Biomedical Engineering Letters最新文献

{"title":"A comprehensive system for eyelid analysis using deep learning: automated measurement of eyelid position and corneal exposure.","authors":"Bokeun Song, Yeon Gyu Han, Sundariya Tsetsegjargal, Kyungmin Cho, Chang-Wook Seo, Sunje Kim, Dongheon Lee","doi":"10.1007/s13534-025-00546-9","DOIUrl":"10.1007/s13534-025-00546-9","url":null,"abstract":"<p><p>Quantitative eyelid analysis is crucial for diagnosing and managing eyelid disorders, such as ptosis, and for surgical planning. However, traditional manual measurements of Marginal Reflex Distance (MRD) and Corneal Exposure Ratio are often error-prone and inconsistent. Existing deep learning approaches using RGB imaging face challenges in accurately delineating eye structures due to insufficient contrast and varying iris pigmentation. This study introduces a comprehensive system integrating infrared (IR) imaging with deep learning-based eye segmentation to overcome these limitations. The system utilizes an IR camera with facial rotation compensation for consistent image acquisition and employs deep learning algorithms optimized for IR imaging characteristics. A dataset of 2,121 IR images from 303 patients (internal validation) and 406 public dataset images (external validation) was used. Multiple models including RITnet and SegFormer were evaluated, with data augmentation techniques applied to improve performance. SegFormer with data augmentation demonstrated the highest performance, achieving mean IoU of 0.9325 for internal validation and 0.9288 for external validation. Our automated system achieved improved precision compared to manual measurements, with MRD1 MSE of 0.362±0.511 mm and MRD2 MSE of 0.286±0.324 mm. This proposed system offers a robust, efficient, and standardized approach for quantitative eyelid analysis, leveraging IR imaging advantages to address the shortcomings of conventional assessment methods and providing clinical-grade accuracy for automated eyelid parameter measurement.</p><p><strong>Supplementary information: </strong>The online version contains supplementary material available at 10.1007/s13534-025-00546-9.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 3","pages":"747-758"},"PeriodicalIF":3.4,"publicationDate":"2026-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13129012/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147822016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of an intraoperative acoustic lung monitoring system for real-time respiratory rate detection in patient receiving mechanical ventilation.","authors":"Kyu-Min Kang, Woo-Jin Kim, Woo-Young Seo, Hong-Cheol Yoon, Junik Park, Hyun-Seok Kim, Sung-Hoon Kim","doi":"10.1007/s13534-025-00527-y","DOIUrl":"10.1007/s13534-025-00527-y","url":null,"abstract":"<p><p>Respiratory monitoring during mechanical ventilation is constrained by procedure complexity, operator dependency, and delayed response. This study aimed to develop an intraoperative lung monitoring system that calculates respiratory rate (RR) from lung sounds acquired via an esophageal stethoscope, providing a simple alternative to conventional monitoring techniques. An acoustic signal processing algorithm was developed to analyze lung sounds and detect RR in real time. The system integrated a microphone sensor into a standard esophageal stethoscope and applied preprocessing, dual-threshold peak detection, signal-to-noise ratio filtering, and RR confirmation logic. To evaluate the system under clinical conditions, three experiments were conducted: to assess measurement accuracy by comparing RR measurements with reference values independently derived from the CO₂ waveform, to compare measurement error by analyzing error metrics in subgroups stratified by ventilation mode, and to measure detection latency relative to a conventional capnography monitor. Lung sound signals were recorded from 37 adults under general anesthesia. The acoustic algorithm showed higher measurement accuracy, with a mean difference of 0.00 ± 0.11 compared to 0.04 ± 0.34 for capnography. Measurement error metrics were similar between volume-controlled and pressure-controlled ventilation modes (mean difference = 0.00 ± 0.12 and 0.00 ± 0.08, respectively). The acoustic system detected RR changes more rapidly (9.76 ± 2.78 s) than capnography (48.12 ± 5.64 s), reflecting significantly shorter detection latency. The proposed system demonstrated higher measurement accuracy than conventional capnography, exhibited similar measurement errors in both ventilation modes, and achieved significantly shorter detection latency. These results support its feasibility for real-time intraoperative respiratory rate monitoring.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 2","pages":"439-449"},"PeriodicalIF":2.8,"publicationDate":"2026-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13013729/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147522029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SSA-DCNet: a cross-session MI-EEG classification network based on deformable convolution and spatial-shift attention.","authors":"Xiuli Du, Hanxing Wang, Meiling Xi, Shaoming Qiu, Yana Lv","doi":"10.1007/s13534-025-00548-7","DOIUrl":"10.1007/s13534-025-00548-7","url":null,"abstract":"<p><p>Brain-computer interfaces (BCIs) based on motor imagery (MI) electroencephalogram (EEG) signals have shown tremendous potential in neurorehabilitation due to their non-invasive acquisition and ease of use. However, the cross-session nature of EEG signals-where recordings from the same subject at different sessions may vary due to fluctuations in physiological state and environmental conditions-presents a significant challenge. Efficient extraction and preservation of temporal and spatial features from EEG signals can capture invariant neural activation patterns while suppressing session-dependent noise and variability, thereby greatly enhancing the robustness of cross‑session motor imagery classification. To address the suboptimal performance of existing models in cross-session MI-EEG classification, this paper proposes Spatial-Shift Attention Deformable Convolution Network-SSA-DCNet, a compact convolutional neural network in which temporal filtering is implemented via a two-dimensional deformable convolution of size 1 × 64, so that the sampling grid dynamically adapts to the non-uniform distributions of informative EEG segments while operating on a 1 × 64 kernel along the temporal axis. Thereafter, a spatial-shift attention architecture expands each intermediate feature map from C to 3 C channels, evenly splits them into three subsets, applies distinct spatial-shift operations to each subset, and finally merges them via a split-attention that recalibrates channel weights to emphasize spatial patterns stable across sessions. On the public BCI Competition IV-2a and 2b datasets, SSA-DCNet achieved classification accuracies of 84.72% and 90.45%, respectively. Moreover, t-SNE visualizations provide intuitive evidence, underscoring its superior discriminative power and robust cross-session generalization.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 3","pages":"769-780"},"PeriodicalIF":3.4,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13129041/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147822161","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Recent strategies for designing adhesive hydrogels to enhance muscle tissue regeneration.","authors":"Yesung Lee, Youbin Park, Mikyung Shin","doi":"10.1007/s13534-025-00540-1","DOIUrl":"10.1007/s13534-025-00540-1","url":null,"abstract":"<p><p>Adhesive hydrogels have seen rapid advancement through the development of innovative adhesion strategies and versatile fabrication techniques, enabling robust tissue integration and localized therapeutic delivery. In regenerative medicine, muscle tissue regeneration has emerged as a key application area where these materials can promote repair, enhance bio-integration, and support functional recovery. This review summarizes recent progress in the design of adhesive hydrogels, including structural strategies for tissue-specific adhesion and fabrication methods for various hydrogel platforms. Key physicochemical properties, representative biomedical applications, and design principles for optimizing muscle regeneration are discussed, providing a concise yet comprehensive overview of their translational potential in clinical settings. Key design principles for optimizing muscle regeneration are discussed, with particular emphasis on their underlying physicochemical properties and representative biomedical applications, providing a concise yet comprehensive overview of their translational potential in clinical settings.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 2","pages":"217-235"},"PeriodicalIF":2.8,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13013798/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147522279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Metasurfaces for biomedical applications: materials design, optical engineering, and system integration.","authors":"Minseok Lee, Seungjin Jeong, Bongjoong Kim","doi":"10.1007/s13534-025-00545-w","DOIUrl":"10.1007/s13534-025-00545-w","url":null,"abstract":"<p><p>Metasurfaces, planar optical elements composed of subwavelength nanostructures, are catalyzing a paradigm shift in biomedical imaging and diagnostics through their unprecedented ability to control light at the nanoscale. This review provides a materials-centric analysis of the critical progression in metasurface design: from foundational all-metal plasmonic systems to all-dielectric platforms, and ultimately, their integration with CMOS technology for biomedical applications. We first examine all-metal metasurfaces, which leverage intense localized surface plasmon resonances (LSPRs) for field-enhanced spectroscopy but are fundamentally constrained by intrinsic Ohmic losses that limit resonance quality factors (Q-factors) and sensing resolution. Hybrid metal-dielectric architectures are then analyzed as a strategy to mitigate these losses while retaining strong field enhancement. A significant performance leap is realized with all-dielectric metasurfaces, where low-loss materials support high-Q Mie resonances and quasi-bound states in the continuum (quasi-BIC), enabling unprecedented sensitivity. This advancement facilitates record-breaking detection limits in label-free biosensing and single-molecule fluorescence analysis. Beyond biosensing, we demonstrate how wavefront engineering principles address long-standing limitations in biomedical imaging, enabling miniaturized, high-performance systems for optical coherence tomography (OCT), multi-contrast microscopy, and endoscopy. Finally, the review addresses the crucial challenge of clinical translation by exploring system-level integration with CMOS technology and culminating in a vision of AI-enabled sensing platforms that promise to transform point-of-care diagnostics and fundamental biological discovery.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 2","pages":"251-281"},"PeriodicalIF":2.8,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13013758/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147522308","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Materials strategies and multidisciplinary approaches for low-temperature preservation of 3D cellular systems.","authors":"Seongsoo Kim, Khandoker Asiqur Rahaman, Jung Hoon Lee, Hyung-Seop Han, Hojeong Jeon, Yongdeok Kim","doi":"10.1007/s13534-025-00541-0","DOIUrl":"10.1007/s13534-025-00541-0","url":null,"abstract":"<p><p>Recent advances in tissue engineering have paved the way for sophisticated three-dimensional (3D) cellular systems, including organoids and microphysiological systems, which hold great promise in fields like regenerative medicine, disease modeling, and drug discovery. Despite these advances, the lack of effective preservation strategies for 3D systems continues to hinder their biobanking, large-scale distribution, and practical implementation. This review explores the key principles and challenges involved in the low-temperature preservation of 3D cellular constructs and presents multidisciplinary strategies to overcome them. While cryopreservation techniques have been widely used in single cell levels, yet its translation to 3D multicellular systems is limited by restricted cryoprotectant diffusion, cytotoxicity, and nonuniform thermal transfer within 3D complex architectures. Recent innovations integrating thermodynamics, biophysics, biochemistry, materials science, and bioengineering have focused on better ice formation control, minimizing cryoprotectant agent toxicity, and enhancing heat and mass transfer. Finally, the review discusses emerging materials-based and system-level innovations, including nanowarming, novel cryoprotectants, and microfluidics- or isochoric-based systems, which collectively aim to achieve scalable, reproducible, and clinically relevant preservation of organoids, microphysiological systems, and other tissue-scale systems, bridging the translational gap between cellular cryopreservation and functional tissue preservation.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 2","pages":"237-250"},"PeriodicalIF":2.8,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13013855/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147522218","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Synergistic bone regeneration through sequential dual-drug delivery.","authors":"Jae Won Jang, Hyunji Kim, Joon Seok Oh, Dongtak Lee, Hye-Jung Choi, Hyunjung Shin, Chang-Nam Son, Xian Jin, Hyo Gi Jung, Yonghwan Kim, Youngjun Seo, Jisung Yoon, Jae Hyung Park, Young Hag Koh, Seung-Hyun Kim, Hak Chang, Dae Sung Yoon","doi":"10.1007/s13534-025-00539-8","DOIUrl":"10.1007/s13534-025-00539-8","url":null,"abstract":"<p><p>The global aging population has raised concerns regarding age-related health issues like osteoporosis and bone fractures. To address these conditions, bone-like scaffolds containing bioactive molecules and biomaterials have been widely studied. However, uncontrolled burst release and delivery of drugs can incur negative side effects. To overcome this issue, a collagen-hydroxyapatite scaffold (COHAS) that can sequentially deliver Bone morphogenetic protein-2 (BMP-2) and Osteoprotegerin fused to the Fc region of immunoglobulin (OPG-Fc) is synthesized. The COHAS comprises a collagen-hydroxyapatite matrix containing BMP-2 and numerous poly-lactic glycolic acid (PLGA) microspheres with OPG-Fc, dispersed in the matrix. The dispersion of PLGA microspheres enables the retardation of OPG-Fc release compared to BMP-2 release. The controlled sequential delivery of BMP-2 and OPG-Fc exhibits synergistic potential in promoting new bone formation by simultaneously activating osteoblasts and deactivating osteoclasts. This investigation revealed that the COHAS co-loaded with BMP-2 and OPG-Fc possesses excellent cell viability and enhanced osteogenic properties in vitro. In vivo assessment via implantation of the drug-loaded COHAS using an 8 mm-calvarial defect rat model demonstrated high efficacy of new bone formation with good biocompatibility. Hence, these findings provide valuable insights for developing therapeutic scaffolds capable of sequential release of multiple drugs, with the potential to extend a cell-free treatment system for bone regeneration.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 3","pages":"705-718"},"PeriodicalIF":3.4,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13129004/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147822157","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-Scale convolutional neural networks integrated with self-attention for motor imagery EEG decoding.","authors":"Shutong Duan, Penghai Li, Ding Yuan, Kun Wang, Dongfang Yu, Longlong Cheng","doi":"10.1007/s13534-025-00542-z","DOIUrl":"10.1007/s13534-025-00542-z","url":null,"abstract":"<p><p>Brain-computer interface (BCI), as a cutting-edge technology with great application prospects, has received widespread attention in recent years. Motor imagery (MI) electroencephalography (EEG) classification is a key component of brain-computer interfaces, widely used in applications such as assisting people with disabilities, controlling devices, and interacting with environments. However, since convolutional neural networks (CNNs) extract only local temporal features, they may be unable to capture the long-term dependencies used for EEG decoding, which can have an impact on the decoding performance. In order to address this problem, this paper proposes a novel deep learning network that combines a multi-scale convolutional neural network with an attention mechanism to capture temporal information and global dependencies. First, a multi-scale structure is designed to extract spatial-temporal information at different scales and multimodal information from both the mean and variance perspectives. Second, a squeeze-excite-compress (SEC) module is used to enhance the feature response of each branch and reduce information redundancy. Finally, an encoder with a multi-head attention mechanism extracts more discriminative features and highlights the most valuable information in MI-EEG data. In addition, this paper uses a data augmentation method of signal reorganization to expand the dataset and further enhance the generalization ability of the network. Our method was evaluated by performing experiments on the BCI Competition IV-2a (BCI-IV-2a) and High Gamma Dataset (HGD) with classification accuracies of 85.26% and 95.86%, respectively. The experimental results show that our method achieves state-of-the-art performance and has great potential to be a new baseline for general EEG decoding.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 3","pages":"719-734"},"PeriodicalIF":3.4,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13129155/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147822159","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"TremorFusion: AI-driven feature extraction for multi-class Parkinson's tremor classification using CSVM and DeepK-CNN.","authors":"Mohammad Sakib, Shoma Khanom, Tachiya Mahamud Nahadi, Asif Mohammad Mithu, Nayeem Ahmed, Maisha Islam, Tania Sarker, N M Chisty, Md Ahad, M Mofazzal Hossain, Feroz Ahmed","doi":"10.1007/s13534-025-00526-z","DOIUrl":"10.1007/s13534-025-00526-z","url":null,"abstract":"<p><p>Parkinson's disease (PD) is a common neurological disorder that, if left untreated, can significantly impair a patient's social and physical functioning. This study evaluates the feasibility of a wrist-worn accelerometer sensor device for detecting and classifying five distinct types of PD tremors. A novel feature extraction technique is proposed for the Cubic Support Vector Machine (CSVM) model, along with a pre-processed tremor spectrum analysis utilizing Non-negative Matrix Factorization (NMF) for a hybrid model that integrates k-Nearest Neighbor (KNN) with a Convolutional Neural Network (CNN), referred to as DeepK-CNN. Poincaré-based analysis was employed to extract time-domain features from tremor data, which were subsequently selected for the CSVM model based on the area under the receiver operating characteristic (ROC) curve. The dataset comprises 90 patients, with 30 exhibiting tremor types I-V, contributing 151 labeled tremor episodes over 15 h of recorded data. Using leave-one-patient-out fivefold cross-validation, the CSVM model demonstrated superior performance, achieving 95.28% mean sensitivity, 95.51% mean specificity, and a mean false alarm rate (FAR) of 0.15/24 h. The DeepK-CNN model achieved an 87.51% mean sensitivity, 91.65% mean specificity, and a mean FAR of 0.64/24 h. Performance evaluation yielded <i>p</i>-values of <0.001 for the CSVM model and <0.05 for the DeepK-CNN model. The proposed method enhances real-time tremor classification, potentially improving personalized treatment strategies and facilitating remote patient assessment for clinical adoption.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 2","pages":"411-437"},"PeriodicalIF":2.8,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13013909/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147522378","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Prediction of mental fatigue states based on the CNN-LSTM architecture for BCG signals.","authors":"Liu Zhichao, Liu Ziqi, Li Xin","doi":"10.1007/s13534-025-00543-y","DOIUrl":"10.1007/s13534-025-00543-y","url":null,"abstract":"<p><p>To achieve feature fusion of ballistocardiogram (BCG) signals for the purpose of predicting mental fatigue states, a prediction model based on convolutional neural network-long short-term memory (CNN-LSTM) architecture is built in this paper. The model combines the BCG signal Heart Rate Variability (HRV) parameter values with multiscale features to achieve accurate prediction of the BCG signal state. The results show that the fatigue classification proposed in this paper is highly consistent with the subjective evaluation, and all the evaluation indexes are better than the traditional methods. It achieved an accuracy of 95%, a precision of 96%, a linear regression of 0.9, and a G-mean of 0.93. In addition, it is able to automatically accelerate the learning of features without human intervention. There is no need to perform a tedious feature extraction process compared to traditional models. The accuracy and real-time performance of mental fatigue monitoring can be improved by using this model.</p>","PeriodicalId":46898,"journal":{"name":"Biomedical Engineering Letters","volume":"16 3","pages":"735-746"},"PeriodicalIF":3.4,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13129116/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147822185","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}