IEEE Transactions on Biomedical Engineering最新文献

{"title":"Robust Epilepsy Diagnosis in Neonates Via Spatio-Temporal Graph Learning and Balanced Test-Time Training.","authors":"Shihan Guan, Regine Le Bouquin Jeannes, Maxime Yochum, Fabienne Poree, Huazhong Shu, Xin Chen, Lei Song, Alain Beuchee, Constantin Gomes, Amar Kachenoura, Laurent Albera","doi":"10.1109/TBME.2026.3748561","DOIUrl":"https://doi.org/10.1109/TBME.2026.3748561","url":null,"abstract":"<p><p>Timely diagnosis of neonatal epilepsy is critical for mitigating long-term neurodevelopmental deficits. However, while neonates exhibiting extensive seizure burden are readily identified, accurately diagnosing epileptic infants with sparse or no overt seizure onset during monitoring remains a formidable challenge using electroencephalogram (EEG) signals alone. Current automated systems predominantly target ictal-specific seizure detection, limiting their clinical utility for risk stratification during seizure-free intervals (interictal), particularly given the pronounced EEG variability across neonates. This work presents a cross-subject deep learning framework that robustly distinguishes epileptic from \"healthy\" neonates by integrating dynamic graph topology with self-supervised adaptation. Our architecture models multichannel EEG as spatio-temporal graphs to capture functional connectivity, while a balanced test-time training mechanism dynamically adapts the model to unseen patients via masked signal reconstruction. This effectively mitigates the domain shifts caused by inter-subject variability without requiring patient-specific labels. Validated on a clinically curated cohort of 140 neonates (70 healthy, 70 epileptic), our approach achieves a segment-level accuracy of 83$%$ and an area under the curve of 0.90. For the primary subject-level endpoint, accuracy increases from 0.79 without adaptation to 0.88 with balanced TTT, using aggregation thresholds selected on inner validation subjects and fixed before evaluation on each held-out fold. The cohort includes anoxic-ischemic encephalopathy, benign neonatal convulsions, and cerebrovascular accident. Analyses restricted to non-ictal windows also support discrimination between the two groups, although all neonates in the epilepsy group experienced seizures elsewhere in their recordings.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149297137","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 Finger Worn Ring for On-Demand Measurement of Absolute Blood Pressure.","authors":"Md Abdul Momin, Shipeng Wang, Mahdi Jazini, Mark Freithaler, Shahin Hashemkhani, Graham Ethan Hoover, Yuyang Li, Inhee Lee, Feng Xiong, Ramakrishna Mukkamala","doi":"10.1109/TBME.2026.3748379","DOIUrl":"10.1109/TBME.2026.3748379","url":null,"abstract":"<p><strong>Objective: </strong>Regular blood pressure (BP) monitoring is needed to manage hypertension. However, cuff devices are not readily accessible, and many adults are living with uncontrolled BP. Furthermore, popular wearable cuffless devices typically require calibration with periodic cuff BP measurements or demographic information. A finger worn ring for cuffless BP measurement without such calibration was investigated.</p><p><strong>Methods: </strong>BP ring concepts were conceived based on the oscillometric cuff principle. The user varies their finger pressures similar to a cuff device by performing ring tightening followed by hand lowering or ring tightening only under visual guidance, while the ring obtains oscillometric measurements from the digital arteries for on-demand BP measurement. A prototype device was built to implement the BP ring concepts. The ring included a mechanical actuator to adjust its diameter and a photoplethysmography sensor, force sensor, and inertial measurement unit to make the measurements. The ring is interfaced to a display via a wired data acquisition system to guide the user in performing the ring tightening/hand lowering. An oscillometric algorithm is applied to compute BP from the measurements. The device was tested in 25 volunteers.</p><p><strong>Results: </strong>The ring tightening-hand lowering BP measurement could be learned by new users. The device yielded root-mean-square-errors of 7.1 and 6.9 mmHg for systolic and diastolic BP against a validated arm cuff device.</p><p><strong>Conclusion: </strong>This study demonstrates the feasibility of absolute BP monitoring using a finger worn ring.</p><p><strong>Significance: </strong>The BP ring concepts could potentially be incorporated in smart rings to help improve hypertension control.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149297112","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":"Triphase MagRing-MPI: Topology Evolution and Vector Space Control for Electronic FFL Scanning.","authors":"Ziwei Chen, Guanghui Li, Zhongwei Bian, Guangxing Zhou, Jian'an Ye, Gen Shi, Zhaoji Miao, Yu An, Jie Tian","doi":"10.1109/TBME.2026.3747177","DOIUrl":"https://doi.org/10.1109/TBME.2026.3747177","url":null,"abstract":"<p><strong>Objective: </strong>To develop and validate a Triphase MagRing magnetic particle imaging (MPI) framework for electronically rotatable and steerable field-free-line (FFL) control based on topology evolution and vector-space control in a large-bore prototype.</p><p><strong>Methods: </strong>A Triphase MagRing-MPI architecture was formulated using the Clarke-Park transform. Three wiring topologies were compared in controllability, spectral behavior, and engineering complexity. A 200-mm-bore air-core prototype was evaluated by magnetic-field mapping, current-spectrum analysis, thermal monitoring, robustness simulation with measured disturbance injection, and head-phantom imaging.</p><p><strong>Results: </strong>The prototype achieved controllable FFL generation, rotation, and steering with a measured gradient of 0.071 T/m. Y-3 showed the cleanest current spectra and the simplest implementation. The prototype remained thermally stable during 35 min of operation under the tested duty cycle. Head-phantom imaging successfully localized tracer targets within a 120 × 120mm$^{2}$ field of view. Preliminary simulations with soft magnetic cores indicated that the central gradient could be increased to 2.15 T/m, suggesting further room for optimization.</p><p><strong>Conclusion: </strong>The proposed framework enabled stable electronically rotatable and steerable FFL operation in the present prototype and identified Y-3 as the preferred wiring topology.</p><p><strong>Significance: </strong>This work provides an engineering validation of MagRing-MPI in a triphase, three-pole-pair configuration, supports future extension to systems with more pole pairs and more phases, and provides an engineering basis for future large-bore MPI system development.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149266570","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":"Toward Wearable Joint Load Monitoring: Concurrent Estimation of Hip, Knee, and Ankle Sagittal Contact Forces from Sparse IMU Measurements.","authors":"Tao Ma, Xinxu Fan, Ningbo Yu, Jianda Han, Jiayin Guan, Tianyang Fan, Tao Sun","doi":"10.1109/TBME.2026.3746974","DOIUrl":"https://doi.org/10.1109/TBME.2026.3746974","url":null,"abstract":"<p><strong>Objective: </strong>Continuous monitoring of lower extremity joint contact forces is critical for rehabilitation medicine, sports injury prevention, and assistive device development. Yet, current wearable solutions often lack multi-joint integration or require cumbersome sensor arrays. This study aims to concurrently estimate hip, knee, and ankle sagittal compressive contact forces using a sparse inertial measurement unit (IMU) configuration across diverse daily activities.</p><p><strong>Methods: </strong>We collected data from 14 healthy adults equipped with four IMU sensors performing various daily movements, with reference contact forces captured using optical motion capture and force platforms. To estimate joint contact forces, we developed a domain-informed multi-task temporal convolutional network (Ada-Att-MTCN). This architecture integrates a shared temporal convolutional network encoder, task-specific attention mechanisms, homoscedastic uncertainty-based dynamic task weighting, and a soft physiological-bound penalty. Performance was evaluated using 14-fold leave-one-subject-out nested cross-validation against 4 baseline methods, with root mean square error (RMSE) and coefficient of determination (R²) as evaluation metrics.</p><p><strong>Results: </strong>On the self-collected dataset, Ada-Att-MTCN achieved the highest accuracy across all activities, with mean RMSEs of 0.328 ± 0.13 BW (hip), 0.380 ± 0.09 BW (knee), and 0.365 ± 0.15 BW (ankle), and an average R2 of 0.822 ± 0.10. On the external dataset, Ada-Att-MTCN likewise performed best among all methods (average R² = 0.754 ± 0.05).</p><p><strong>Conclusion: </strong>The proposed framework captures complex inter-joint dynamics, enabling robust simultaneous load estimation with drastically reduced hardware complexity.</p><p><strong>Significance: </strong>This approach provides a practical, non-invasive solution for transitioning lab-grade multi-joint load monitoring to wearable deployment in rehabilitation and sports medicine.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149266578","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":"Multimodal Analysis of Brain Network Abnormalities in Autism Spectrum Disorder: a Graph Neural Network-based Fusion Investigation of EEG and fNIRS.","authors":"Yuhang Li, Lizi Zhang, Lei Hou, Lin Wan, Gang Zhu, Guang Yang, Rihui Li","doi":"10.1109/TBME.2026.3746308","DOIUrl":"https://doi.org/10.1109/TBME.2026.3746308","url":null,"abstract":"<p><strong>Objective: </strong>This study aimed to identify brain network abnormalities of children with Autism Spectrum Disorder (ASD) by analyzing multimodal neuroimaging data through graph neural networks (GNNs), laying a technical foundation for future machine learning-based neuroimaging analysis.</p><p><strong>Methods: </strong>We employed combined electroencephalogram (EEG) and functional near-infrared spectroscopy (fNIRS) to capture brain activity of 58 ASD children and 63 age- and gender-matched typically developing (TD) children. The concurrent EEG and fNIRS data were fused using an enhanced GNN to identify ASD with high accuracy. Key connections and regions of interest (ROIs) were identified for each group, followed by graph theory metrics (e.g., local efficiency) and topological data analysis (TDA) to quantify group differences.</p><p><strong>Results: </strong>The proposed GNN model achieved 84% accuracy, outperforming the single-modality approach. Analysis of subgraph created by the model showed that the two groups exhibited different neural patterns in motor-language integration and lateralization, a key basis for GNN models to distinguish between the two groups. Individuals with ASD may also have topological defects such as lower average local efficiency and a lack of high-order structures.</p><p><strong>Conclusion: </strong>The proposed framework effectively captures quantitative insights ranging from specific connection-level variations to network-level differences, identifying candidate discriminative connectivity patterns within this cohort.</p><p><strong>Significance: </strong>This study establishes a multimodal deep learning framework that identifies candidate discriminative topological signatures from fused EEG-fNIRS data, providing preliminary within-cohort classification evidence and an exploratory topological analysis framework.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149244199","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":"The Simpler, the Better: A Lightweight Hierarchical Token Averager for Single-Label Microscopic Specimen Image Classification.","authors":"Dichao Liu, Kenji Suzuki","doi":"10.1109/TBME.2026.3745692","DOIUrl":"https://doi.org/10.1109/TBME.2026.3745692","url":null,"abstract":"<p><strong>Objective: </strong>Single-label microscopic specimen image classification (SMSIC) depends on subtle local morphology, yet many natural-image architectures devote substantial capacity to global interaction. We aim to develop an efficient local-first architecture tailored to this setting.</p><p><strong>Methods: </strong>We propose the Hierarchical Spatially Merged Token Averager (HSMTA), which reorganizes neighboring spatial elements and progressively compresses them into higher-level descriptors before a simple final average. HSMTA-SRR is introduced as an SRR-inspired, sparsity-regularized analytical counterpart for examining stage-wise activation selectivity and intra-class compactness.</p><p><strong>Results: </strong>Across SIPaKMeD, C-NMC, DSCC, and DeFungi, HSMTA achieves 98.1%, 76.6%, 94.5%, and 93.0% accuracy, respectively, with 1.1 M parameters and 0.4 GMACs. It outperforms five recent lightweight baselines on all four datasets and remains competitive with higher-capacity token-mixer models on DeFungi. Controlled comparisons with CNN+GAP, one-shot patch averaging, MIL-style pooling, and post-formation token mixers attribute the gain to hierarchical local descriptor formation rather than averaging alone. HSMTA-SRR further exposes stage-wise activation sparsification and intra-class compactness, providing model-centric diagnostics of representation formation.</p><p><strong>Conclusion: </strong>Local hierarchical descriptor formation followed by simple averaging provides a favorable accuracy-efficiency trade-off for morphology-dominant SMSIC tasks.</p><p><strong>Significance: </strong>HSMTA offers a compact specimen-level classifier for high-throughput digital microscopy workflows, where per-instance cost accumulates over many cells or local regions. The code is available at https://github.com/Dichao-Liu/HSMTA.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149238279","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":"Vascular Anatomy-aware Self-supervised Pre-training for X-ray Angiogram Analysis.","authors":"De-Xing Huang, Chaohui Yu, Xiao-Hu Zhou, Tian-Yu Xiang, Qin-Yi Zhang, Mei-Jiang Gui, Rui-Ze Ma, Chen-Yu Wang, Nu-Fang Xiao, Fan Wang, Zeng-Guang Hou","doi":"10.1109/TBME.2026.3745895","DOIUrl":"https://doi.org/10.1109/TBME.2026.3745895","url":null,"abstract":"<p><strong>Objective: </strong>X-ray angiography is the gold standard imaging modality for cardiovascular diseases. However, current deep learning approaches for X-ray angiogram analysis are severely constrained by the scarcity of annotated data. While large-scale self-supervised learning (SSL) has emerged as a promising solution, its potential in this domain remains largely unexplored, primarily due to the lack of effective SSL frameworks and large-scale datasets.</p><p><strong>Methods: </strong>To bridge this gap, we introduce a Vascular anatomy-aware Masked Image Modeling (VasoMIM) framework that explicitly integrates domain-specific anatomical knowledge. Specifically, VasoMIM comprises two key designs: an anatomy-guided masking strategy and an anatomical consistency loss. The former strategically masks vessel-containing patches to compel the model to learn robust vascular semantics, while the latter preserves structural consistency of vessels between the original and reconstructed images, enhancing the discriminability of the learned representations. In conjunction with VasoMIM, we curate XA-170K, the largest X-ray angiogram pre-training dataset to date.</p><p><strong>Results: </strong>We validate VasoMIM on four downstream tasks across six datasets, where it demonstrates superior transferability and achieves state-of-the-art performance compared to existing methods.</p><p><strong>Conclusion: </strong>Incorporating vascular anatomical knowledge into self-supervised pre-training enables more discriminative and transferable representations for X-ray angiogram analysis.</p><p><strong>Significance: </strong>These findings highlight the significant potential of VasoMIM as a foundation model for advancing a wide range of X-ray angiogram analysis tasks. VasoMIM and XA-170K are available on GitHub and Hugging Face, respectively.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149238245","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":"An Improved Fabrication Method for Flexible COC-Based Implantable Neural Probes with Reliability Validation.","authors":"Taekyung Lee, Joowon Lee, Seoyoung Hwang, Jisung Kim, Sang Beom Jun, Changhoon Baek, Jongmo Seo","doi":"10.1109/TBME.2026.3745757","DOIUrl":"https://doi.org/10.1109/TBME.2026.3745757","url":null,"abstract":"<p><p>Implantable neural interfaces require mechanically flexible and electrochemically stable materials to ensure long-term reliability, but conventional polymer substrates suffer from high moisture permeability and weak adhesion, leading to limited long-term stability. Cyclic olefin copolymer (COC) provides improved barrier properties, however, further optimization in flexibility, process scalability, and electrochemical performance remains important. Building on prior COC-based electrode developments, this study presents process updates and a comprehensive functional evaluation of COC-based neural probes. T-die-cast COC films show reduced thickness variation and increased flexibility, and femtosecond laser micromachining enables precise site opening with dimensional scalability. IrO$_{mathrm{x}}$ electroplating reduces impedance to 681.04 $Omega$ at 1 kHz and increases CSC$_{mathrm{c}}$ to 23.51 mC/cm$^{2}$, while the probe maintained stable performance under repeated bending and folding. Accelerated soaking confirms insulation reliability, with leakage remaining below 10$^{-9}$ A for 320 days at 75$^circ$C. Ex vivo retinal stimulation experiments showed evidence of evoked neural activity, supporting functional feasibility. Collectively, these findings indicate that refined COC processing and fabrication improvements address practical constraints of compression molded COC electrodes, enabling flexible, reliable, and electrochemically stable neural probes and broadening the practical applicability of COC.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149238225","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 Sensitive BLE-Enabled Wearable fNIRS Biosensor for Real-Time Monitoring and Mapping of Task-Evoked Oxygen Influxes During On-Screen Cognitive Tasks.","authors":"Rajesh Mandal, Gayathri Sai Prabhakaran, Arjun V Kowshik, Bhargav Makam Balaji, Balaji Jayaprakash","doi":"10.1109/TBME.2026.3745751","DOIUrl":"https://doi.org/10.1109/TBME.2026.3745751","url":null,"abstract":"<p><strong>Objective: </strong>To develop and validate a compact Bluetooth Low Energy (BLE)-enabled wearable functional near-infrared spectroscopy (fNIRS) device for real-time hemodynamic monitoring and proposes a task-evoked metric for identifying transient oxygenation events during cognitive engagement.</p><p><strong>Method: </strong>We designed a four-channel wearable fNIRS biosensor that integrates optical sensing, signal acquisition, and wireless transmission into a compact, cable-free platform using off-the-shelf electronic components. The device was characterized through optical power linearity, ADC voltage input, pulsed light acquisition, and signal-to-noise ratio analysis. Its performance was validated using forearm vascular occlusion and breath-hold experiments, followed by cortical recordings during breath-hold tasks, finger tapping, dynamic visual stimulation, and a screen-based spatial learning and memory task. We further formulated task-evoked instances of differential oxygen influx (tEIDO), computed from the instantaneous differential response between $Delta$[HbO] and $Delta$[Hb], as a hemodynamic index of task-locked neurovascular activity.</p><p><strong>Result: </strong>The device reliably detected expected oxy- and deoxy-hemoglobin changes during vascular and breath-hold validation experiments and captured task-evoked responses over motor, visual, and prefrontal cortices. The tEIDO metric identified transient differential oxygenation events during visual stimulation and revealed task-aligned oxygenation dynamics. During spatial navigation, differential oxygenation maps showed hemodynamic modulation around the target region only after memory acquisition.</p><p><strong>Significance: </strong>The proposed BLE-enabled fNIRS platform, together with tEIDO, offers a compact and sensitive approach for untethered real-time cortical hemodynamic monitoring, with potential applications in mobile neuroimaging, neuroergonomics, cognitive-state monitoring, and brain-computer interface systems.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149238230","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":"Function-based Deep Learning Autoencoder for Passive Shimming in a Portable MRI Magnet with Enhanced Field Homogeneity.","authors":"Jiannan Zhou, Xia Xiao, Guoquan Chen, Pengfei Zhao, Yanwei Pang, Zhenchang Wang","doi":"10.1109/TBME.2026.3745374","DOIUrl":"10.1109/TBME.2026.3745374","url":null,"abstract":"<p><p>High-field magnetic resonance imaging (MRI) offers excellent sensitivity and spatial resolution, yet its substantial cost and limited mobility restrict deployment in decentralized or resource-constrained environments. Portable MRI systems offer a complementary imaging modality to high-field MRI, but their performance depends critically on the uniformity of the main magnetic field (B0). Passive shimming, achieved by placing ferromagnetic materials within the magnet bore, is widely used to correct spatial field variations. Existing optimization strategies- including linear programming and evolutionary algorithms-can improve B homogeneity, but often struggle to achieve an effective trade-off among field strength, uniformity, and shim mass. To address this challenge, a function-based deep learning autoencoder (FDLAE) is proposed to optimize the spatial distribution of passive shims. The framework operates without a precompiled dataset and performs online self-supervised learning to iteratively refine the shim configuration. This approach improves B0 homogeneity from 4873 ppm to 547 ppm (parts per million) and increases the average field strength from 80.402 mT to 80.948 mT. Compared with linear programming (LP), it achieved lower inhomogeneity, reduced the shim mass by 0.29 kg, and shortened the optimization time from approximately 1.2 to 0.5 h. MRiLab simulations also showed a reduction in image NRMSE from 0.889 to 0.543. Relative to genetic-algorithm-based shimming, the resulting field uniformity is improved by nearly a factor of five. These results demonstrate that FDLAE may provide an efficient and comparatively robust passive-shimming approach for portable MRI magnets.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.4,"publicationDate":"2026-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149227742","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}