Biomedical Signal Processing and Control最新文献

{"title":"Parkinson’s disease classification and prediction via adaptive sparse learning from multiple modalities","authors":"","doi":"10.1016/j.bspc.2024.107061","DOIUrl":"10.1016/j.bspc.2024.107061","url":null,"abstract":"<div><div>Parkinson’s disease (PD) is a prevalent neurodegenerative disorder where early clinical diagnosis is critical for patients. In this study, we introduce a new embedded adaptive sparse learning method that integrates multimodal data, labeling, and clinical score information for early PD diagnosis. Based on the manifold learning theory, the proposed method utilizes a low-dimensional manifold to represent the structure of high-dimensional data. We carry out similarity learning and feature selection from different modalities while using the <span><math><mrow><msub><mi>l</mi><mrow><mn>2</mn><mo>,</mo><mi>p</mi></mrow></msub></mrow></math></span> specification for adaptive sparsity control. Also, we dynamically learn the similarity between features to find more effective information in features. an effective optimization iterative algorithm is proposed to solve this problem. To validate the effectiveness of the proposed method, we conducted extensive experiments on the Parkinson’s Progression Markers Initiative (PPMI) database, including PD vs. normal controls (NC), scans without evidence of dopaminergic deficit (SWEDD) vs. NC, PD vs. SWEDD, and PD vs. SWEDD vs. NC. Using baseline data for experiments, our method achieved accuracies of 82.38%, 85.56%, 86.47%, and 65.54%, respectively. When using 12-month data, the accuracies were 77.43%, 95.26%, 95.35%, and 77.98%, respectively. Overall, our method outperformed the other methods. Additionally, the performance of the three deep models using the feature subsets selected by our method surpassed that achieved using the original data, further validating the effectiveness of our method.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142433457","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":"SDLU-Net: A similarity-based dynamic linking network for the automated segmentation of abdominal aorta aneurysms and branching vessels","authors":"","doi":"10.1016/j.bspc.2024.106991","DOIUrl":"10.1016/j.bspc.2024.106991","url":null,"abstract":"<div><div>Accurate preoperative measurement of abdominal aortic aneurysms (AAAs) and associated vascular structures is indispensable for surgical planning. However, the precise extraction of complex multi-branch vascular structures from CTA images presents significant challenges. These challenges stem from the high individual variation in vascular structure, the difficulty of separating important organ branches from the surrounding tissue area, and the unclear boundaries between main branches and organ branches. To overcome these obstacles, the SDLU-Net network architecture is introduced. Firstly, a novel spatial positional encoding structure, named “dual position encoding”, is presented to preserve the spatial relationships of vascular anatomical structures throughout the network operations. Secondly, a similarity-based dynamic attention linking module is proposed to effectively distinguish vessels from tissues, and the main branches from other branches. Furthermore, dual skip connections and channel attention modules are integrated into the decoding layer to enhance the information flow within the network. Lastly, a hybrid loss function was introduced to address the issue of sample imbalance, thereby improving the segmentation results of organ branches. Extensive experiments have demonstrated that SDLU-Net shows excellent segmentation performance in multi-branch blood vessels of the abdominal aorta, highlighting its substantial clinical potential for surgical planning involving complex multi-branch structures.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142433456","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":"SMANet: Superpixel-guided multi-scale attention network for medical image segmentation","authors":"","doi":"10.1016/j.bspc.2024.107062","DOIUrl":"10.1016/j.bspc.2024.107062","url":null,"abstract":"<div><div>Medical image segmentation plays a crucial role in assisting diagnosis. However, the inherent low contrast and noise in medical images make it challenging to achieve accurate medical image segmentation. To address this problem, we propose a superpixel-guided multi-scale attention network (SMANet) for segmenting medical images accurately. Superpixel segmentation could effectively divide medical images into different regions based on image gradient information. Accordingly, a superpixel-guided fusion attention module is proposed to utilize the regional division information provided by superpixel segmentation and further optimize the features in spatial and channel dimensions. In the encoder stage, an inverted pyramid feature extraction architecture is constructed to take advantage of texture information in shallow features, effectively solving the problem of information loss caused by sampling. In the proposed multi-scale feature joint decoder, multi-scale features are effectively enhanced and integrated to reconstruct image details guided by high-level features. Specifically, the full-scale feature attention module is embedded into multi-scale skip connections to contribute to the sufficient expression of important semantic information in features. Besides, we redesign the classic decoder to make full use of to the semantic information of deep features to guide feature fusion. Extensive experiments based on different public datasets and proposed neck vessels ultrasound dataset (USdata) prove the superiority of SMANet in terms of generalization, qualitative and quantitative performance.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142433455","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":"Robust multi-modal COVID-19 medical image registration using dense deep learning descriptor model","authors":"","doi":"10.1016/j.bspc.2024.107007","DOIUrl":"10.1016/j.bspc.2024.107007","url":null,"abstract":"<div><div>In medical image processing, multi-modal medical image registration is a challenging task due to the varied image characteristics. Because of the Non-functional strength relation and the erratic intricate deformations among images. To overcome these issues, this paper proposed an enhanced residual dense learning data descriptor for multi-modal COVID-19 image registration. In this work, input images are taken from the COVID-19 X-ray and CT Chest Images Dataset. Initially, the input images are pre-processed using the boosted switching bilateral filter (BSBF), in which the best median value is examined using a Sorted Quadrant Median Vector (SQMV). Then, the Directed Edge Enhancer (DEE) algorithm is used for the edge enhancement process. These pre-processed images are provided as the input of a deep learning based multi-scale feature extraction module to diminish the mutual interference of features and make it easier to train the network model. Data Adaptive Descriptor (DAD) is provided for structural representation, and the self-similarity metrics of the reference and floating images are examined by the Sum of Squared Differences (SSD). The goal function for image registration is made to the final deformation field based on SSD. Here, the simulation is performed by using a Python tool. The accuracy value of the proposed method in the COVID-19 X-ray and CT Chest images dataset is 96%, and the MSE value is 0.03%. Compared with other existing methods, our proposed method produces better performance. The proposed model is more efficient by using the hybrid deep learning methodology.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142434136","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":"SMRFnet: Saliency multi-scale residual fusion network for grayscale and pseudo color medical image fusion","authors":"","doi":"10.1016/j.bspc.2024.107050","DOIUrl":"10.1016/j.bspc.2024.107050","url":null,"abstract":"<div><div>Currently, multimodal medical images are widely used in the medical field, such as surgical planning, remote guidance, and medical teaching. However, the information of single-modal medical images is limited, making it difficult for doctors to obtain information from multiple perspectives and gain a more comprehensive understanding of the patient’s condition. To overcome this difficulty, many multimodal medical image fusion algorithms have been proposed. However, existing fusion algorithms have drawbacks such as weak edge strength, detail loss or color distortion. To overcome these shortcomings, a saliency multi-scale residual fusion network (SMRFnet) is proposed and applied to the fusion of grayscale and pseudo color medical images. Firstly, MRSFnet extracts saliency features through the VGG network. Then, the saliency features are added together to obtain the fusion features. Finally, the fusion features are fed into a multi-scale residual network to decode into the fusion image. The experiment shows that the proposed algorithm preserves more important saliency information and details in the fusion images compared to the reference algorithms. In addition, the proposed algorithm has more details and objective indicators.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142434435","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 multimodal emotion classification method considering micro-expression information and simulating human visual attention mechanism","authors":"","doi":"10.1016/j.bspc.2024.107036","DOIUrl":"10.1016/j.bspc.2024.107036","url":null,"abstract":"<div><div>Micro-expressions, as fleeting and involuntary facial expressions, genuinely reflect an individual’s emotional state. However, the characteristics of micro-expressions, primarily manifested in subtle movements of specific facial regions, affect the accuracy of recognition. Moreover, micro-expression recognition methods relying solely on visual information are limited. To address these issues, this paper proposes a Periphery Attention Fusion Network (PAFN) for micro-expression recognition. The method integrates facial expressions, electroencephalogram (EEG) time series, and spatial sequence information, aiming to enhance the accuracy and reliability of micro-expression recognition through multimodal information fusion. PAFN consists of three key modules: the Three-Dimensional Construction Module (3DCM) for constructing the three-dimensional features of EEG signals; the Preliminary Convolutional Preprocessing Module (PCP) applying dynamic convolution and depthwise separable convolution techniques for preliminary extraction of facial features; and the Periphery Self-Attention Memory Module (PSAM) combined with Long Short-Term Memory (LSTM) networks, generating peripheral positional encoding and adjusting attention weights to simulate the human visual attention mechanism, reducing interference and focusing on key information. Experimental results on the DEAP dataset and the micro-expression PEG dataset demonstrate that the PAFN method’s recognition accuracy exceeds 96%, outperforming existing methods, confirming its efficiency and advancement in the field of micro-expression recognition.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142434135","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":"Involution-based HarmonyNet: An efficient hyperspectral imaging model for automatic detection of neonatal health status","authors":"","doi":"10.1016/j.bspc.2024.106982","DOIUrl":"10.1016/j.bspc.2024.106982","url":null,"abstract":"<div><h3>Background and Objective</h3><div>Neonatal health is critical for early infant care, where accurate and timely diagnoses are essential for effective intervention. Traditional methods such as physical exams and laboratory tests may lack the precision required for early detection. Hyperspectral imaging (HSI) provides non-invasive, detailed analysis across multiple wavelengths, making it a promising tool for neonatal diagnostics. This study introduces HarmonyNet, an involution-based HSI model designed to improve the accuracy and efficiency of classifying neonatal health conditions.</div></div><div><h3>Methods</h3><div>Data from 220 neonates were collected at the Neonatal Intensive Care Unit of Selçuk University, comprising 110 healthy infants and 110 diagnosed with conditions such as respiratory distress syndrome (RDS), pneumothorax (PTX), and coarctation of the aorta (AORT). The HarmonyNet model incorporates involution kernels and residual blocks to enhance feature extraction. The model’s performance was evaluated using metrics such as overall accuracy, precision, recall, and area under the curve (AUC). Ablation studies were conducted to optimize hyperparameters and network architecture.</div></div><div><h3>Results</h3><div>HarmonyNet achieved an AUC of 98.99%, with overall accuracy, precision and recall rates of 90.91%, outperforming existing convolution-based models. Its low parameter count and computational efficiency proved particularly advantageous in low-data scenarios. Ablation studies further demonstrated the importance of involution layers and residual blocks in improving classification accuracy.</div></div><div><h3>Conclusions</h3><div>HarmonyNet represents a significant advancement in neonatal diagnostics, offering high accuracy with computational efficiency. Its non-invasive nature can contribute to improved health outcomes and more efficient medical interventions. Future research should focus on expanding the dataset and exploring the model’s potential in multi-class classification tasks.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142434137","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":"ZFNet and deep Maxout network based cancer prediction using gene expression data","authors":"","doi":"10.1016/j.bspc.2024.107038","DOIUrl":"10.1016/j.bspc.2024.107038","url":null,"abstract":"<div><div>The earlier predictions of cancer types are highly essential. Currently, gene expression data (GED) is employed for effective and earlier diagnosing of cancer. The GED allows the models to effectively learn the problem and acts as the most efficient strategy for extracting relevant and new data. Various researchers have implemented many techniques for cancer prediction but the accuracy is required to be improved for the early prediction. Here, the Zeiler and Fergus network fused Deep Maxout Network (ZF-maxout Net) is designed for cancer prediction using GED. At first, input GED data is taken from a certain dataset. Then, data transformation is performed in input GED utilizing Box-Cox transformation. After that, feature fusion (FF) is conducted using the Deep Neural Network (DNN) with Kulczynski similarity. Finally, the cancer prediction is done by utilizing ZF-maxout Net. Moreover, ZF-maxout Net is an amalgamation of the Zeiler and Fergus network (ZFNet) and Deep Maxout Network (DMN). In addition, ZF-mCanceraxout Net obtained a maximal accuracy of 92.7 %, a True Positive Rate (TPR) of 95.8 % and a minimal False Negative Rate (FNR) of 49.9 %.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142434436","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":"Fusing temporal-frequency information with Contrast Learning on Graph Convolution Network to decoding EEG","authors":"","doi":"10.1016/j.bspc.2024.106986","DOIUrl":"10.1016/j.bspc.2024.106986","url":null,"abstract":"<div><div>Decoding human electroencephalogram (EEG) signals and identifying the brain’s response to external stimuli is a challenging task, but it is crucial for understanding the brain’s information processing mechanisms and developing brain like intelligent computers. Previous studies have used neural networks to analyze the spatiotemporal features of various EEG regions for EEG emotion recognition, but there have been few studies characterizing different frequency bands of EEG signals, and little attention has been paid to the issue of fuzzy emotion labeling in continuous emotion models. Based on these two issues, this study proposes a Graph Convolutional Network (FCLGCN) method to collect time and frequency band information of EEG, and solves the problem of fuzzy emotional boundaries through contrastive learning. FCLGCN achieved high recognition accuracy on DEAP and SEED datasets. According to the experimental results, the connection between the frontal and temporal lobes on both sides of the brain becomes tighter during emotional changes.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142428725","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":"AUDSER: Auto-detect and self-recovery reversible steganography algorithm for biological signals","authors":"","doi":"10.1016/j.bspc.2024.106974","DOIUrl":"10.1016/j.bspc.2024.106974","url":null,"abstract":"<div><h3>Objective</h3><div>As stated in guidelines of the Health Insurance Portability and Accountability Act, 1996, protecting medical data is important in this digital age. In this paper, a new AUDSER: Auto-detect and Self-recovery steganography algorithm is presented tested on single-channel biological signals.</div></div><div><h3>Methods</h3><div>The raw signal (α) was first transformed (β) into Legendre domain (LD) using the Legendre-transformation, and then, secret bit embedding was done using a variable-key (VK), generating stego-data in LD (<span><math><mover><mi>β</mi><mo>∼</mo></mover></math></span>), followed by inverse-Legendre-transformation, generating stego-data in time-domain (<span><math><mover><mi>α</mi><mo>∼</mo></mover></math></span>). At this stage, a Hash-data, was computed using <span><math><mover><mi>β</mi><mo>∼</mo></mover></math></span> and <span><math><mover><mi>α</mi><mo>∼</mo></mover></math></span>. The subsequent sample-steganography (SS) utilized VK, which was continuously updated, making it very challenging to compromise the security chain. Finally, another information, τ, pertaining to <span><math><mover><mi>β</mi><mo>∼</mo></mover></math></span> and <span><math><mover><mi>α</mi><mo>∼</mo></mover></math></span> was updated to <span><math><mover><mi>α</mi><mo>∼</mo></mover></math></span>, to generate final stego-data,<span><math><mover><mi>α</mi><mo>̂</mo></mover></math></span>. Violation of τ indicated occurrence of error during extraction process and a self-recovery architecture for n^th erroneous sample was initiated using a particle swarm optimization, Hash data and VK of (n-1)^th SS.</div></div><div><h3>Results</h3><div>The extraction algorithm was exact reverse algorithm of encryption, and the reconstruction error was found to be reduced after extraction. As a consequence, this suggested algorithm results in partially-reversible steganography and produced a minimum percent-root-mean-square-difference of 0.31% and 1%, respectively, for ECG and PPG signals, available in Physionet.</div></div><div><h3>Conclusion</h3><div>This algorithm is capable of inserting maximum 3 bits/sample and has the ability of auto-detection and self-correction of manipulation in stego-data, without human intervention.</div></div><div><h3>Significance</h3><div>This algorithm can be implemented in any time-series data irrespective of signal features resulting in very low distortion error.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":null,"pages":null},"PeriodicalIF":4.9,"publicationDate":"2024-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142428673","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}