Zhenyou Liu , Zhanlong Zhang , Wei He , Yuxin Fang , Shaohua Hu
{"title":"Study on novel internal and external electrode configuration and optimization methods for lung impedance imaging in intubated patients","authors":"Zhenyou Liu , Zhanlong Zhang , Wei He , Yuxin Fang , Shaohua Hu","doi":"10.1016/j.bspc.2025.108809","DOIUrl":"10.1016/j.bspc.2025.108809","url":null,"abstract":"<div><div>This study introduces a novel electrode configuration and optimization method for monitoring pulmonary ventilation in intubated patients using electrical impedance tomography (EIT). By positioning electrodes within the esophagus or trachea through existing intubation, the effectiveness of lung ventilation imaging is significantly enhanced. Furthermore, the optimization of external electrode arrays for lung impedance imaging with internal electrodes is investigated. Electrode placement was optimized as a variable based on the principles of focused imaging, constructing numerical sensitivity field matrices for the lung region and overall sensitivity field condition numbers as optimization objectives. A fast elitist multi-objective genetic algorithm (NSGA-III) was employed to optimize the distribution of peripheral electrode arrays. Experimental results from thoracic physical models demonstrated that, with the same total number of electrodes, the introduction of internal electrodes combined with the optimization of external arrays reduced the average relative error (RE) of image reconstruction by 66.31 %, increased the reconstruction correlation coefficient (CC) by 50.72 %, and enhanced the Structural Similarity Index Measure (SSIM) of reconstruction by 85.96 %. This research presents a non-invasive method for introducing internal electrodes, significantly improving the capability to acquire information from the central thoracic region and the accuracy of pulmonary ventilation imaging in intubated patients. Additionally, the study advances the optimization process for impedance electrode arrays.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108809"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265944","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}
Huiyun Zhang , Zilong Pang , Puyang Zhao , Gaigai Tang , Lingfeng Shen , Guanghui Wang
{"title":"Pre-attentive speech signal processing with adaptive routing for emotion recognition","authors":"Huiyun Zhang , Zilong Pang , Puyang Zhao , Gaigai Tang , Lingfeng Shen , Guanghui Wang","doi":"10.1016/j.bspc.2025.108782","DOIUrl":"10.1016/j.bspc.2025.108782","url":null,"abstract":"<div><div>Emotion recognition from speech is essential for various applications in human–computer interaction, customer service, healthcare, and entertainment. However, developing robust and reproducible Speech emotion recognition (SER) systems is challenging due to the complexity of emotions and variability in speech signals. In this paper, we first define the concept of reproducibility in the context of deep learning models. We then introduce SpeechNet, a novel deep learning model designed to enhance reproducibility and robustness in SER. SpeechNet integrates multiple advanced components: speech recall, speech attention, and speech signal refinement modules to effectively capture temporal dependencies and emotional cues in speech signal. Additionally, it incorporates a pre-attention mechanism and a modified routing technique to improve feature emphasis and processing efficiency. We also explore effective acoustic feature fusion technique. Extensive experiments on several benchmark datasets demonstrate that the SpeechNet model achieves better performance and reproducibility compared to existing models. By addressing reproducibility and robustness, SpeechNet sets a new standard in SER, facilitating reliable and practical applications.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108782"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265948","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}
Juntao Ding , Qian Liu , Bingo Wing-Kuen Ling , Wenli Li
{"title":"Heart rate estimation based on joint convolutional neural network and conditional generative adversarial network via heart rate variabilities and other features extracted from photoplethysmograms","authors":"Juntao Ding , Qian Liu , Bingo Wing-Kuen Ling , Wenli Li","doi":"10.1016/j.bspc.2025.108831","DOIUrl":"10.1016/j.bspc.2025.108831","url":null,"abstract":"<div><div>Accurate heart rate estimation is crucial for diagnosing and preventing cardiovascular diseases. Traditional methods rely on electrocardiograms (ECGs), which require attaching multiple electrodes to the body, making the process cumbersome and limiting its practicality. In contrast, photoplethysmograms (PPGs) offer a simpler and more convenient way to capture cardiovascular information, including heart rate. However, PPG-based heart rate measurements often differ from ECG-based ones due to timing differences in signal peaks and heart rate variability (HRV). To address these challenges, this paper proposes a novel approach combining Convolutional Neural Networks (CNNs) and Conditional Generative Adversarial Networks (CGANs) for heart rate estimation using HRV and other features extracted from PPGs. The CNN first estimates the heart rate, and the CGAN refines this estimation. The CGAN’s generator uses conditional information to produce more accurate heart rates, while the discriminator, equipped with residual blocks and a self-attention mechanism, classifies differences between actual and generated heart rates through a multi-layer convolutional network. This design mitigates gradient vanishing and enhances model stability, allowing the system to capture complex relationships between CNN-estimated heart rates and ECG-measured ones. To further improve accuracy, a perceptual loss function based on conditional information is used to minimize errors between estimated and actual heart rates. Simulation results show significant improvements in Pearson’s correlation coefficient (ρ), Frechet distance (FD), root mean square error (RMSE), and mean absolute distortion (MAD), demonstrating the method’s effectiveness and reliability. This approach has practical applications in wearable health devices, enabling continuous and non-invasive heart rate monitoring for early detection and management of cardiovascular conditions.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108831"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145266490","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}
Jiahao Xie , Saiqi He , Youyao Fu , Xin Tao , Shiqing Zhang , Jiangxiong Fang , Xiaoming Zhao , Guoyu Wang , Zhaohui Yang , Hongsheng Lu
{"title":"ThyHisTer: A new thyroid histopathology image dataset for ternary classification of thyroid cancer","authors":"Jiahao Xie , Saiqi He , Youyao Fu , Xin Tao , Shiqing Zhang , Jiangxiong Fang , Xiaoming Zhao , Guoyu Wang , Zhaohui Yang , Hongsheng Lu","doi":"10.1016/j.bspc.2025.108819","DOIUrl":"10.1016/j.bspc.2025.108819","url":null,"abstract":"<div><div>Thyroid cancer is a common type of endocrine cancer, and its incidence rate has been increasing year by year. Due to the scarcity of publicly accessible histopathology image datasets for thyroid cancer diagnosis, it is difficult to develop automatic Computer-aided Diagnostic (CAD) systems for enhancing the accuracy of thyroid cancer diagnosis. To address this issue, this work aims to construct a novel publicly accessible <strong>thy</strong>roid <strong>his</strong>topathology image dataset for <strong>ter</strong>nary classification of thyroid cancer, namely ThyHisTer. Furthermore, to present a benchmarking performance evaluation on the ThyHisTer dataset, this work explores the performance of various deep learning methods on thyroid cancer classification tasks. Additionally, this work proposes a new lightweight deep learning model called SeSepViT for thyroid cancer classification, which integrates the advantages of Squeeze and Excitation (SE) networks and Separable Vision Transformer (SepViT). This work conducts extensive experiments on the collected ThyHisTer dataset, and utilize various deep learning methods to validate the performance of thyroid cancer classification. Experimental results show that the proposed SeSepViT achieves highly comparable performance to other used deep learning methods on thyroid cancer classification tasks, and simultaneously exhibits relatively lower computational cost. The release of ThyHisTer is expected to facilitate the application of advanced deep learning methods for automatic thyroid cancer diagnosis, thereby assisting doctors in early detecting thyroid cancer in clinical practice. The code is available on <span><span>https://github.com/beatttt/ThyHisTer</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108819"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145264969","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}
K.Venkatesh Guru , Vignesh Janarthanan , M. Jaganathan , V. Senthil kumar
{"title":"A multi-modal ensemble framework for breast cancer segmentation and classification using genetic U-Net and HBoost","authors":"K.Venkatesh Guru , Vignesh Janarthanan , M. Jaganathan , V. Senthil kumar","doi":"10.1016/j.bspc.2025.108827","DOIUrl":"10.1016/j.bspc.2025.108827","url":null,"abstract":"<div><div>Breast cancer remains a leading cause of mortality among women worldwide, and early, accurate diagnosis is essential to improving treatment outcomes. Traditional segmentation and classification models often fail in ultrasound imaging due to noisy inputs, imbalanced class distributions, and computational inefficiencies. To address these challenges, we propose M2G-HBoost, a multi-modal ensemble framework explicitly designed for robustness in noisy and unbalanced data conditions. The framework integrates cosine similarity–based Graph Convolutional Network (GCN) augmentation to enrich feature diversity by modeling global topological and attribute relationships, M2GCNet for joint spatial and channel-wise dependency modeling, Genetic U-Net optimized via evolutionary algorithms for parameter-efficient high-accuracy segmentation, and HBoost, a heterogeneous boosting ensemble, for resilient classification. This modular design was chosen to employ complementary strengths: GCN augmentation for data diversity, M2GCNet for rich feature extraction, Genetic U-Net for segmentation precision with low complexity, and HBoost for classification stability under imbalance. On the BUSI dataset, M2G-HBoost achieved a Dice score of 93.58 % and IoU of 91.81 %, outperforming CBAM-RIUnet (76.25 % Dice, 80.73 % IoU), PCA-UNet (80.47 %, 84.12 %), EDCNN (85.99 %, 86.24 %), ELRL-E (87.14 %, 88.67 %), and SegEIR-Net (91.07 %, 89.49 %). In classification, the model reached 96.45 % (benign), 98.53 % (malignant), and 98.87 % (normal), exceeding AdaBoost, XGBoost, and Gradient Boosting. These results demonstrate the superiority and clinical applicability of the proposed method, offering a robust, accurate, and efficient solution for breast cancer segmentation and classification in ultrasound imaging.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108827"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265275","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}
Yan Wang , Fa Yang , Xiaoying Pan , Hao Wang , Xiaopan Xu , Yiran Pan , Kun Yang , Ge Ma , Zhangchao Hao , Huanxiang Liu , Peng Yang
{"title":"Improving uneven exposure using color characteristics as a priori information in endoscopic images","authors":"Yan Wang , Fa Yang , Xiaoying Pan , Hao Wang , Xiaopan Xu , Yiran Pan , Kun Yang , Ge Ma , Zhangchao Hao , Huanxiang Liu , Peng Yang","doi":"10.1016/j.bspc.2025.108825","DOIUrl":"10.1016/j.bspc.2025.108825","url":null,"abstract":"<div><div>Endoscopic examinations often encounter color discrepancies in images due to uneven lighting and variations in color temperature of the light source, which directly affect the accuracy of disease screening and diagnosis. Traditional methods for enhancing low-quality endoscopic images primarily concentrate on correcting underexposed images, but do not simultaneously address properly exposed regions or suppress overexposed regions in the image. To address these issues, this study proposes a novel network structure, EndoUEI, which leverages color priors to improve over / underexposure of endoscopic images. The EndoUEI integrates an encoder-decoder framework with embedded color features, a Retinex exposure correction module, and an exposure fusion module. Within the encoder-decoder, an efficient multiscale attention module is employed to enhance feature representation and capture both short- and long-range pixel dependencies. Multi-scale color information are employed as color features to guide the segmenting of unevenly exposed areas. This approach effectively mitigates uneven exposure levels in low-quality endoscopic images while enriching the overall color information. On the colonoscopic dataset, the Peak Signal-to-Noise Ratio reached 21.06, while on the real nasopharyngeal dataset, the Natural Image Quality Evaluator was only 7.35, and the parameter number of the model is only 0.646 M. These results demonstrate that the EndoUEI significantly enhances the image quality of colonoscopy and nasopharyngoscopy while maintaining a minimal parameter count, thereby holding greater clinical applications.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108825"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145266500","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":"DWTFrTV-DEAL: Denoising and deep ensemble learning for arrhythmia classification using ECG signals","authors":"Sumita Lamba , Manoj Diwakar","doi":"10.1016/j.bspc.2025.108751","DOIUrl":"10.1016/j.bspc.2025.108751","url":null,"abstract":"<div><div>Electrocardiogram (ECG) signals adeptly capture the intricate electrical dynamics of the cardiac system, offering a means to assess its physiological well-being. To address the rising cardiovascular diseases (CVDs), various Machine Learning (ML)/Deep Learning (DL) techniques are employed to improve the accuracy, speed, and robustness of classifying Arrhythmia. While considerable attention has been given to designing architectures and selecting data sets, the significance of pre-processing data and its classification must be underscored. The presence of diverse disturbances during the acquisition of ECG signals influences the accurate feature extraction, which leads to lower classification accuracy. Hence, the paper introduces an efficient method of denoising and classification of ECG signals to substantially enhance the accuracy of using DL models for arrhythmia classification. The proposed work in the paper delineates two distinct methodologies, one for denoising ECG signals and another for classification. The initial approach involves mitigating noise in the ECG pattern through Discrete Wavelet Transform using Fractional Total Variation (DWTFrTV), while the subsequent step entails classifying the denoised signals utilizing a combination of two stacked learner classifiers. The base learner (set of optimized classifiers) has three models and the meta learner is the regression model. The proposed model is trained and tested over the MIT-BIH Arrhythmia dataset where the results and comparative analysis with MIT-BIH Normal Sinus Rhythm Database and St. Petersburg INCART Database demonstrates that the proposed model yields a notable improvement with an average accuracy of 99.9% and MAE: 0.05 (for Training) and average accuracy 99.7% and MAE: 0.09 (for Testing) over recent existing models.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108751"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145266492","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}
Zhengwei Huang , Shidong Lian , Chunsheng Zhang , Xiaoyong Li
{"title":"Application of graph convolutional spatial temporal fusion model based on gated recurrent unit and attention mechanism in inertial signal human activity recognition","authors":"Zhengwei Huang , Shidong Lian , Chunsheng Zhang , Xiaoyong Li","doi":"10.1016/j.bspc.2025.108745","DOIUrl":"10.1016/j.bspc.2025.108745","url":null,"abstract":"<div><div>The study of human activity recognition systems based on wearable sensors is crucial in medical rehabilitation training, game entertainment, sports analysis and other fields. A new-style hybrid deep learning model based on graph convolutional network, gated recurrent unit and attention mechanism is first proposed to recognize human activity in real-time. In order to demonstrate the superiority of the proposed hybrid model, this paper trains and verifies the proposed model, convolutional neural network model and graph convolutional network model on the open source daily life activity dataset. The experiment results indicate that the accuracy and average recall rate of the proposed model is 97.8% and 97.3% respectively, which higher than convolutional neural network model and graph convolutional network model. In addition, the accuracy of the proposed model also exceeds other state-of-the-art recognition models. Therefore, the proposed novel hybrid model with higher accuracy and stronger robustness is favorable to the practice application of human activity recognition system based on wearable sensors.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108745"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145264970","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":"ERNet: A deep framework for detection and classification of lung cancer from histopathological images","authors":"Prem Chand Yadava, Subodh Srivastava","doi":"10.1016/j.bspc.2025.108817","DOIUrl":"10.1016/j.bspc.2025.108817","url":null,"abstract":"<div><div>Lung cancer is a potent condition that impacts the mortality rate. The conventional assessment of lung cancer includes microscopic biopsies. The process comprises a labor-intensive visual assessment that is subjective and necessitates the expertise of a professional pathologist. However, noise typically affects the readable features in the low-resolution digital biopsy images. Moreover, the incidence of noise affects both inter- and intra-observer interpretations by pathologists. To overcome these issues, a novel enhanced RetinaNet (ERNet) has been proposed to detect and classify the lung abnormalities in a single framework. The proposed ERNet integrates a convolutional block attention module for the refined feature extraction. Additionally, the proposed ERNet employs a generalized intersection over union bounding box loss function to precisely localize abnormalities. The proposed method utilizes LC25000 lung histopathological images for its development. To improve, and denoise the lung biopsies image datasets, ant colony fourth-order partial differential equation has been applied. The comparative qualitative, and quantitative study has been presented with respect to existing methodologies such as faster regional convolutional neural network, single shot detector, RetinaNet and detection transformer. The quantitative assessments are evaluated in terms of accuracy, true positive rate, true negative rate, precision, F-score, Jaccard index, and Dice coefficient. The following values are obtained: 98.73%, 98.04%, 98.45%, 0.98, 0.98, 0.99, 0.98, 0.98, and 0.99, respectively. The results of qualitative, quantitative with ablation analysis exhibit that the proposed method surpasses the outcomes of the other pre-existing methods.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108817"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265276","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":"Adaptive degradation-aware medical image enhancement for multi-modal diagnostics","authors":"Parthasarathy Palani, Bharath Subramani, Magudeeswaran Veluchamy","doi":"10.1016/j.bspc.2025.108822","DOIUrl":"10.1016/j.bspc.2025.108822","url":null,"abstract":"<div><div>Medical image enhancement plays a vital role in the rapid development of medical technology, focusing on improving medical image quality and clarity to support accurate diagnosis and effective treatment. However, poor medical imaging, such as imbalanced intensity or non-uniform illumination, brings significant challenges to automated diagnosis analysis and screening of diseases. This paper proposes a novel adaptive degradation-aware medical image enhancement to improve the quality of poorly illuminated medical images captured under a low-light enclosed intestinal environment while conserving the critical pathological details. In this work, a novel complementary illumination adjustment function is employed to improve the brightness of dark regions while preventing overexposure in bright areas. Then, the proposed method incorporates a guided and bilateral filter to improve delicate clinical details and anatomical structures while suppressing contrast degradation of medical images. Experiment results comprehensively illustrate the performance of contrast enhancement in clinical diagnosis by effectively preserving color information and structure details. Extensive experimental evaluation demonstrates the superior performance of the proposed method in terms of qualitative and different quantitative metrics compared to other recent existing methods.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108822"},"PeriodicalIF":4.9,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145265312","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}