Biomedical Signal Processing and Control最新文献

{"title":"Parkinson’s disease detection from voice signals using adaptive frequency attribute topology","authors":"Tao Zhang ,&nbsp;Jing Tian ,&nbsp;Zaifa Xue ,&nbsp;Xiaonan Guo","doi":"10.1016/j.bspc.2025.107592","DOIUrl":"10.1016/j.bspc.2025.107592","url":null,"abstract":"<div><div>Dysphonia is one of the early symptoms of Parkinson’s disease (PD). The voice features extracted from voice signals can be effectively used for PD detection. Among them, the energy features extracted based on attribute topology have achieved preliminary results by combining time–frequency domain and structure representation. However, these energy features are susceptible to temporal fluctuation and are sensitive to weak coupling, which reduces their stability and representativeness. To solve these problems, this paper proposes the connected structural feature based on adaptive frequency attribute topology (CS-AFAT). Firstly, the energy statistic windows are established according to the distribution of spectral lines in the spectrogram for the aggregated statistics of energy information in time domain, which can effectively eliminate the influence of time disturbances. Secondly, the attribute topology is established to preform visualized statistics on the energy variation information of each point in the window. Finally, to reduce the influence of weak coupling in the topology, the adaptive threshold mechanism is designed to remove the edge with low coupling strength in the topology, so as to obtain more stable and representative features for PD classification. The results on two different language datasets show that the highest classification accuracy of CS-AFAT is 92.41% and 96.67%. The advantage of the proposed CS-AFAT is that it can obtain classification performance superior to or comparable to advanced energy features while having low-dimensional feature representation, which verifies the effectiveness and advancement of the CS-AFAT feature extracted by combining global energy information statistics and adaptive threshold mechanism.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107592"},"PeriodicalIF":4.9,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155328","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":"Holographic convolutional attention neural network for motor imagery decoding based on EEG temporal–spatial frequency features","authors":"Qingsong Ai ,&nbsp;Yuang Liu ,&nbsp;Quan Liu ,&nbsp;Li Ma ,&nbsp;Kun Chen","doi":"10.1016/j.bspc.2025.107526","DOIUrl":"10.1016/j.bspc.2025.107526","url":null,"abstract":"<div><div>As brain–computer interface (BCI) technology advances, numerous researchers begin to employ deep learning techniques for the interpretation and categorization of electroencephalography (EEG) signals. Motor imagery brain–computer interface (MI-BCI) is a kind of technology to realize human–computer interaction by decoding EEG signal, which has a wide application prospect in the fields of neural rehabilitation and auxiliary equipment control. However, due to complexity and dynamic characteristics of EEG signals and interference of background noise, it remains a significant challenge to achieve high-precision and robust dynamic imagery classification. Traditional feature extraction methods tend to focus only on time-domain features or frequency-domain features, and rarely combine the two to analyze together. Traditional attention mechanisms are limited in capturing complex spatial–temporal dependencies and consume large computational resources. To solve these problems, we propose a holographic convolutional attention neural network based on temporal–spatial frequency features for motor imagery (MI) classification. Specifically, a convolutional filter bank is constructed to distill the temporal–spatial features of EEG signals, and then a frequency domain feature module based on fast fourier transform has been introduced to capture the frequency domain features of EEG signals, and then average pooling and variance pooling are used to capture different multimodal information. Subsequently, holographic convolution attention is designed to obtain local features through convolution operations, and then features are weighted and aggregated by attention weights, combined with context vectors, which can adaptively modulate the influence of individual local features on the overarching information, aiding the model in capturing the distant interdependencies between temporal and spatial dimensions. To assess the feasibility of the suggested approach, an extensive series of experiments is conducted using two publicly accessible datasets. The findings indicate that the mean performance of the suggested model in the BCIC-IV-2a four-classification reaches 80.59%, which is excellent to some current methods. The average classification accuracy of BCIC-IV-2b binary classification reaches 87.69%, which proves the validity of our model.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107526"},"PeriodicalIF":4.9,"publicationDate":"2025-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143154917","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":"MRI intracranial Neoplasm classification using hybrid LOA-based deep learning classifier","authors":"J.Catherina Mary ,&nbsp;M. Suganthi","doi":"10.1016/j.bspc.2025.107560","DOIUrl":"10.1016/j.bspc.2025.107560","url":null,"abstract":"<div><div>The detection and segmentation of brain tumors in MRI images are essential for accurate diagnosis and effective treatment planning. Current methodologies often struggle to maintain accuracy while efficiently processing large datasets. This research integrates advanced techniques in image preprocessing, feature extraction, and classification to improve brain tumor detection and segmentation. This proposed research employs Adaptive Median-Gaussian Filtering to reduce noise while preserving edges and fine details. For segmentation, we combine Kernel Density Estimation (KDE) with Fuzzy C-Means (FCM) clustering (KDFCM) to utilize local density variations for more precise results. Feature extraction is conducted using Histogram of Oriented Gradients (HOG), which is optimized with Particle Swarm Optimization (PSO) for efficient and robust image analysis. The classification utilizes the ResNeSt architecture with split-attention mechanisms, and the Lion Optimization Algorithm (LOA) fine-tunes the model parameters. The simulations were conducted using MATLAB 2021a. Experimental results demonstrate the effectiveness of the proposed methods, achieving high accuracy, precision, recall, and F-measure across various datasets, including the X-ray image dataset, FigShare and BTC Magnetic Resonance Imaging (MRI) datasets. The proposed algorithm achieved 99.41 % accuracy, 99.57 % precision, 99.12 % recall, and 99.52 % F-measure for FigShare Data, and 99.72 % accuracy, 99.85 % precision, 99.25 % recall, and 99.67 % F-measure for BTC Data, outperforming existing DL technology. This research highlights the potential of these methods for reliable and efficient brain tumor detection.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107560"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155293","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel CNN-ViT-based deep learning model for early skin cancer diagnosis","authors":"Ishak Pacal ,&nbsp;Burhanettin Ozdemir ,&nbsp;Javanshir Zeynalov ,&nbsp;Huseyn Gasimov ,&nbsp;Nurettin Pacal","doi":"10.1016/j.bspc.2025.107627","DOIUrl":"10.1016/j.bspc.2025.107627","url":null,"abstract":"<div><div>Skin cancer is a serious global health issue where early detection is crucial for effective treatment and improved patient outcomes. However, accurate diagnosis is challenging due to the variety of subtypes and imaging complexities. This study introduces an innovative deep learning model based on the MetaFormer architecture, optimized specifically for skin cancer. The Proposed Model features a hybrid design that replaces traditional self-attention methods with novel focal self-attention mechanisms, enhancing its ability to identify critical regions, reduce noise, and extract features more effectively, ultimately boosting diagnostic accuracy. To evaluate the model’s generalization capabilities, it was tested on two benchmark datasets: ISIC 2019, which includes a diverse set of dermatological images across eight skin cancer classes, and HAM10000, widely used in dermatological research. The model achieved outstanding results, including an accuracy of 0.9254, precision of 0.9041, recall of 0.8768, and an F1-score of 0.8886 on ISIC 2019, and an accuracy of 0.9501, precision of 0.9470, recall of 0.9211, and an F1-score of 0.9334 on HAM10000. The Proposed Model surpasses existing methods in the field, outperforming ten advanced CNN models and twenty state-of-the-art ViT models under the same training and evaluation conditions. With a lightweight design of just 35.01 million parameters, it is optimized for real-time and mobile applications, making it highly practical for clinical use. Its reliable performance ensures accurate diagnoses, which are essential for early intervention and treatment, addressing a critical need in modern healthcare.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107627"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155330","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":"FNBUI-NET: A multi-task model for fetal nasal bone ultrasound image defect detection and classification","authors":"Yapeng Li ,&nbsp;Zhonghua Liu ,&nbsp;Jiansong Zhang ,&nbsp;Pan Zeng ,&nbsp;Yuling Fan ,&nbsp;Longxiang Feng ,&nbsp;Peizhong Liu ,&nbsp;Guorong Lyu ,&nbsp;Xiuming Wu","doi":"10.1016/j.bspc.2025.107586","DOIUrl":"10.1016/j.bspc.2025.107586","url":null,"abstract":"<div><div>In prenatal ultrasound diagnosis, the accurate identification and timely assessment of the fetal nasal bone are crucial for monitoring the growth and development of the fetus’s face. However, the small size and complex anatomy of the fetal nasal bone make it difficult for doctors to determine if there are any defects. Motivated by this critical and intricate ultrasound task, a multi-task deep learning model, FNBUI Network (FNBUI-NET), was designed to detect fetal nasal bone defects and classify fetal nasal bone ultrasound images (FNBUI) in real-time. The model reconsiders and balances the relationship between inference speed and computational accuracy in object detection algorithms, innovatively proposing the integration of the ShuffleNetV2 (Shuffle_Block) network with the Content-Aware ReAssembly of Features (CARAFE) module to process data tensors of varying sizes from clinical settings, and employed Coordinate Attention (CA) for feature compression encoding to extract consistency constraints within chaotic feature distributions. The FNBUI dataset is also proposed to support the research on fetal nasal bone detection and classification, which 1104 ultrasound images with three objects (i.e., normal, dysplasia, absence) of fetal nasal bones from 731 mid-pregnancy women for training and testing the performance of FNBUI-NET. The experiments show that FNBUI-NET can accurately classify fetal ultrasound images and detect nasal bone defects, achieving an [email protected] of 0.985 in nasal bone detection and an accuracy of 0.981 in classification, outperforming 18 popular current methods. It provides a valuable tool for less experienced and younger sonologists, assisting them in the examination and diagnosis of fetal nasal bone defects.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107586"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155327","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":"ACANet: Adaptive Contour Aware Nucleus Segmentation Network","authors":"Yulin Chen,&nbsp;Qian Huang,&nbsp;Zhijian Wang,&nbsp;Meng Geng","doi":"10.1016/j.bspc.2025.107575","DOIUrl":"10.1016/j.bspc.2025.107575","url":null,"abstract":"<div><div>Accurate segmentation and analysis of nucleus are crucial for cancer diagnosis and prognosis. For nuclear images with blurred boundaries and low color contrast, existing methods often rely on additional contour-prediction branches and final feature fusion modules. However, these approaches encounter challenges such as the loss of contour information during the downsampling process and the introduction of redundant information and noisy artifacts during feature fusion. To overcome these limitations, we propose an Adaptive Contour Aware Nucleus Segmentation Network (ACANet), which integrates Sobel convolution and Fourier transform to enhance the contours of the nuclei. Additionally, a novel multi-scale adaptive feature learning module is employed to extract multi-scale discriminative features of nuclei. Finally, a loss function guided by Grad-CAM is proposed to achieve better segmentation performance and enhance model interpretability. Experimental results demonstrate that the proposed architecture achieves superior accuracy in nucleus segmentation tasks.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107575"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155334","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":"Dual consistency semi-supervised learning for 3D medical image segmentation","authors":"Lin Wei ,&nbsp;Runxuan Sha ,&nbsp;Yucheng Shi ,&nbsp;Qingxian Wang ,&nbsp;Lei Shi ,&nbsp;Yufei Gao","doi":"10.1016/j.bspc.2025.107568","DOIUrl":"10.1016/j.bspc.2025.107568","url":null,"abstract":"<div><div>Semi-supervised learning (SSL) methods can utilize a large amount of unlabeled data to alleviate the problem of scarce labeled data. Consistency regularization as a classic SSL strategy, has achieved significant results in this field. However, due to the low contrast of organ contours in medical images, existing approaches are prone to uncertain predictions at the segmentation boundaries and the hard-to-identify blurry areas. To address these issues, we propose a dual consistency semi-supervised medical image segmentation network (DC-Net). Specifically, it can reduce the prediction uncertainty to obtain low-entropy decision boundaries by performing consistency prediction under model-level and task-level perturbations. We first design a cross-consistency loss between the segmentation map and the pseudo-labels across different models, aiming to encourage the models to maintain consistent predictions in blurry areas. Then, consistency constraints are applied between the segmentation task and the geometric perception task to construct the geometric contours of the target, thereby obtaining more precise boundary distance information. The experiments on four public medical image datasets (including LA, Pancreas CT, ACDC and PROMISE2012) demonstrate that DC-Net has achieved the state-of-the-art performance over other advanced semi-supervised methods.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107568"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155342","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":"Spatio-temporal graph Bert network for EEG emotion recognition","authors":"Jingjie Yan ,&nbsp;Chengkun Du ,&nbsp;Na Li ,&nbsp;Xiaoyang Zhou ,&nbsp;Ying Liu ,&nbsp;Jinsheng Wei ,&nbsp;Yuan Yang","doi":"10.1016/j.bspc.2025.107576","DOIUrl":"10.1016/j.bspc.2025.107576","url":null,"abstract":"<div><div>EEG data presents a topological structure in spatial. In order to effectively capture the temporal and spatial characteristics of brain data, this paper proposes a Spatio-temporal Graph Bert network (STGB) and applies it to the emotion recognition of Electroencephalogram (EEG) signals. The STGB network learns the EEG features from the spatial and the temporal domains respectively. In the spatial domain, the adjacency matrix is constructed to model the graph of EEG signals, and then the spatial domain features of EEG signals are extracted by using the Graph Bert network through the steps of subgraph partitioning, node embedding, node feature updating based on attention mechanism and node clustering. In the temporal domain, the spatial domain features of EEG signals obtained from each period are connected by the Long Short-Term Memory network (LSTM) to learn the temporal correlation of the EEG signals, so as to complete the EEG emotion recognition task. Experiments on SEED dataset and DEAP dataset prove that STGB can complete the learning of EEG features more comprehensively and accurately, and achieve a higher emotion recognition rate.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107576"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143154717","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":"Extended Spatiotemporal Separation for noise reduction in magnetocardiography","authors":"Yuheng Zhou ,&nbsp;Yifan Jia ,&nbsp;Jiaqi Liang ,&nbsp;Dong Xu ,&nbsp;Min Xiang","doi":"10.1016/j.bspc.2025.107534","DOIUrl":"10.1016/j.bspc.2025.107534","url":null,"abstract":"<div><div>Magnetocardiography (MCG) is a fast and noninvasive method for diagnosing heart diseases but is very susceptible to external noise. MCG using optically pumped magnetometers (OPMs) has the advantages of not needing cooling and high time resolution. The Spatiotemporal Signal Space Separation method can effectively suppress noise outside the heart in the MCG. However, the number of MCG sensors is relatively small, and sensors are usually arranged flat, which limits the use of spatiotemporal signal space separation in MCG. Therefore, this paper proposed a new method called Extended Spatiotemporal Separation (ESTS) based on Signal Space Separation (SSS), which is suitable for smaller and flat-arranged channels. Besides, we proposed a real-time automatic ESTS method to remove destructive huge noise while ensuring less distortion. In addition, we conducted simulations and experimental verifications to prove the effectiveness of ESTS in MCG noise reduction. We compared it with ordinary SSS and standard denoising methods like Independent Component Analysis (ICA). The results show that when faced with a huge noise for MCG with few sensor channels, ESTS has a higher signal-to-noise ratio (SNR) and can better restore the waveforms. As a result, this paper provided new ideas for clinical real-time noise reduction of MCG and unshielded MCG measurement.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107534"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155290","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":"Integrating Gaussian mixture model with adjacent spatial adaptive transformer multi-stage network for magnetic resonance image reconstruction","authors":"Tong Hou ,&nbsp;Hongqing Zhu ,&nbsp;Jiahao Liu ,&nbsp;Ning Chen ,&nbsp;Jiawei Yan ,&nbsp;Bingcang Huang ,&nbsp;Weiping Lu ,&nbsp;Suyi Yang ,&nbsp;Ying Wang","doi":"10.1016/j.bspc.2025.107577","DOIUrl":"10.1016/j.bspc.2025.107577","url":null,"abstract":"<div><div>Accurate and fast magnetic resonance imaging (MRI) reconstruction using undersampled data is crucial for practical applications. This paper proposes a novel multi-stage network, termed IGT-Net, which integrates the Gaussian mixture model (GMM) with an adjacent spatial adaptive transformer (ASAT). Specifically, IGT-Net consists of a compressive sensing sampling initialization module (CSS-IM) and a reconstruction group. The CSS-IM is designed to equivalently mimic the sampling process and transmit MRI images of different sampling rates to the reconstruction group. The reconstruction group consists of the proposed transformer-based iterative shrinkage threshold algorithm (TrISTA) and transformer-based Gaussian mixture model (TrGMM), which reconstruct clear MRI images more efficiently. In TrISTA, the dynamic gradient descent module (DGDM) achieves fast and stable convergence of results throughout the iteration process. The proposed transformer-based proximal mapping module (TPMM) utilizes a designed transformer to fully integrate adjacent spatial features for MRI reconstruction, thereby avoiding unclear results that may arise from neglecting adjacent stage information. Meanwhile, TrGMM incorporates both GMM and ASAT to leverage maximum likelihood estimation for weight updates and artifact removal. Additionally, a plug-and-play ASAT is designed to effectively extract adjacent spatial features. Experiments conducted on two public datasets demonstrate that IGT-Net outperforms state-of-the-art methods and significantly improves the quality of reconstructed images.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"104 ","pages":"Article 107577"},"PeriodicalIF":4.9,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143155326","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}