Biomedical Physics & Engineering Express最新文献

{"title":"TFDISNet: Temporal-frequency domain-invariant and domain-specific feature learning network for enhanced auditory attention decoding from EEG signals.","authors":"Zhongcai He, Yongxiong Wang","doi":"10.1088/2057-1976/ae09b2","DOIUrl":"10.1088/2057-1976/ae09b2","url":null,"abstract":"<p><p>Auditory Attention Decoding (AAD) from Electroencephalogram (EEG) signals presents a significant challenge in brain-computer interface (BCI) research due to the intricate nature of neural patterns. Existing approaches often fail to effectively integrate temporal and frequency domain information, resulting in constrained classification accuracy and robustness. To address these shortcomings, a novel framework, termed the Temporal-Frequency Domain-Invariant and Domain-Specific Feature Learning Network (TFDISNet), is proposed to enhance AAD performance. A dual-branch architecture is utilized to independently extract features from the temporal and frequency domains, which are subsequently fused through an advanced integration strategy. Within the fusion module, shared features, common across both domains, are aligned by minimizing a similarity loss, while domain-specific features, essential for the task, are preserved through the application of a dissimilarity loss. Additionally, a reconstruction loss is employed to ensure that the fused features accurately represent the original signal. These fused features are then subjected to classification, effectively capturing both shared and unique characteristics to improve the robustness and accuracy of AAD. Experimental results show TFDISNet outperforms state-of-the-art models, achieving 97.1% accuracy on the KUL dataset and 88.2% on the DTU dataset with a 2 s window, validated across group, subject-specific, and cross-subject analyses. Component studies confirm that integrating temporal and frequency features boosts performance, with the full TFDISNet surpassing its variants. Its dual-branch design and advanced loss functions establish a robust EEG-based AAD framework, setting a new field standard.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145124100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"EigenU-Net: integrating eigenvalue decomposition of the Hessian into U-Net for 3D coronary artery segmentation.","authors":"Cathy Ong Ly, Chris McIntosh","doi":"10.1088/2057-1976/ae08bb","DOIUrl":"10.1088/2057-1976/ae08bb","url":null,"abstract":"<p><p><i>Objective</i>. Coronary artery segmentation is critical in medical imaging for the diagnosis and treatment of cardiovascular disease. However, manual segmentation of the coronary arteries is time-consuming and requires a high level of training and expertise.<i>Approach</i>. Our model, EigenU-Net, presents a novel approach to coronary artery segmentation of cardiac computed tomography angiography (CCTA) images that seeks to directly embed the geometrical properties of tubular structures, i.e. arteries, into the model. To examine the local structure of objects in the image we have integrated a closed-form solution of the eigenvalues of the Hessian matrix of each voxel for input into an U-Net based architecture.<i>Main results</i>. We demonstrate the feasibility and potential of our approach on the public IMAGECAS dataset consisting of 1000 CCTAs. The best performing model at 87% centerline Dice was EigenU-Net with Gaussian pre-filtering of the images.<i>Significance</i>. We were able to directly integrate the calculation of eigenvalues into our model EigenU-Net, to capture more information about the structure of the coronary vessels. EigenU-Net was able to segment regions that were overlooked by other models.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145085123","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning based classification of imagined speech electroencephalogram data from the amplitude and phase spectrum of frequency domain EEG signal.","authors":"Meenakshi Bisla, Radhey Shyam Anand","doi":"10.1088/2057-1976/ae04ee","DOIUrl":"10.1088/2057-1976/ae04ee","url":null,"abstract":"<p><p>Imagined speech classification involves decoding brain signals to recognize verbalized thoughts or intentions without actual speech production. This technology has significant implications for individuals with speech impairments, offering a means to communicate through neural signals. The prime objective of this work is to propose an innovative machine learning (ML) based classification methodology that combines electroencephalogram (EEG) data augmentation using a sliding window technique with statistical feature extraction from the amplitude and phase spectrum of frequency domain EEG segments. This work uses an EEG dataset recorded from a 64-channel device during the imagination of long words, short words, and vowels with 15 human subjects. First, the raw EEG data is filtered between 1 Hz and 100 Hz, then segmented using a sliding window-based data augmentation technique with a window size of 100 and 50% overlap. The Fourier Transform is applied to each windowed segment to compute the amplitude and phase spectrum of the signal at each frequency point. The next step is to extract 50 statistical features from the amplitude and phase spectrum of frequency domain segments. Out of these, the 25 most statistically significant features are selected by applying the Kruskal-Walli's test. The extracted feature vectors are classified using six different machine learning based classifiers named support vector machine (SVM), K nearest neighbor (KNN), Random Forest (RF), XGBoost, LightGBM, and CatBoost. The CatBoost classifier outperforms other machine learning classifiers by achieving the highest accuracy of 91.72 ± 1.52% for long words classification, 91.68 ± 1.54% for long versus short word classification, 88.05 ± 3.07% for short word classification, and 88.89 ± 1.97% for vowel classification. The performance of the proposed model is assessed using five performance evaluation metrics: accuracy, F1-score, precision, recall, and Cohen's kappa. Compared to the existing literature, this study has achieved a 5%-7% improvement with the CatBoost classifier and extracted feature matrix.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145028897","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quantitative Photo-Acoustic Imaging based on data normalisation : application to the reconstruction of the opto-mechanical properties of the intervertebral disc.","authors":"Antoine Capart, Roman Allais, Julien Wojak, Olivier Boiron, Anabela Da Silva","doi":"10.1088/2057-1976/ae0b76","DOIUrl":"https://doi.org/10.1088/2057-1976/ae0b76","url":null,"abstract":"<p><p>The inverse problem in quantitative photoacoustic imaging (QPAI), particularly in optical inversion, presents significant challenges for accurate image reconstruction due to the nonlinearity of photoacoustic signal. In this study, we introduce a novel reconstruction strategy specifically tailored for imaging the intervertebral disc (IVD), a biphasic tissue primarily composed of water and collagen. This work offers two key contributions: (1) the development of a new model-based optical inversion method that leverages the ratio of multi-wavelength photoacoustic measurements to define the cost function-circumventing the need for prior knowledge of the Grüneisen parameter, which is typically difficult to estimate; and (2) a first demonstration of QPAI's potential to non-invasively probe the biochemical composition of the IVD, particularly its water-to-collagen ratio. &#xD; A non-linear model-based reconstruction approach was implemented, using the adjoint variables method to express gradients, and Monte Carlo simulations to solve the Radiative Transfer Equation for both the forward and adjoint problems. Several scenarios were investigated, and a method was proposed that does not rely on prior knowledge of the Grüneisen coefficient. A numerical study was conducted using a digitised pig's disc and opto-mechanical properties sourced from the literature, under measurement configurations compatible with experimental setups. The results show that the derived ratio-based cost function in the model-based optical inversion scheme significantly enhances the reconstructions quality, particularly when combined with appropriate regularisation, thereby validating the feasibility and robustness of the approach. Tridimensional reconstructions of the two chromophores content in the IVD, performed under the highly restricted configuration of a single illumination, were obtained with less than 5% error, along with the reconstruction of the Grüneisen coefficient with less than 15% error, up to 2 cm in depth (noiseless synthetic data) with the proposed method, with only three different wavelength measurements required.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145147561","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated quantification and feature extraction of nuclei in diffuse large B-cell lymphoma using advanced imaging techniques.","authors":"Chee Chin Lim, Gei Ki Tang, Faezahtul Arbaeyah Hussain, Qi Wei Oung, Aidy Irman Yajid, Yen Fook Chong, Siti Nur Syahirah Mohamad Redwan","doi":"10.1088/2057-1976/ae06ab","DOIUrl":"10.1088/2057-1976/ae06ab","url":null,"abstract":"<p><p>Diffuse Large B-Cell Lymphoma is a subtype of non-Hodgkin lymphoma that occurs worldwide; around 80 thousand new cases were recorded in 2022. The accurate diagnosis and subtyping of non-Hodgkin lymphoma pose significant challenges that necessitate expertise, extensive experience, and meticulous morphological analysis. To address these challenges, a study was developed to segment and classify the slides of diffuse large B-cell lymphoma. This study utilizes a dataset consisting of 108 images of H&E-stained slides, including MYC-positive, MYC-negative, and normal slides. The images are initially pre-processed using colour deconvolution to highlight haematoxylin-stained nuclei before they are segmented to detect a nuclear area using the watershed algorithm and extract a morphological feature of nuclei. Results indicate that area showed a significant difference (P < 0.05), highlighting variations in nuclear size among groups. In contrast, perimeter, diameter, and circularity showed no significant differences. Colour analysis revealed significant differences in the standard deviation of the L component in LAB space and all RGB standard deviations, while the mean LAB and RGB values showed no significant differences.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145069005","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identifying EEG-Based Neurobehavioral Risk Markers of Gaming Addiction Using Machine Learning and Iowa Gambling Task.","authors":"Denis Kornev, Roozbeh Sadeghian, Amir Gandjbakhche, Julie Giustiniani, Siamak Aram","doi":"10.1088/2057-1976/ae0b75","DOIUrl":"https://doi.org/10.1088/2057-1976/ae0b75","url":null,"abstract":"<p><p>Internet Gaming Disorder (IGD), Gaming Disorder (GD), and Internet Addiction represent behavioral patterns with significant psychological and neurological consequences. Affected individuals often disengage from routine activities and exhibit distress upon interruption of gaming, impacting family life and overall well-being. Timely and objective detection methods are essential. This study investigates EEG-based biomarkers in healthy participants, aiming to classify them into two groups based on behavioral patterns observed during the Iowa Gambling Task (IGT). EEG and IGT data were collected simultaneously, with IGT serving as a cognitive challenge to induce decision-making under uncertainty and risk. EEG signals were segmented into event-related potentials (ERPs), pre-processed, and used to extract temporal features. Advanced signal transformation techniques, including Fast Fourier Transform (FFT), Power Spectral Density (PSD), Autocorrelation Function (ACF), and Wavelet Transforms, were employed to build the feature space. Machine Learning (ML) and Deep Learning (DL) classifiers, particularly Random Forest (RF) and Convolutional Neural Networks (CNN), were trained and validated, achieving a classification accuracy of 93%. This approach offers early detection of abnormal decision-making behavior and distinguishes participants based on neurophysiological responses, enhancing diagnostic speed and objectivity. The study emphasizes methodological transparency, ethical compliance, and data availability, providing a replicable framework for future investigations into behavioral biomarkers of gaming-related disorders.&#xD.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145147565","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DCDSN: dual-color domain siamese network for multi-classification of pathological artifacts.","authors":"Wei-Long Ding, Jin-Long Liu, Wei Zhu, Li-Feng Xu, Chun-Nian Wang","doi":"10.1088/2057-1976/ae0592","DOIUrl":"https://doi.org/10.1088/2057-1976/ae0592","url":null,"abstract":"<p><p>Pathological images are prone to artifacts during scanning and preparation, which can compromise diagnostic accuracy. Therefore, robust artifact detection is essential for improving image quality and ensuring reliable pathological assessments. However, existing methods often struggle with the wide variability of artifact types, leading to high computational cost and poor feature representation. Furthermore, most methods rely on a single-color domain, resulting in weak color perception in Hematoxylin-Eosin (H&E) stained images and reduced ability to distinguish artifacts from normal tissue. To address these issues, we propose a Dual-Color-Domain Siamese Network (DCDSN) that leverages RGB and HSV color domains. By minimizing representation discrepancies between two color domains via Siamese network similarity learning, our method achieves better feature alignment and enhances pathological feature representation. Additionally, to improve efficiency, we integrate a lightweight MobileViT-XS backbone with transfer learning, significantly reducing computational cost. And we introduce a Dynamic Snake Convolution-based feature mapper to enhance the network's sensitivity to subtle artifact features and reduce misclassification between artifacts and diseased tissues. Experiment results show that DCDSN achieves a 90.97% accuracy, outperforming the AR-Classifier baseline. Notably, it reduces parameter counts and computation by 74.39% and 96.55% respectively, demonstrating strong performance with lower resource consumption in artifact detection tasks.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":"11 5","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145129989","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A 3D multi-task network for the automatic segmentation of CT images featuring hip osteoarthritis.","authors":"Hongjie Wang, Xiaogang Zhang, Shihong Li, Xiaolong Zheng, Yali Zhang, Qingyun Xie, Zhongmin Jin","doi":"10.1088/2057-1976/ae0593","DOIUrl":"10.1088/2057-1976/ae0593","url":null,"abstract":"<p><p>Total hip arthroplasty (THA) is the primary treatment for end-stage hip osteoarthritis, with successful outcomes depending on precise preoperative planning that requires accurate segmentation and reconstruction of periarticular bone of the hip joint. However, patients with hip osteoarthritis typically exhibit pathological characteristics, including joint space narrowing, femoroacetabular impingement, osteophyte formation. These changes present significant challenges for traditional manual or semi-automatic segmentation methods. To address these challenges, this study proposed a novel 3D UNet-based multi-task network to achieve rapid and accurate segmentation and reconstruction of the periarticular bone in hip osteoarthritis patients. The bone segmentation main network incorporated the Transformer module during the encoder to effectively capture spatial anatomical features, while a boundary-optimization branch was designed to address segmentation challenges at the acetabular-femoral interface. These branches were jointly optimized through a multi-task loss function, with an oversampling strategy introduced to enhance the network's feature learning capability for complex structures. The experimental results showed that the proposed method achieved excellent performance on the test set with hip osteoarthritis. The average Dice coefficient was 0.945 (0.96 for femur, 0.93 for hip), with an overall precision of 0.95 and recall of 0.97. In terms of the boundary matching metrics, the average surface distance (ASD) and the 95% Hausdorff distance (HD95) were 0.58 mm and 3.55 mm, respectively. The metrics showed that the proposed automatic segmentation network achieved high accuracy in segmenting the periarticular bone of the hip joint, generating reliable 2D masks and 3D models, thereby demonstrating significant potential for supporting THA surgical planning.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145032707","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DBCM-net:dual backbone cascaded multi-convolutional segmentation network for medical image segmentation.","authors":"Xiuwei Wang, Biyuan Li, Jinying Ma, Lianhao Huo, Xiao Tian","doi":"10.1088/2057-1976/ae0482","DOIUrl":"https://doi.org/10.1088/2057-1976/ae0482","url":null,"abstract":"<p><p>Medical image segmentation plays a vital role in diagnosis, treatment planning, and disease monitoring. However, endoscopic and dermoscopic images often exhibit blurred boundaries and low contrast, presenting a significant challenge for precise segmentation. Moreover, single encoder-decoder architectures suffer from inherent limitations, resulting in the loss of either fine-grained details or global context. Some dual-encoder models yield inaccurate results due to mismatched receptive fields and overly simplistic fusion strategies. To overcome these issues, we present the Dual Backbone Cascaded Multi-Convolutional Segmentation Network (DBCM-Net). Our approach employs a Multi-Axis Vision Transformer and a Vision Mamba encoder to extract semantic features at multiple scales, with a cascaded design that enables information sharing between the two backbones. We introduce the Global and Local Fusion Attention Block (GLFAB) to generate attention masks that seamlessly integrate global context with local detail, producing more precise feature maps. Additionally, we incorporate a Depthwise Separable Convolution Attention Module (DSCAM) within the encoders to strengthen the model's ability to capture critical features. A Feature Refinement Fusion Block (FRFB) is further applied to refine these feature maps before subsequent processing. The cascaded network architecture synergistically combines the complementary strengths of both encoders. We rigorously evaluated our model on three distinct datasets, achieving Dice coefficients of 94.93% on the CVC-ClinicDB polyp dataset, 91.93% on ISIC2018, and 92.73% on ACDC, each surpassing current state-of-the-art methods. Extensive experiments demonstrate that the proposed method excels in segmentation accuracy and preserves edge details effectively.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":"11 5","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145074413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generalizable 2D medical image segmentation via wavelet-guided spatial-frequency fusion network.","authors":"Xiang Pan, Zhihao Shi, Herong Zheng, Qiuyu Li","doi":"10.1088/2057-1976/ae01a9","DOIUrl":"10.1088/2057-1976/ae01a9","url":null,"abstract":"<p><p>Medical image segmentation faces significant challenges in cross-domain scenarios due to variations in imaging protocols and device-specific artifacts. While existing methods leverage either spatial-domain features or global frequency transforms (e.g., DCT, FFT), they often fail to effectively integrate multi-scale structural cues with localized frequency signatures, leading to degraded performance under domain shifts. To address this limitation, we propose a novel framework that unifies spatial and wavelet-frequency representations through wavelet-guided fusion. Our approach introduces two key innovations: (1) a wavelet-guided multi-scale attention mechanism that decomposes features into directional subbands to capture domain-invariant structural patterns, and (2) an adaptive lateral fusion strategy that dynamically aligns frequency-refined decoder features with spatially enhanced skip connections. By leveraging the inherent localization and directional sensitivity of wavelet transforms, our method achieves superior preservation of anatomical boundaries across domains. Comprehensive evaluations on dermoscopy, ultrasound, and microscopy datasets demonstrate state-of-the-art performance across both seen and unseen domains. Compared to previous methods, WGSF-Net improves the Dice score by up to 1.5% on dermoscopy, 2.0% on ultrasound, and 13.9% on microscopy in unseen settings. These results validate that wavelet-guided spatial-frequency fusion effectively enhances generalization in 2D medical image segmentation.</p>","PeriodicalId":8896,"journal":{"name":"Biomedical Physics & Engineering Express","volume":" ","pages":""},"PeriodicalIF":1.6,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144940859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}