"},"PeriodicalIF":4.9,"publicationDate":"2025-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243367","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}

引用次数: 0

Optimized dynamic global structure enhanced multi-channel graph neural network based automatic cataract disease classification 基于优化动态全局结构增强多通道图神经网络的白内障疾病自动分类

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-09 DOI: 10.1016/j.bspc.2025.108125

Saritha Balu , Mrinal R. Bachute , K. Venkatraman , John Augustine Parvathinathan

{"title":"Optimized dynamic global structure enhanced multi-channel graph neural network based automatic cataract disease classification","authors":"Saritha Balu , Mrinal R. Bachute , K. Venkatraman , John Augustine Parvathinathan","doi":"10.1016/j.bspc.2025.108125","DOIUrl":"10.1016/j.bspc.2025.108125","url":null,"abstract":"<div><div>The most prevalent condition that causes vision distortion is cataracts. The greatest method to reduce the danger and prevent blindness is to detect cataracts accurately and timely detection. Research interest in the cataract detection systems based on artificial intelligence is recently increased. In this manuscript, Optimized Dynamic Global Structure Enhanced Multi-channel Graph Neural Network depend Automatic Cataract Disease Classification (DSEMGNN-CACD-SETOA) is proposed. The input fundus images are obtained using glaucoma dataset and the image is given to pre-processing. The input images are pre-preprocessing utilizing Generalized Multi-kernel Maximum Correntropy Kalman Filter (GMMCKF) to resize and normalize the image. The pre-processed imagery is provided to the categorization. Finally, the pre-processed imageries are provided to Dynamic Global Structure Enhanced multi-channel Graph Neural Network (DSEMGNN) to classify cataract disease as Referable Glaucoma and Non-referable Glaucoma. The Stock Enhancing Trading Optimization Algorithm (SETOA) is proposed for improving the weight parameter of DSEMGNN for cataract disease classification. The proposed KOAC-GCIGNN-AcME-SBOA technique is implemented on Python. The proposed approach attains 28%, 30.78% and 25.29% higher accuracy, 15.08%, 20.58%, and 15.25% higher precision when comparing with the existing methods like GLA-Net: A global–local attention network for automated cataract categorization (GLAN-CNN-ACC), cataract grading technique depending on deep convolutional neural networks and stacking ensemble learning (CGM-DCNN-SEL),automated cataract detection scheme with deep learning for fundus imageries (ACD-DNN-FI)respectively.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108125"},"PeriodicalIF":4.9,"publicationDate":"2025-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144239588","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}

引用次数: 0

Optimized attention Induced multi head convolutional neural network with Densenet201 for cervical cancer diagnosis 应用Densenet201优化注意诱导多头卷积神经网络用于宫颈癌诊断

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-09 DOI: 10.1016/j.bspc.2025.108166

T.S. Sheela Shiney , S. Albert Jerome

{"title":"Optimized attention Induced multi head convolutional neural network with Densenet201 for cervical cancer diagnosis","authors":"T.S. Sheela Shiney , S. Albert Jerome","doi":"10.1016/j.bspc.2025.108166","DOIUrl":"10.1016/j.bspc.2025.108166","url":null,"abstract":"<div><div>Cervical cancer is the fourth most common disease globally, highlighting the importance of early detection for effective treatment. Although the Pap smear test is the gold standard for detecting cervical cancer, its effectiveness depends on the expertise and dedication of the physicians. In this paper, Optimized Attention Induced Multi Head Convolutional Neural Network with Densenet201 for Cervical Cancer Diagnosis (AIMHCNN-Densenet201-CCD) is proposed. Initially, the input images are gathered from SIPaKMeD and Medical Scan Classification Dataset. Then, the input data is pre-processed using Regularized Bias-aware Ensemble Kalman filter (RBEKF) to crop and rotate input images. The pre-processed images are fed to Modified Spline-Kernelled Chirplet Transform (MSKCT) to extract the morphological features such as Shape, Colour, Structure and Size. Afterwards, the extracted features are fed into Multi-Head Convolutional Neural Network with Attention Induced and Densenet201 (AIMHCNN-Densenet201) for diagnosing Cervical Cancer like Dyskeratotic, Metaplastic, Koilocytotic, Parabasal and Superficial-Intermediate. Finally, Dove Swarm Optimization (DSO) is proposed to optimize the weight parameter of AIMHCNN-Densenet201 classifier that precisely diagnoses the Cervical Cancer. The proposed AIMHCNN-Densenet201-CCD method is implemented and analyzed using performance metrics such as accuracy, precision, specificity, f1-score, sensitivity, error rate and computation time. The proposed approach attains 29.82 %, 21.24 %, 18.97 % higher accuracy and 24.75 %, 32.57 %, and 29.69 % higher precision compared with existing methods respectively.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108166"},"PeriodicalIF":4.9,"publicationDate":"2025-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144239589","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}

引用次数: 0

Multi-scale information residual network: Deep residual network of prostate cancer segmentation based on multi scale information guidance 多尺度信息残差网络：基于多尺度信息引导的前列腺癌分割深度残差网络

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-09 DOI: 10.1016/j.bspc.2025.108132

Xinyi Chen , Xiang Liu , Yunjie Yu , Yunyu Shi , Yuke Wu , Zhenglei Wang , Shuohong Wang

{"title":"Multi-scale information residual network: Deep residual network of prostate cancer segmentation based on multi scale information guidance","authors":"Xinyi Chen , Xiang Liu , Yunjie Yu , Yunyu Shi , Yuke Wu , Zhenglei Wang , Shuohong Wang","doi":"10.1016/j.bspc.2025.108132","DOIUrl":"10.1016/j.bspc.2025.108132","url":null,"abstract":"<div><div>Magnetic resonance imaging (MRI) is a key tool in prostate cancer screening and diagnosis, with automatic segmentation of the cancer crucial for accurate staging and treatment. Nevertheless, the accurate segmentation of prostate cancer remains a challenging subject. In order to address this challenge, a two-stage segmentation method is employed. In the initial stage, the prostate tissue is delineated as the region of interest. Subsequently, in the second stage, the precise segmentation of prostate cancer is achieved on the aforementioned region of interest. In order to accurately segment the region of interest, we propose MSR-Net (Multi-scale information residual network), which employs an MSR-block, designed based on MLKA convolution, to extract multi-scale information, combines the group attention mechanism to enhance the multi-scale features, and uses the improved CGA feature fusion module to fuse deep and shallow features. The feature map is then upsampled using DySample. The experiments conducted on the Prostatex dataset for the segmentation of prostate cancer were based on the Dice metric. The results demonstrate an improvement of 5.2% (60.5% vs. 55.3%) in comparison to the second-best performing segmentation network (Unet). The application of the two-stage segmentation method has a considerable impact, with a 10.4% improvement (45.3% vs 55.7%) on the baseline when two-stage segmentation is employed for prostate cancer in comparison to direct segmentation of prostate cancer. Furthermore, the network was trained and tested on the prostate segmentation and lung nodule segmentation datasets, achieving the highest dice scores of 0.937 and 0.764, respectively.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108132"},"PeriodicalIF":4.9,"publicationDate":"2025-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243368","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}

引用次数: 0

SIT-SAM: A semantic-integration transformer that adapts the Segment Anything Model to zero-shot medical image semantic segmentation SIT-SAM：一种将任意片段模型应用于零镜头医学图像语义分割的语义集成转换器

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-08 DOI: 10.1016/j.bspc.2025.108086

Wentao Shi , Junjun He , Yiqing Shen

{"title":"SIT-SAM: A semantic-integration transformer that adapts the Segment Anything Model to zero-shot medical image semantic segmentation","authors":"Wentao Shi , Junjun He , Yiqing Shen","doi":"10.1016/j.bspc.2025.108086","DOIUrl":"10.1016/j.bspc.2025.108086","url":null,"abstract":"<div><div>Segment Anything Model (SAM) demonstrates zero-shot instance segmentation capabilities through prompt-guided interaction. However, its application to 3D medical imaging remains limited due to insufficient semantic understanding of complex anatomical structures. Current SAM variants attempt to address this challenge through architectural modifications and fine-tuning ; however, these approaches often compromise SAM’s original zero-shot capabilities. To bridge this gap, we introduce the semantic integration Transformer for SAM (SIT-SAM), an innovative post-processing framework that enhances SAM’s instance-level masks with semantic comprehension of anatomical structures. Our approach preserves SAM’s valuable zero-shot capabilities while introducing semantic awareness. Specifically, SIT-SAM comprises of three functional blocks: (1) the original SAM for instance mask generation, (2) a semantic integration transformer that combines hierarchical multi-scale feature extraction to capture both fine anatomical details and global context while leveraging instance mask geometry for enhanced anatomical structure understanding, (3) a cognitive science-inspired memory module for learning from limited training data. Evaluation on the TotalSegmentator dataset demonstrates SIT-SAM’s superior performance, achieving 90.55% accuracy, substantially outperforming the fine-tuned baseline <em>i</em>.<em>e</em>. SAM-Med3D with fully convolutional network (FCN) prediction head by 52.69%. SIT-SAM also exhibits robustness in data-constrained environments, delivering a 2.43% improvement with single-point prompt and maintaining effectiveness with multiple prompts, showing a 0.78% gain using ten point prompts. Code is available at <span><span>https://github.com/wentao0429/SIT-SAM</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108086"},"PeriodicalIF":4.9,"publicationDate":"2025-06-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144239591","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}

引用次数: 0

FOC-Net: A lightweight network combining full 1 × 1 convolutions with wavelet and attention mechanisms for lung nodule segmentation FOC-Net：将全1 × 1卷积与小波和注意机制相结合的轻量级网络，用于肺结节分割

IF 4.9 2区医学

Biomedical Signal Processing and Control Pub Date : 2025-06-08 DOI: 10.1016/j.bspc.2025.108088

Xingao Li, Hongmin Deng, Xuan Tang

{"title":"FOC-Net: A lightweight network combining full 1 × 1 convolutions with wavelet and attention mechanisms for lung nodule segmentation","authors":"Xingao Li, Hongmin Deng, Xuan Tang","doi":"10.1016/j.bspc.2025.108088","DOIUrl":"10.1016/j.bspc.2025.108088","url":null,"abstract":"<div><div>Accurate segmentation and analysis of lung nodules are essential for the formulation of effective treatment strategies. However, existing high-performance segmentation algorithms generally depend on substantial computational resources, presenting significant challenges for resource-constrained medical devices, especially edge devices. This study proposes a lightweight and efficient segmentation network called FOC-Net, which replaces large convolutional kernels by using 1 <span><math><mo>×</mo></math></span> 1 convolutions combined with spatial shifting operations. FOC-Net builds a U-shaped encoder–decoder architecture based on the shift convolution residual block (SCR-Block) and three other key modules: the wavelet-based downsampling (WBD) module, which preserves detailed information and suppresses noise, thereby reducing information loss during the downsampling process; the channel-prior spatial attention (CPSA) module, which makes the model focus on lung nodule regions; and the weight-aware feature fusion (WAFF) module, which augments the model’s ability to capture contextual information. Experiments conducted on the LIDC-IDRI dataset demonstrate that the proposed model outperforms other state-of-the-art methods in lung nodule segmentation tasks, achieving a Dice similarity coefficient (DSC) of 92.06% and a Jaccard index (JI) of 85.37%, while maintaining a parameter count of only 0.64 million with GFLOPs of 3.15. Further experiments on the ISIC-2018 skin disease dataset validate the model’s generalization capability, with similar results: a DSC of 89.36% and a JI of 80.77%, still outperforming other state-of-the-art methods.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"110 ","pages":"Article 108088"},"PeriodicalIF":4.9,"publicationDate":"2025-06-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243629","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}

引用次数: 0