Biomedical Signal Processing and Control最新文献_第6页

RePrediction of chemical characteristics of individuals with HIV infection using an image processing-centric optimized deep learning-based CNN model 使用以图像处理为中心优化的基于深度学习的CNN模型重新预测HIV感染者的化学特征

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-13 DOI: 10.1016/j.bspc.2026.109816

Yusuf Alaca , Nursel Karaoğlan , Erdal Başaran , Yüksel Çelik

Accurate prediction of HIV-related chemical properties is of critical importance for computational drug discovery and bioinformatics applications. In this study, an image processing–centric deep learning framework is proposed to predict HIV chemical activity using molecular images automatically generated from SMILES representations. Multi-scale deep features are extracted using two complementary convolutional neural network architectures, namely Xception and ResNet50, and subsequently fused to capture both low-level structural patterns and high-level molecular representations. The main novelty of this work lies in the integration of CNN-based molecular image feature extraction with the Manta Ray Foraging Optimization (MRFO) algorithm. The MRFO algorithm is employed to perform optimization-driven feature selection and classifier hyperparameter tuning, aiming to improve both predictive accuracy and generalization capability. The optimized feature set is finally classified using a support vector machine (SVM), enabling robust discrimination between active and inactive HIV-related compounds. Experimental evaluations conducted on the benchmark HIV SMILES dataset demonstrate that the proposed framework achieves superior and stable performance, reaching an accuracy of 81.16% and a ROC-AUC of 0.87, outperforming several state-of-the-art machine learning and deep learning approaches reported in the literature. These results confirm that combining molecular image representations with optimization-guided deep learning provides an effective and reliable strategy for HIV chemical property prediction.

准确预测hiv相关的化学性质对于计算药物发现和生物信息学应用至关重要。在这项研究中，提出了一个以图像处理为中心的深度学习框架，利用smile表示自动生成的分子图像来预测HIV的化学活性。使用两个互补的卷积神经网络架构（即Xception和ResNet50）提取多尺度深度特征，并随后融合以捕获低级结构模式和高级分子表示。本工作的主要新颖之处在于将基于cnn的分子图像特征提取与蝠鲼觅食优化（MRFO）算法相结合。采用MRFO算法进行优化驱动的特征选择和分类器超参数调优，以提高预测精度和泛化能力。最后使用支持向量机（SVM）对优化后的特征集进行分类，从而实现对活性和非活性hiv相关化合物的鲁棒区分。在基准HIV SMILES数据集上进行的实验评估表明，所提出的框架具有优越且稳定的性能，准确率达到81.16%，ROC-AUC为0.87，优于文献中报道的几种最先进的机器学习和深度学习方法。这些结果证实，将分子图像表示与优化引导的深度学习相结合，为HIV化学性质预测提供了一种有效可靠的策略。

{"title":"RePrediction of chemical characteristics of individuals with HIV infection using an image processing-centric optimized deep learning-based CNN model","authors":"Yusuf Alaca , Nursel Karaoğlan , Erdal Başaran , Yüksel Çelik","doi":"10.1016/j.bspc.2026.109816","DOIUrl":"10.1016/j.bspc.2026.109816","url":null,"abstract":"<div><div>Accurate prediction of HIV-related chemical properties is of critical importance for computational drug discovery and bioinformatics applications. In this study, an image processing–centric deep learning framework is proposed to predict HIV chemical activity using molecular images automatically generated from SMILES representations. Multi-scale deep features are extracted using two complementary convolutional neural network architectures, namely Xception and ResNet50, and subsequently fused to capture both low-level structural patterns and high-level molecular representations. The main novelty of this work lies in the integration of CNN-based molecular image feature extraction with the Manta Ray Foraging Optimization (MRFO) algorithm. The MRFO algorithm is employed to perform optimization-driven feature selection and classifier hyperparameter tuning, aiming to improve both predictive accuracy and generalization capability. The optimized feature set is finally classified using a support vector machine (SVM), enabling robust discrimination between active and inactive HIV-related compounds. Experimental evaluations conducted on the benchmark HIV SMILES dataset demonstrate that the proposed framework achieves superior and stable performance, reaching an accuracy of 81.16% and a ROC-AUC of 0.87, outperforming several state-of-the-art machine learning and deep learning approaches reported in the literature. These results confirm that combining molecular image representations with optimization-guided deep learning provides an effective and reliable strategy for HIV chemical property prediction.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109816"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192616","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

PulseAI: An automated machine learning-based augmentation index detector for arterial stiffness monitoring from cuff-based measurements PulseAI：一种基于机器学习的自动增强指数检测器，用于通过袖带测量来监测动脉刚度

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-14 DOI: 10.1016/j.bspc.2026.109840

Alessio Tamborini , Arian Aghilinejad , Morteza Gharib

Arterial stiffness is a fundamental characteristic of circulatory physiology and a well-established predictor of cardiovascular risk and mortality. However, routine clinical assessment remains limited by the need for dual-site measurements. To address this challenge, we developed a machine learning algorithm – PulseAI – for automated fiducial point detection on brachial cuff waveforms for single-site assessment of arterial stiffness. PulseAI was trained and evaluated using a clinical dataset comprising 5,215 waveforms from 145 heterogeneous subjects. Performance was assessed on fiducial point predictions accuracy (inflection point, t_i, and dicrotic notch, t_n) and downstream pulse waveform analysis (PWA) metrics. Our multi-channel convolutional neural network (PulseAI) reported a median [IQR] on mean absolute error for fiducial point detection of 5 [3, 10] ms. PulseAI demonstrated high accuracy in predicting t_i (r = 0.913, p < 0.0001) and t_n (r = 0.939, p < 0.0001), with an average prediction error of 12.6 ms and 6.2 ms for t_i and t_n, respectively. While the t_n results are comparable to other academic models reporting ∼10 ms errors, our approach provides both fiducial point indices from a single model. PWA features derived from PulseAI closely matched those derived from human-annotated labels, including systolic pressure–time integral (r = 0.988, p < 0.0001), augmentation index (AIx) (r = 0.990, p < 0.0001), and end systolic pressure (r = 0.998, p < 0.0001). AIx tertiles showed statistically significant association with height-adjusted pulse transit time (p < 0.05), used as a surrogate of arterial stiffness, demonstrating the model’s sensitivity to stiffness-related changes. These findings demonstrate that PulseAI enables accurate fiducial point detection and represents a clinically viable tool for automated, single-site monitoring of arterial stiffness.

动脉僵硬是循环生理学的一个基本特征，也是心血管风险和死亡率的一个公认的预测指标。然而，常规的临床评估仍然受到双部位测量需求的限制。为了解决这一挑战，我们开发了一种机器学习算法PulseAI，用于对肱袖带波形进行自动基准点检测，用于单点动脉刚度评估。PulseAI使用临床数据集进行训练和评估，该数据集包括来自145名异质受试者的5,215个波形。性能评估的基准预测精度（拐点，ti和二向散陷，tn）和下游脉冲波形分析（PWA）指标。我们的多通道卷积神经网络（PulseAI）的基准点检测的平均绝对误差中位数[IQR]为5 [3,10]ms. PulseAI在预测ti （r = 0.913, p < 0.0001）和tn （r = 0.939, p < 0.0001）方面显示出较高的准确性，ti和tn的平均预测误差分别为12.6 ms和6.2 ms。虽然tn的结果与其他报告约10 ms误差的学术模型相当，但我们的方法提供了来自单个模型的两个基点指数。PulseAI获得的PWA特征与人工标注标签的PWA特征非常吻合，包括收缩压-时间积分（r = 0.988, p < 0.0001）、增强指数（AIx）（r = 0.990, p < 0.0001）和收缩压结束（r = 0.998, p < 0.0001）。AIx tertiles与高度调整后的脉冲传递时间（p < 0.05）有统计学意义（p < 0.05），该时间被用作动脉刚度的替代指标，表明该模型对刚度相关变化的敏感性。这些发现表明，PulseAI能够实现精确的基准点检测，是临床上可行的自动化、单点动脉僵硬监测工具。

{"title":"PulseAI: An automated machine learning-based augmentation index detector for arterial stiffness monitoring from cuff-based measurements","authors":"Alessio Tamborini , Arian Aghilinejad , Morteza Gharib","doi":"10.1016/j.bspc.2026.109840","DOIUrl":"10.1016/j.bspc.2026.109840","url":null,"abstract":"<div><div>Arterial stiffness is a fundamental characteristic of circulatory physiology and a well-established predictor of cardiovascular risk and mortality. However, routine clinical assessment remains limited by the need for dual-site measurements. To address this challenge, we developed a machine learning algorithm – PulseAI – for automated fiducial point detection on brachial cuff waveforms for single-site assessment of arterial stiffness. PulseAI was trained and evaluated using a clinical dataset comprising 5,215 waveforms from 145 heterogeneous subjects. Performance was assessed on fiducial point predictions accuracy (inflection point, t<sub>i</sub>, and dicrotic notch, t<sub>n</sub>) and downstream pulse waveform analysis (PWA) metrics. Our multi-channel convolutional neural network (PulseAI) reported a median [IQR] on mean absolute error for fiducial point detection of 5 [3, 10] ms. PulseAI demonstrated high accuracy in predicting t<sub>i</sub> (r = 0.913, p < 0.0001) and t<sub>n</sub> (r = 0.939, p < 0.0001), with an average prediction error of 12.6 ms and 6.2 ms for t<sub>i</sub> and t<sub>n</sub>, respectively. While the t<sub>n</sub> results are comparable to other academic models reporting ∼10 ms errors, our approach provides both fiducial point indices from a single model. PWA features derived from PulseAI closely matched those derived from human-annotated labels, including systolic pressure–time integral (r = 0.988, p < 0.0001), augmentation index (AIx) (r = 0.990, p < 0.0001), and end systolic pressure (r = 0.998, p < 0.0001). AIx tertiles showed statistically significant association with height-adjusted pulse transit time (p < 0.05), used as a surrogate of arterial stiffness, demonstrating the model’s sensitivity to stiffness-related changes. These findings demonstrate that PulseAI enables accurate fiducial point detection and represents a clinically viable tool for automated, single-site monitoring of arterial stiffness.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109840"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192591","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

Population information exchange differential evolution with CMAES for multi-threshold segmentation of breast cancer pathologic images 群体信息交换差分进化与CMAES在乳腺癌病理图像多阈值分割中的应用

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-09 DOI: 10.1016/j.bspc.2026.109796

Zhi Liu, Weibin Chen, Huiling Chen

The critical role of accurate segmentation in breast cancer diagnosis motivates this research. We propose a novel approach that integrates an advanced optimization algorithm to significantly improve the segmentation precision of pathological images. Specifically, we first improve the Differential Evolution (DE) algorithm by incorporating the Population Information Exchange (PIE) strategy and the Covariance Matrix Adaptation Evolution Strategy (CMAES) to propose TDE, The PIE mechanism improves the global search ability of the algorithm through random population information exchange, effectively suppressing premature convergence. The CMAES strategy generates high-quality solutions based on covariance learning, significantly improving the accuracy of local exploitation. Subsequently, TDE is coupled with Rényi entropy, forming the core of our segmentation methodology for breast histopathology images. We initially verified the optimization capability of TDE on the IEEE CEC 2017 functions set. Statistical validations through the Wilcoxon signed-rank test and Friedman test confirmed the superiority of TDE. Following this, we evaluated the multi-threshold segmentation model combining TDE and Rényi entropy on 9 breast cancer pathology images, demonstrating that the TDE algorithm outperforms peer algorithms. The comprehensive experimental results affirm TDE’s outstanding performance in both optimization benchmarks and medical image segmentation applications.

准确分割在乳腺癌诊断中的关键作用激发了这项研究。我们提出了一种集成先进优化算法的新方法，以显着提高病理图像的分割精度。具体而言，我们首先将种群信息交换（PIE）策略与协方差矩阵适应进化策略（CMAES）相结合，对差分进化（DE）算法进行改进，提出了差分进化（DE）算法，PIE机制通过随机种群信息交换提高了算法的全局搜索能力，有效抑制了过早收敛。CMAES策略基于协方差学习生成高质量的解，显著提高了局部开发的准确性。随后，将TDE与rsamnyi熵相结合，形成乳腺组织病理学图像分割方法的核心。我们在IEEE CEC 2017函数集上初步验证了TDE的优化能力。通过Wilcoxon sign -rank检验和Friedman检验的统计验证证实了TDE的优越性。在此基础上，我们对9张乳腺癌病理图像进行了TDE和rsamnyi熵相结合的多阈值分割模型的评估，结果表明TDE算法优于同类算法。综合实验结果证实了TDE在优化基准测试和医学图像分割应用中的出色表现。

{"title":"Population information exchange differential evolution with CMAES for multi-threshold segmentation of breast cancer pathologic images","authors":"Zhi Liu, Weibin Chen, Huiling Chen","doi":"10.1016/j.bspc.2026.109796","DOIUrl":"10.1016/j.bspc.2026.109796","url":null,"abstract":"<div><div>The critical role of accurate segmentation in breast cancer diagnosis motivates this research. We propose a novel approach that integrates an advanced optimization algorithm to significantly improve the segmentation precision of pathological images. Specifically, we first improve the Differential Evolution (DE) algorithm by incorporating the Population Information Exchange (PIE) strategy and the Covariance Matrix Adaptation Evolution Strategy (CMAES) to propose TDE, The PIE mechanism improves the global search ability of the algorithm through random population information exchange, effectively suppressing premature convergence. The CMAES strategy generates high-quality solutions based on covariance learning, significantly improving the accuracy of local exploitation. Subsequently, TDE is coupled with Rényi entropy, forming the core of our segmentation methodology for breast histopathology images. We initially verified the optimization capability of TDE on the IEEE CEC 2017 functions set. Statistical validations through the Wilcoxon signed-rank test and Friedman test confirmed the superiority of TDE. Following this, we evaluated the multi-threshold segmentation model combining TDE and Rényi entropy on 9 breast cancer pathology images, demonstrating that the TDE algorithm outperforms peer algorithms. The comprehensive experimental results affirm TDE’s outstanding performance in both optimization benchmarks and medical image segmentation applications.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109796"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191909","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

The application of convolutional neural networks for brain age prediction: A systematic review 卷积神经网络在脑年龄预测中的应用综述

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-14 DOI: 10.1016/j.bspc.2026.109812

Yutong Wu, Lan Lin, Chen Zhang

Deep learning (DL) has revolutionized neuroimaging, particularly with convolutional neural networks (CNNs) leading the charge, delivering unprecedented accuracy in brain age prediction. This systematic review meticulously investigates the strides made in CNNs for this critical task, pivotal to understanding variations in brain development and aging. Our exhaustive literature search spanning 2018 to 2024 using PRISMA principles and protocol identified 71 studies meeting stringent criteria, forming the basis of our analysis. Our examination delves deep into the neuroimaging data powering brain age prediction via CNNs, highlighting diverse model architectures and their real-world applications. We provide an intricate analysis of these architectures and methodologies, highlighting the shift towards 3D CNNs and the integration of multimodal neuroimaging data. Additionally, the review stresses the significance of model generalization and the hurdles posed by dataset limitations. Moreover, we underscore the imperative for models capable of generalizing effectively across varied demographics, underscoring the potential of CNNs to propel personalized medicine forward. This review encapsulates not only the current landscape but also points towards future directions, advocating for continued innovation to unlock the full potential of CNNs in deciphering the complexities of brain aging and beyond.

深度学习（DL）已经彻底改变了神经成像，特别是卷积神经网络（cnn）引领了这一潮流，在大脑年龄预测方面提供了前所未有的准确性。这篇系统综述细致地调查了cnn在这一关键任务上取得的进展，这对理解大脑发育和衰老的变化至关重要。我们使用PRISMA原则和协议对2018年至2024年的文献进行了详尽的检索，确定了71项符合严格标准的研究，构成了我们分析的基础。我们的研究深入研究了通过cnn进行脑年龄预测的神经成像数据，突出了不同的模型架构及其在现实世界中的应用。我们对这些架构和方法进行了复杂的分析，强调了向3D cnn的转变和多模态神经成像数据的集成。此外，本文还强调了模型泛化的重要性以及数据集限制带来的障碍。此外，我们强调了能够有效概括不同人口统计数据的模型的必要性，强调了cnn推动个性化医疗向前发展的潜力。这篇综述不仅概括了当前的情况，也指出了未来的方向，主张继续创新，以释放cnn在破译大脑衰老等复杂性方面的全部潜力。

{"title":"The application of convolutional neural networks for brain age prediction: A systematic review","authors":"Yutong Wu, Lan Lin, Chen Zhang","doi":"10.1016/j.bspc.2026.109812","DOIUrl":"10.1016/j.bspc.2026.109812","url":null,"abstract":"<div><div>Deep learning (DL) has revolutionized neuroimaging, particularly with convolutional neural networks (CNNs) leading the charge, delivering unprecedented accuracy in brain age prediction. This systematic review meticulously investigates the strides made in CNNs for this critical task, pivotal to understanding variations in brain development and aging. Our exhaustive literature search spanning 2018 to 2024 using PRISMA principles and protocol identified 71 studies meeting stringent criteria, forming the basis of our analysis. Our examination delves deep into the neuroimaging data powering brain age prediction via CNNs, highlighting diverse model architectures and their real-world applications. We provide an intricate analysis of these architectures and methodologies, highlighting the shift towards 3D CNNs and the integration of multimodal neuroimaging data. Additionally, the review stresses the significance of model generalization and the hurdles posed by dataset limitations. Moreover, we underscore the imperative for models capable of generalizing effectively across varied demographics, underscoring the potential of CNNs to propel personalized medicine forward. This review encapsulates not only the current landscape but also points towards future directions, advocating for continued innovation to unlock the full potential of CNNs in deciphering the complexities of brain aging and beyond.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109812"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146193105","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

Ensemble-MIL: deep learning-based ensemble framework for biomarker prediction from histopathological images in colorectal cancer 集成- mil：基于深度学习的集成框架，用于从结直肠癌的组织病理学图像中预测生物标志物

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-13 DOI: 10.1016/j.bspc.2026.109759

Geng-Yun Tien , Yu-Chia Chen , Liang-Chuan Lai , Tzu-Pin Lu , Mong-Hsun Tsai , Eric Y. Chuang , Hsiang-Han Chen

Recent studies have explored histopathological whole slide images (WSIs) for predicting colorectal cancer (CRC) biomarkers, aiming to create cost-effective and efficient diagnostic tools. However, achieving strong predictive performance and generalizability across datasets remains a challenge. Here, we introduce a deep learning-based ensemble framework, Ensemble-MIL, designed to robustly predict key CRC biomarkers, including BRAF V600E, KRAS mutations, and MSI-H status, with improved cross-dataset performance. We employed two independent CRC datasets: TCGA-COAD for model training and internal evaluation, and CPTAC-COAD as an external test set to assess generalizability. All WSIs were preprocessed and divided into small image patches. A tumor detection model was applied to identify tumor regions, and patch-level; features were extracted via SimCLR, a contrastive learning method. These features were utilized to train three multiple instance learning (MIL) models: Att-MIL, Tran-MIL, and GNN-MIL. The models were then integrated into the final Ensemble-MIL framework. In internal testing with TCGA-COAD, the proposed method achieved area under the curve (AUC) scores of 0.90, 0.87, and 0.64 for MSI-H, BRAF, and KRAS, respectively. In external testing on CPTAC-COAD, it achieved AUCs of 0.78, 0.76, and 0.61 for MSI-H, BRAF, and KRAS, respectively, outperforming previous results. This framework offers a scalable and effective solution for image-based biomarker screening and demonstrates strong potential for clinical application, particularly in resource-limited settings. The code is available at https://github.com/chenh2lab/Ensemble-MIL.

最近的研究探索了组织病理学全切片图像（WSIs）预测结直肠癌（CRC）生物标志物的方法，旨在创造经济有效的诊断工具。然而，实现强大的预测性能和跨数据集的通用性仍然是一个挑战。在这里，我们引入了一个基于深度学习的集成框架，ensemble - mil，旨在稳健地预测关键的CRC生物标志物，包括BRAF V600E、KRAS突变和MSI-H状态，并提高了跨数据集的性能。我们使用了两个独立的CRC数据集：TCGA-COAD用于模型训练和内部评估，CPTAC-COAD作为外部测试集来评估泛化性。对所有wsi进行预处理并分割成小图像块。采用肿瘤检测模型对肿瘤区域、斑块级进行识别；通过对比学习方法SimCLR提取特征。这些特征被用来训练三种多实例学习（MIL）模型：at -MIL， trans -MIL和GNN-MIL。然后将这些模型集成到最终的Ensemble-MIL框架中。在TCGA-COAD内部测试中，该方法对MSI-H、BRAF和KRAS的曲线下面积（AUC）得分分别为0.90、0.87和0.64。在CPTAC-COAD的外部测试中，MSI-H、BRAF和KRAS的auc分别为0.78、0.76和0.61，优于之前的结果。该框架为基于图像的生物标志物筛选提供了一种可扩展且有效的解决方案，并显示出强大的临床应用潜力，特别是在资源有限的环境中。代码可在https://github.com/chenh2lab/Ensemble-MIL上获得。

{"title":"Ensemble-MIL: deep learning-based ensemble framework for biomarker prediction from histopathological images in colorectal cancer","authors":"Geng-Yun Tien , Yu-Chia Chen , Liang-Chuan Lai , Tzu-Pin Lu , Mong-Hsun Tsai , Eric Y. Chuang , Hsiang-Han Chen","doi":"10.1016/j.bspc.2026.109759","DOIUrl":"10.1016/j.bspc.2026.109759","url":null,"abstract":"<div><div>Recent studies have explored histopathological whole slide images (WSIs) for predicting colorectal cancer (CRC) biomarkers, aiming to create cost-effective and efficient diagnostic tools. However, achieving strong predictive performance and generalizability across datasets remains a challenge. Here, we introduce a deep learning-based ensemble framework, Ensemble-MIL, designed to robustly predict key CRC biomarkers, including BRAF V600E, KRAS mutations, and MSI-H status, with improved cross-dataset performance. We employed two independent CRC datasets: TCGA-COAD for model training and internal evaluation, and CPTAC-COAD as an external test set to assess generalizability. All WSIs were preprocessed and divided into small image patches. A tumor detection model was applied to identify tumor regions, and patch-level; features were extracted via SimCLR, a contrastive learning method. These features were utilized to train three multiple instance learning (MIL) models: Att-MIL, Tran-MIL, and GNN-MIL. The models were then integrated into the final Ensemble-MIL framework. In internal testing with TCGA-COAD, the proposed method achieved area under the curve (AUC) scores of 0.90, 0.87, and 0.64 for MSI-H, BRAF, and KRAS, respectively. In external testing on CPTAC-COAD, it achieved AUCs of 0.78, 0.76, and 0.61 for MSI-H, BRAF, and KRAS, respectively, outperforming previous results. This framework offers a scalable and effective solution for image-based biomarker screening and demonstrates strong potential for clinical application, particularly in resource-limited settings. The code is available at <span><span>https://github.com/chenh2lab/Ensemble-MIL</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109759"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192619","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

DS-Mamba: Dynamic snake visual state space model for vessel segmentation DS-Mamba：用于血管分割的动态蛇形视觉状态空间模型

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-11 DOI: 10.1016/j.bspc.2026.109783

Zixuan Liu , Yao Cheng , Zhaoqin Huang , Wanqiang Cai , Kidiyo Kpalma , Dengwang Li , Hua Lu

Accurately segmenting vascular networks holds significant clinical implications for disease diagnosis and analysis. However, the intrinsically elongated, serpentine, and multi-scale nature of these structures poses a significant challenge, with existing methods often struggling to preserve both global connectivity and local morphological fidelity. To address this challenge, we propose a novel deep-learning architecture, termed Dynamic Snake Mamba (DS-Mamba), inspired by the sinuous morphology of vessels. DS-Mamba first leverages a Mamba backbone, composed of Residual Visual State Space (ResVSS) blocks, to establish a topologically coherent global representation of the vascular network. Subsequently, Dynamic Snake Convolutions (DSC) are strategically embedded to enhance the feature extraction of local serpentine details. To further improve its capabilities, the architecture incorporates three key components: (1) a Multi-scale Information Mamba Fusion (MIMF) mechanism that aggregates features from all encoder stages; (2) a Snake Tokenized Kolmogorov-Arnold Network (STK) at the bottleneck to manage complex feature interactions; and (3) Global–Local Information Fusion (GLIF) blocks that merge the global context with serpentine details. The efficacy of DS-Mamba was validated through comprehensive experiments on eight diverse tubular structure datasets. Results demonstrate that our approach not only achieves state-of-the-art performance in connectivity and morphological fidelity but also exhibits superior accuracy in segmenting thin, low-contrast vessels and robust resilience against high-intensity image noise. Furthermore, rigorous capacity-controlled ablation studies confirm that the performance gains stem from the synergistic architectural design rather than parameter scaling. Finally, inference speed analysis verifies the model’s feasibility for real-time clinical applications.

准确分割血管网络对疾病的诊断和分析具有重要的临床意义。然而，这些结构固有的细长、蛇形和多尺度性质带来了重大挑战，现有的方法往往难以保持全局连通性和局部形态保真度。为了解决这一挑战，我们提出了一种新的深度学习架构，称为动态蛇曼巴（DS-Mamba），灵感来自血管的弯曲形态。DS-Mamba首先利用由残余视觉状态空间（ResVSS）块组成的曼巴主干来建立血管网络的拓扑一致的全局表示。随后，策略性地嵌入动态蛇形卷积（DSC）来增强局部蛇形细节的特征提取。为了进一步提高其功能，该架构包含三个关键组件：(1)多尺度信息曼巴融合（MIMF）机制，该机制聚合了所有编码器阶段的功能；(2)在瓶颈处使用蛇形token化Kolmogorov-Arnold网络（STK）来管理复杂的特征交互；(3)全局-局部信息融合（GLIF）块，将全局上下文与蛇形细节合并。通过8个不同管状结构数据集的综合实验，验证了DS-Mamba的有效性。结果表明，我们的方法不仅在连通性和形态保真度方面实现了最先进的性能，而且在分割薄、低对比度血管方面表现出卓越的准确性，并且对高强度图像噪声具有强大的恢复能力。此外，严格的容量控制消融研究证实，性能的提高源于协同的建筑设计，而不是参数缩放。最后，通过推理速度分析验证了该模型在实时临床应用中的可行性。

{"title":"DS-Mamba: Dynamic snake visual state space model for vessel segmentation","authors":"Zixuan Liu , Yao Cheng , Zhaoqin Huang , Wanqiang Cai , Kidiyo Kpalma , Dengwang Li , Hua Lu","doi":"10.1016/j.bspc.2026.109783","DOIUrl":"10.1016/j.bspc.2026.109783","url":null,"abstract":"<div><div>Accurately segmenting vascular networks holds significant clinical implications for disease diagnosis and analysis. However, the intrinsically elongated, serpentine, and multi-scale nature of these structures poses a significant challenge, with existing methods often struggling to preserve both global connectivity and local morphological fidelity. To address this challenge, we propose a novel deep-learning architecture, termed Dynamic Snake Mamba (DS-Mamba), inspired by the sinuous morphology of vessels. DS-Mamba first leverages a Mamba backbone, composed of Residual Visual State Space (ResVSS) blocks, to establish a topologically coherent global representation of the vascular network. Subsequently, Dynamic Snake Convolutions (DSC) are strategically embedded to enhance the feature extraction of local serpentine details. To further improve its capabilities, the architecture incorporates three key components: (1) a Multi-scale Information Mamba Fusion (MIMF) mechanism that aggregates features from all encoder stages; (2) a Snake Tokenized Kolmogorov-Arnold Network (STK) at the bottleneck to manage complex feature interactions; and (3) Global–Local Information Fusion (GLIF) blocks that merge the global context with serpentine details. The efficacy of DS-Mamba was validated through comprehensive experiments on eight diverse tubular structure datasets. Results demonstrate that our approach not only achieves state-of-the-art performance in connectivity and morphological fidelity but also exhibits superior accuracy in segmenting thin, low-contrast vessels and robust resilience against high-intensity image noise. Furthermore, rigorous capacity-controlled ablation studies confirm that the performance gains stem from the synergistic architectural design rather than parameter scaling. Finally, inference speed analysis verifies the model’s feasibility for real-time clinical applications.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109783"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192774","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

A novel diagnosis framework of melanoma skin cancer using adaptive and attention-based deep network with mamba-efficient-UNet-aided abnormality segmentation 基于mamba-efficient- unet辅助异常分割的适应性和基于注意力的深度网络的黑色素瘤皮肤癌诊断新框架

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-11 DOI: 10.1016/j.bspc.2026.109751

R. Lakshmi , B. Arthi

Melanoma is one of the most harmful dermatologic cancers worldwide. Recently, medical sector has undergone a significant evolution through the utilization of automatic detection frameworks that assist medical professionals for identifying cancer tissues growth more precisely. However, machine learning approaches struggle to present accurate diagnosis results because of their inability to determine the visual similarity among the malignant as well as benign tissues. Addressing this limitation, involves deep learning techniques in medical sector, due to their competence in detection tasks. However, imbalanced databases and image occlusion negatively impacted the accuracy of the detection model. Thus, a promising diagnosis framework is designed for melanoma skin cancer. Initially, required images were aggregated from available datasets. Furthermore, input images were subjected into Mamba-Efficient-UNet (MEUNet) for segmenting abnormalities. Segmented images were fed into Adaptive and Attention-based Efficient Net B7 with Long Short Term Memory (AAENB7-LSTM) layer for diagnosing melanoma skin cancer. Here, the AAENB7-LSTM parameters are tuned by Modified Random Variable-based Dollmaker Optimization Algorithm (MRV-DOA). Lastly, efficiency assessment of developed model is carried out by analyzing with other related techniques to showcase its effectiveness in melanoma skin cancer. The findings of recommended method attain 98.1% in terms of F1-score on dataset 1 with batch size 32. In addition, efficiency of suggested network is analyzed with existing classifiers to demonstrate 98.2% for precision with batch size 8 on dataset 2 than other existing CNN, MFEUsLNet, Grand-CAM, and AENB7-LSTM model.

黑色素瘤是世界上最有害的皮肤癌症之一。最近，医疗部门通过利用自动检测框架，帮助医疗专业人员更准确地识别癌症组织的生长，经历了重大的演变。然而，机器学习方法难以提供准确的诊断结果，因为它们无法确定恶性组织和良性组织之间的视觉相似性。解决这一限制涉及医疗部门的深度学习技术，因为它们在检测任务中的能力。然而，数据库不平衡和图像遮挡对检测模型的准确性产生了负面影响。因此，为黑色素瘤皮肤癌设计了一个有希望的诊断框架。最初，从可用的数据集中聚合所需的图像。此外，对输入图像进行Mamba-Efficient-UNet （MEUNet）分割异常。将分割后的图像输入到具有长短期记忆的Adaptive and Attention-based Efficient Net B7 （AAENB7-LSTM）层，用于黑色素瘤皮肤癌的诊断。在这里，AAENB7-LSTM参数通过基于改进随机变量的模型优化算法（MRV-DOA）进行调整。最后，通过与其他相关技术的分析，对所开发的模型进行有效性评估，展示其在黑色素瘤皮肤癌中的有效性。推荐的方法在批大小为32的数据集1上的f1得分达到98.1%。此外，使用现有的分类器对所建议网络的效率进行了分析，在数据集2上，批大小为8的网络的准确率比其他现有的CNN、MFEUsLNet、grande - cam和AENB7-LSTM模型高98.2%。

{"title":"A novel diagnosis framework of melanoma skin cancer using adaptive and attention-based deep network with mamba-efficient-UNet-aided abnormality segmentation","authors":"R. Lakshmi , B. Arthi","doi":"10.1016/j.bspc.2026.109751","DOIUrl":"10.1016/j.bspc.2026.109751","url":null,"abstract":"<div><div>Melanoma is one of the most harmful dermatologic cancers worldwide. Recently, medical sector has undergone a significant evolution through the utilization of automatic detection frameworks that assist medical professionals for identifying cancer tissues growth more precisely. However, machine learning approaches struggle to present accurate diagnosis results because of their inability to determine the visual similarity among the malignant as well as benign tissues. Addressing this limitation, involves deep learning techniques in medical sector, due to their competence in detection tasks. However, imbalanced databases and image occlusion negatively impacted the accuracy of the detection model. Thus, a promising diagnosis framework is designed for melanoma skin cancer. Initially, required images were aggregated from available datasets. Furthermore, input images were subjected into Mamba-Efficient-UNet (MEUNet) for segmenting abnormalities. Segmented images were fed into Adaptive and Attention-based Efficient Net B7 with Long Short Term Memory (AAENB7-LSTM) layer for diagnosing melanoma skin cancer. Here, the AAENB7-LSTM parameters are tuned by Modified Random Variable-based Dollmaker Optimization Algorithm (MRV-DOA). Lastly, efficiency assessment of developed model is carried out by analyzing with other related techniques to showcase its effectiveness in melanoma skin cancer. The findings of recommended method attain 98.1% in terms of F1-score on dataset 1 with batch size 32. In addition, efficiency of suggested network is analyzed with existing classifiers to demonstrate 98.2% for precision with batch size 8 on dataset 2 than other existing CNN, MFEUsLNet, Grand-CAM, and AENB7-LSTM model.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109751"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192792","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

Intelligent diagnosis of aortic lesion using non-contrast CT: An integrated approach combining deep learning and morphological characteristics 非对比CT对主动脉病变的智能诊断：一种深度学习与形态学特征相结合的综合方法

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-12 DOI: 10.1016/j.bspc.2026.109823

Mingliang Yang , Aoxue Mei , Xiaolin Guo , Jianxing Hu , Yue Zhang , Xiangbing Bian , Jiayu Huang , Sulian Su , Jinhao Lyu , Xin Lou

Traditional diagnosis of aortic lesions typically relies on CT angiography (CTA), which provides detailed vascular structures but requires contrast agents. Non-contrast CT (NCCT) is more accessible for routine use but suffers from inherently low soft-tissue contrast, making visual assessment and manual delineation of non-calcified lesions challenging and subjective. In this study, we propose a hybrid algorithm designed to perform comprehensive segmentation and classification of various aortic lesions, including aneurysms, dissections, lumen stenosis, vessel wall calcification, and normal conditions, directly from NCCT images. The method focuses on utilizing advanced deep learning segmentation techniques along with knowledge of aortic morphology to accurately diagnose aortic lesions. Data from three centers (n = 435) were used for algorithm development, and data from three additional centers (n = 493) were used for testing. Detailed comparisons with the baseline method showed that our proposed deep learning approach outperformed other methods, achieving an accuracy of 89.7%, a sensitivity of 79.4%, a specificity of 97.1%, and an F1 score of 0.790 for multiple lesion diagnosis tasks. Further integration of morphological features improved the diagnostic accuracy to 90.3% and the overall performance metric, the F1 score, to 0.803. These experimental results demonstrate the feasibility of diagnosing aortic lesions from NCCT and validate the effectiveness of combining deep learning with morphological characteristics.

主动脉病变的传统诊断通常依赖于CT血管造影（CTA），它提供了详细的血管结构，但需要造影剂。非对比CT （NCCT）更容易用于常规使用，但其固有的软组织对比度较低，使得视觉评估和手工描绘非钙化病变具有挑战性和主观性。在这项研究中，我们提出了一种混合算法，旨在对各种主动脉病变进行全面的分割和分类，包括动脉瘤、夹层、管腔狭窄、血管壁钙化和正常情况，直接来自NCCT图像。该方法着重于利用先进的深度学习分割技术以及主动脉形态知识来准确诊断主动脉病变。来自三个中心（n = 435）的数据用于算法开发，另外三个中心（n = 493）的数据用于测试。与基线方法的详细比较表明，我们提出的深度学习方法优于其他方法，在多病变诊断任务中，准确率为89.7%，灵敏度为79.4%，特异性为97.1%，F1评分为0.790。形态学特征的进一步整合将诊断准确率提高到90.3%，整体性能指标F1得分提高到0.803。这些实验结果证明了NCCT诊断主动脉病变的可行性，验证了深度学习与形态学特征相结合的有效性。

{"title":"Intelligent diagnosis of aortic lesion using non-contrast CT: An integrated approach combining deep learning and morphological characteristics","authors":"Mingliang Yang , Aoxue Mei , Xiaolin Guo , Jianxing Hu , Yue Zhang , Xiangbing Bian , Jiayu Huang , Sulian Su , Jinhao Lyu , Xin Lou","doi":"10.1016/j.bspc.2026.109823","DOIUrl":"10.1016/j.bspc.2026.109823","url":null,"abstract":"<div><div>Traditional diagnosis of aortic lesions typically relies on CT angiography (CTA), which provides detailed vascular structures but requires contrast agents. Non-contrast CT (NCCT) is more accessible for routine use but suffers from inherently low soft-tissue contrast, making visual assessment and manual delineation of non-calcified lesions challenging and subjective. In this study, we propose a hybrid algorithm designed to perform comprehensive segmentation and classification of various aortic lesions, including aneurysms, dissections, lumen stenosis, vessel wall calcification, and normal conditions, directly from NCCT images. The method focuses on utilizing advanced deep learning segmentation techniques along with knowledge of aortic morphology to accurately diagnose aortic lesions. Data from three centers (n = 435) were used for algorithm development, and data from three additional centers (n = 493) were used for testing. Detailed comparisons with the baseline method showed that our proposed deep learning approach outperformed other methods, achieving an accuracy of 89.7%, a sensitivity of 79.4%, a specificity of 97.1%, and an F1 score of 0.790 for multiple lesion diagnosis tasks. Further integration of morphological features improved the diagnostic accuracy to 90.3% and the overall performance metric, the F1 score, to 0.803. These experimental results demonstrate the feasibility of diagnosing aortic lesions from NCCT and validate the effectiveness of combining deep learning with morphological characteristics.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109823"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192790","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

Neural synchrony and attention dynamics during naturalistic video viewing: a gender comparison using EEG and deep learning approaches 自然视频观看过程中的神经同步和注意动力学：使用脑电图和深度学习方法的性别比较

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-15 Epub Date: 2026-02-09 DOI: 10.1016/j.bspc.2026.109763

Tengis Tserendondog , Bat-Erdene Gotov , Uurtsaikh Luvsansambuu , Dong-Sung Pae , Hansaem Park

This study introduces a hybrid electroencephalographic (EEG) framework that integrates classical analyses with an interpretable attention-driven deep learning model to examine gender-related neural differences during naturalistic video viewing. Seventy-four adults (37 males, 37 females) watched a 117-second silent video while 14-channel EEG signals were recorded. Classical measures, including inter-subject correlation (ISC) and wavelet-based time–frequency mapping, revealed that males exhibited significantly higher ISC than females (0.065 ± 0.040 vs. 0.031 ± 0.033; t(72) = 3.88, p = 0.0002, Cohen’s d = 0.91), indicating stronger inter-brain synchrony during rapid scene transitions. Spectral analysis further demonstrated stronger frontal beta power in males, associated with top-down control, and enhanced parietal alpha activity in females, linked to sensory integration. To move beyond aggregated metrics, a hierarchical attention-based autoencoder was employed to reconstruct EEG signals while assigning time-resolved attention weights. Temporal attention profiles were computed in 4-second windows and compared across genders using Welch’s t-tests with FDR correction. Six windows (28, 32, 36, 60, 76, 80 s) showed significant group differences (q < 0.05), demonstrating dynamic gender-specific attentional shifts. These divergences aligned with ISC and spectral peaks, indicating that the deep-learning model captured the same engagement-relevant segments identified by classical metrics, but with higher temporal precision. By linking physiological markers (ISC and frequency dynamics) with interpretable temporal salience, the framework provides a coherent multi-scale account of how males and females differentially process continuous visual stimuli. The approach advances methodological transparency in EEG-based deep learning and supports applications in personalized media design, adaptive learning environments, neuromarketing, and gender-aware brain–computer interfaces.

本研究引入了一种混合脑电图（EEG）框架，该框架将经典分析与可解释的注意力驱动深度学习模型相结合，以研究自然视频观看过程中与性别相关的神经差异。74名成年人（37名男性，37名女性）观看了一段117秒的无声视频，同时记录了14个通道的脑电图信号。经典测量方法包括受试者间相关（ISC）和基于小波的时频映射，结果显示男性的ISC显著高于女性（0.065±0.040 vs 0.031±0.033;t(72) = 3.88, p = 0.0002, Cohen’s d = 0.91），表明在快速场景转换过程中大脑间同步更强。光谱分析进一步表明，男性的额叶β能量更强，与自上而下的控制有关，而女性的顶叶α活性增强，与感觉整合有关。为了超越聚合指标，在分配时间分辨注意力权重的同时，采用基于分层注意的自编码器重构脑电信号。在4秒窗口内计算时间注意概况，并使用带有FDR校正的Welch’s t检验比较性别间的差异。6个窗口（28、32、36、60、76、80 s）组间差异显著（q < 0.05），显示出动态的性别注意转移。这些差异与ISC和光谱峰一致，表明深度学习模型捕获了与经典指标识别的相同的参与相关片段，但具有更高的时间精度。通过将生理标记（ISC和频率动态）与可解释的时间显著性联系起来，该框架提供了一个连贯的多尺度说明男性和女性如何不同地处理连续视觉刺激。该方法提高了基于脑电图的深度学习方法的透明度，并支持在个性化媒体设计、自适应学习环境、神经营销和性别意识脑机接口中的应用。

{"title":"Neural synchrony and attention dynamics during naturalistic video viewing: a gender comparison using EEG and deep learning approaches","authors":"Tengis Tserendondog , Bat-Erdene Gotov , Uurtsaikh Luvsansambuu , Dong-Sung Pae , Hansaem Park","doi":"10.1016/j.bspc.2026.109763","DOIUrl":"10.1016/j.bspc.2026.109763","url":null,"abstract":"<div><div>This study introduces a hybrid electroencephalographic (EEG) framework that integrates classical analyses with an interpretable attention-driven deep learning model to examine gender-related neural differences during naturalistic video viewing. Seventy-four adults (37 males, 37 females) watched a 117-second silent video while 14-channel EEG signals were recorded. Classical measures, including inter-subject correlation (ISC) and wavelet-based time–frequency mapping, revealed that males exhibited significantly higher ISC than females (0.065 ± 0.040 vs. 0.031 ± 0.033; t(72) = 3.88, p = 0.0002, Cohen’s d = 0.91), indicating stronger inter-brain synchrony during rapid scene transitions. Spectral analysis further demonstrated stronger frontal beta power in males, associated with top-down control, and enhanced parietal alpha activity in females, linked to sensory integration. To move beyond aggregated metrics, a hierarchical attention-based autoencoder was employed to reconstruct EEG signals while assigning time-resolved attention weights. Temporal attention profiles were computed in 4-second windows and compared across genders using Welch’s t-tests with FDR correction. Six windows (28, 32, 36, 60, 76, 80 s) showed significant group differences (q < 0.05), demonstrating dynamic gender-specific attentional shifts. These divergences aligned with ISC and spectral peaks, indicating that the deep-learning model captured the same engagement-relevant segments identified by classical metrics, but with higher temporal precision. By linking physiological markers (ISC and frequency dynamics) with interpretable temporal salience, the framework provides a coherent multi-scale account of how males and females differentially process continuous visual stimuli. The approach advances methodological transparency in EEG-based deep learning and supports applications in personalized media design, adaptive learning environments, neuromarketing, and gender-aware brain–computer interfaces.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"119 ","pages":"Article 109763"},"PeriodicalIF":4.9,"publicationDate":"2026-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192861","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

CycleGAN-based prosody and spectrum modeling for Mandarin touch-controlled Electrolaryngeal speech enhancement 基于cyclegan的普通话触控电喉语音增强韵律和频谱建模

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control

Pub Date : 2026-06-01 Epub Date: 2026-02-06 DOI: 10.1016/j.bspc.2026.109746

Jie Zhou , Li Wang , Fengji Li , Shaochuan Zhang , Fei Shen , Fan Fan , Tao Liu , Xiaohong Chen , Haijun Niu

The application of Electrolarynx (EL) for tonal language laryngectomees remains challenging due to the difficulty in achieving tonal completion without useful fundamental frequency (F0) information. This study proposes a novel Mandarin EL speech enhancement framework by integrating the prior F0 information provided by finger movements, combined with the Cycle-Consistent Adversarial Network (CycleGAN) and Continuous Wavelet Transform (CWT). For prosody modeling, we exploit the hierarchical structure inherent in Mandarin prosody by using CWT decomposition coefficients as a feature representation of F0. For spectral conversion, we extract Mel-frequency cepstral coefficients (MCEP) as spectral features. These two feature sets were trained separately using the CycleGAN model. In results, acoustic feature analysis indicates that the four tones after converted are closer to normal tones in both F0 value and F0 contour. The spectrogram of the converted speech is also more similar to that of normal speech, and compensates for low-frequency energy missing below 500 Hz. Both subjective and objective evaluations demonstrate the effectiveness of the proposed method in Mandarin EL speech enhancement. This study also provides a novel approach for EL speech enhancement in other tonal languages. And it may provide valuable insights and guidance for future improvement in tonal EL devices development and EL speech enhancement.

由于没有有用的基频（F0）信息难以实现音调补全，因此电喉（EL）在声调语言喉切除术中的应用仍然具有挑战性。本研究结合循环一致对抗网络（CycleGAN）和连续小波变换（CWT），将手指运动提供的先验F0信息整合在一起，提出了一种新的普通话EL语音增强框架。对于韵律建模，我们利用汉语韵律固有的层次结构，使用CWT分解系数作为F0的特征表示。对于频谱转换，我们提取Mel-frequency倒谱系数（MCEP）作为频谱特征。这两个特征集分别使用CycleGAN模型进行训练。结果声学特征分析表明，转换后的四个音调在F0值和F0轮廓上都更接近正常音调。转换后的语音频谱图也更接近于正常语音，并补偿了500 Hz以下的低频能量缺失。主观和客观的评价都证明了该方法在普通话英语语音增强中的有效性。本研究也为其他声调语言的EL语音增强提供了一种新的方法。为今后声调语音器件的开发和语音增强提供有价值的见解和指导。

{"title":"CycleGAN-based prosody and spectrum modeling for Mandarin touch-controlled Electrolaryngeal speech enhancement","authors":"Jie Zhou , Li Wang , Fengji Li , Shaochuan Zhang , Fei Shen , Fan Fan , Tao Liu , Xiaohong Chen , Haijun Niu","doi":"10.1016/j.bspc.2026.109746","DOIUrl":"10.1016/j.bspc.2026.109746","url":null,"abstract":"<div><div>The application of Electrolarynx (EL) for tonal language laryngectomees remains challenging due to the difficulty in achieving tonal completion without useful fundamental frequency (F0) information. This study proposes a novel Mandarin EL speech enhancement framework by integrating the prior F0 information provided by finger movements, combined with the Cycle-Consistent Adversarial Network (CycleGAN) and Continuous Wavelet Transform (CWT). For prosody modeling, we exploit the hierarchical structure inherent in Mandarin prosody by using CWT decomposition coefficients as a feature representation of F0. For spectral conversion, we extract Mel-frequency cepstral coefficients (MCEP) as spectral features. These two feature sets were trained separately using the CycleGAN model. In results, acoustic feature analysis indicates that the four tones after converted are closer to normal tones in both F0 value and F0 contour. The spectrogram of the converted speech is also more similar to that of normal speech, and compensates for low-frequency energy missing below 500 Hz. Both subjective and objective evaluations demonstrate the effectiveness of the proposed method in Mandarin EL speech enhancement. This study also provides a novel approach for EL speech enhancement in other tonal languages. And it may provide valuable insights and guidance for future improvement in tonal EL devices development and EL speech enhancement.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"118 ","pages":"Article 109746"},"PeriodicalIF":4.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146174420","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