IEEE Transactions on Biomedical Engineering最新文献

{"title":"Neural Spelling: A Spell-Based BCI System for Language Neural Decoding.","authors":"Xiaowei Jiang, Jinzhao Zhou, Yiqun Duan, Ziyi Zhao, Yu-Cheng Chang, Thomas Do, Chin-Teng Lin","doi":"10.1109/TBME.2026.3691322","DOIUrl":"https://doi.org/10.1109/TBME.2026.3691322","url":null,"abstract":"<p><strong>Objective: </strong>Brain-computer interfaces (BCIs) support the study of communication-oriented neural decoding by translating neural activity into text, yet existing non-invasive systems rarely cover the full alphabet in handwriting-based settings.</p><p><strong>Methods: </strong>We propose a novel non-invasive EEG-based BCI framework, Curriculum-based Neural Spelling (CNS), that decodes all 26 English letters by first learning neural patterns associated with handwriting trajectories. A Generative AI (GenAI) module based on large language models (LLMs) is then integrated to transform noisy letter-level neural predictions into sentence-level outputs under explicit neural constraints.</p><p><strong>Results: </strong>The proposed system achieves robust letter-level decoding and improved sentence-level reconstruction under controlled offline evaluation, outperforming conventional EEGNet and hybrid CNN-RNN baselines. GenAI correction further reduces word error rates and enhances decoding fluency.</p><p><strong>Conclusion: </strong>Combining EEG-based neural spelling with generative language modeling supports the study of full-alphabet decoding and improves sentence-level linguistic metrics in a controlled non-invasive EEG setting, but does not by itself establish clinical or real-world usability.</p><p><strong>Significance: </strong>This work demonstrates how integrating GenAI with neural decoding can bridge the gap between noisy signal-level predictions and coherent language-level outputs, establishing a system-level framework for full-alphabet neural spelling and adaptive language-level correction under non-invasive EEG constraints.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147856279","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":"Transformer-Based Modular Motion Generation Through Shoulder-Arm Decoupling for Upper Limb Rehabilitation.","authors":"Muhammad Fawad Khan, Naveed Ahmad Khan, Fahad Hussain, Prashant K Jamwal, Shahid Hussain","doi":"10.1109/TBME.2026.3691335","DOIUrl":"https://doi.org/10.1109/TBME.2026.3691335","url":null,"abstract":"<p><p>The shoulder girdle is one of the most complex components of the upper limb, and when coupled with the arm, modeling and prediction become even more challenging. To address this, we propose a modular motion-planning framework that explicitly decouples the shoulder girdle from the arm. This separation simplifies modeling, enhances interpretability, and improves adaptability for rehabilitation scenarios. By independently targeting scapular dynamics, the framework enables better generalization of motion prediction across subjects. For human-motion modeling, a Transformer-based deep learning architecture is employed to capture nonlinear dependencies between joint angles and scapular motion. The model accepts joint-specific features as input and predicts shoulder girdle configurations, which are then integrated with arm trajectories to reconstruct complete upper limb motion. In the exoskeleton-mapping stage, a machine learning framework translates predicted human motion into the configuration space of a 6- Degree of freedom (DOF) rehabilitation exoskeleton. This ensures that generated trajectories are physically realizable and clinically suitable. By isolating human-motion prediction from robot mapping, the framework remains modular, scalable, and resilient to subject-specific variability, making it ideal for personalized rehabilitation. The methodology was evaluated by comparing Transformer-based predictions with both experimental data, reinforcement learning models and long short-term memory across multiple rehabilitation tasks. Quantitative analyses included statistical measures (F1-Score, ANOVA, T-test) and kinematic error metrics (RMSE, DTW). Results demonstrated that the Transformer model achieved higher accuracy and better temporal alignment with experimental trajectories. Combining shoulder-arm separation with Transformer-based learning provides an effective and clinically relevant solution for generating human-like motion in upper limb rehabilitation.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147856212","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":"Effect of Insertion Parameters on Insertion Force and Tissue Damage During Rigid Neural Probe Implantation.","authors":"Zebin Jiang, Mohammad Kafi Kangi, Xiang Liu, Saumya Nigam, James R Siegenthaler, Erin K Purcell, Ping Wang, Yan Gong, Wen Li","doi":"10.1109/TBME.2026.3691664","DOIUrl":"https://doi.org/10.1109/TBME.2026.3691664","url":null,"abstract":"<p><strong>Objective: </strong>The implantation of neural probes is critical for precise recording and stimulation of target neurons. However, the implantation of rigid neural probes, involves risks such as tissue damage and foreign body reactions, which can lead to probe failure and irreversible brain injury. Previous studies have employed force response and crack formation during probe insertion to understand the mechanical dynamics of implantation. While many researchers have explored the impact of different probe parameters on the implantation and long-term biological responses, the study of mechanical parameters during insertion remains incomplete. In particular, there are ongoing debates surrounding the quantitative impact of insertion speed on potential tissue damage. This study investigates the interaction effects of insertion speed, insertion depth, and probe geometry parameters on insertion force and insertion-induced damage during probe implantation. Tungsten and boron-doped diamond (BDD) probes were used as representative examples in this research. Peak insertion force and crack size were quantitatively evaluated in both agarose hydrogels and brain tissues, taking insertion direction and relatively wide speed range into account. Our results revealed a previously unreported fourth-order relationship between insertion speed and peak force within a certain insertion depth range, which can be understood as a Taylor-series approximation of the underlying rate- and state-dependent friction behavior within the experimental velocity regime. Meanwhile, crack analysis further showed an inverse relationship between crack size and insertion speed. These findings offer valuable insights into the mechanics of probe implantation with the goal of further improving the safety and reliability of neural implants.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147856273","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":"EMBC Special Issue: CogniFuse and Multimodal Deformers: An Extended Study on Benchmarking and Modeling Biosignal Fusion.","authors":"Anthony Richardson, Michael Beetz, Tanja Schultz, Felix Putze","doi":"10.1109/TBME.2026.3691190","DOIUrl":"https://doi.org/10.1109/TBME.2026.3691190","url":null,"abstract":"<p><p>Human physiological signals reflect complex biological processes and provide important insights into physical and mental states. Extracting such information from biosignal data collected during everyday activities holds great potential for real-time monitoring of physical and mental states, but is challenging due to noise and artifacts. To address this, we introduce CogniFuse, the first publicly available multi-task benchmark for multimodal biosignal fusion in such unconstrained environments. In addition, we develop a comprehensive benchmarking pipeline that emphasizes comparability, reproducibility, accessibility, and usability, while demonstrating robustness across architectures, tasks, and model sizes. For many biosignals, particularly electrophysiological signals, information in different frequency bands is critical for assessing physiological states. Motivated by this, we propose a family of Multimodal Deformer models that capture multi-level power features along with both long- and short-term temporal dependencies across multiple biosignal modalities. In particular, our Multi-Channel Deformer achieves the highest average benchmark score, outperforming all models of comparison. By advancing multimodal biosignal fusion in everyday settings, this work supports real-time monitoring of physical and mental states outside highly controlled clinical conditions. To ensure full transparency and reproducibility, and to facilitate future research, all code and data are made publicly available.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147837278","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":"EMBC Special Issue: Calibrated Uncertainty for Trustworthy Clinical Gait Analysis Using Probabilistic Multiview Markerless Motion Capture.","authors":"Seth Donahue, Irina Djuraskovic, Kunal Shah, Fabian Sinz, Ross Chafetz, R James Cotton","doi":"10.1109/TBME.2026.3691128","DOIUrl":"https://doi.org/10.1109/TBME.2026.3691128","url":null,"abstract":"<p><strong>Objective: </strong>This study evaluates the calibration and accuracy of a probabilistic MMMC method. Video-based human movement analysis holds potential for movement assessment in clinical practice and research. However, the clinical implementation and trust of multi-view markerless motion capture (MMMC) require that, in addition to being accurate, these systems produce reliable confidence intervals to indicate how accurate they are for any individual. Building on our prior work utilizing variational inference to estimate estimate biomechanical variables against clinical gold-standards.</p><p><strong>Methods: </strong>We analyzed data from 68 participants across two institutions, validating the model against an instrumented walkway and standard marker-based motion capture. We measured the calibration of the confidence intervals using the Expected Calibration Error (ECE).</p><p><strong>Results: </strong>the model demonstrated reliable calibration, yielding ECE values generally $< 0.1$ for both step and stride length and bias-corrected gait kinematics. We observed a median step and stride length error of $sim 16$ mm and $sim 12$ mm respectively, with median bias-corrected kinematic errors ranging from $1.5^{circ }$ to $3.8^{circ }$ across lower extremity joints. Consistent with the calibrated ECE, the magnitude of the model's predicted uncertainty correlated strongly with observed error measures.</p><p><strong>Conclusion: </strong>These findings indicate that, as designed, the probabilistic model reconstruction quantifies epistemic uncertainty, shown by low absolute error and calibrated uncertainty.</p><p><strong>Significance: </strong>These findings highlight the potential to identify unreliable outputs without the need for concurrent ground-truth instrumentation through the use of this probabilistic model.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147837299","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":"TransVort: A Temporally-Coherent Physics-Guided Neural Network for Super-Resolving and Denoising 4D Flow MRI of Cerebrospinal Fluid.","authors":"Neal M Patel, Moses J Hamm, Sriram Baireddy, A J Schwichtenberg, Edward J Delp, Vitaliy L Rayz","doi":"10.1109/TBME.2026.3690957","DOIUrl":"https://doi.org/10.1109/TBME.2026.3690957","url":null,"abstract":"<p><strong>Objective: </strong>To enhance the diagnostic utility of 4D flow MRI in assessing cerebrospinal fluid (CSF) dynamics by super-resolving and denoising measured velocities using temporally coherent, physics-guided neural networks (PGNN).</p><p><strong>Methods: </strong>Synthetic 4D flow MRI was generated from 40 computational fluid dynamics (CFD) simulations across 10 ventricular geometries. These simulations were used to generate paired synthetic 4D flow MRI and high-resolution velocity fields used for supervised training. Here, we compare a previously developed temporally independent network (div-mDCSRN-Flow) using divergence-based regularization with two novel temporal PGNNs (tempo-mDCSRN-Flow using divergence-regularization and TransVort additionally constrained by the vorticity transport equation).</p><p><strong>Results: </strong>In application to synthetic 4D flow MRI of a double gyre flow showed the temporal PGNNs improve vorticity estimation. Similarly, both temporal methods improved estimation of vorticity and time-averaged wall shear stress (TAWSS) of synthetic 4D flow MRI in the 3rd and 4th ventricle. While using temporal PGNNs improves velocity and vorticity quantification across temporal resolutions, TransVort demonstrated additional improvement at fine temporal resolutions. Application of TransVort to in vivo 4D flow MRI of CSF flow captured vortex formation and dissipation in the 4th ventricle over the cardiac cycle.</p><p><strong>Conclusion: </strong>Leveraging the temporal information of 4D flow MRI improves reconstruction of high-resolution velocity fields. This leads to better estimation of gradient-based flow metrics such as vorticity and TAWSS, which are associated with neurodegenerative and neurovascular diseases.</p><p><strong>Significance: </strong>Augmenting the accuracy of 4D flow MRI increases its potential for adoption as a clinical tool for diagnosing and monitoring disorders of neurofluid dynamics.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147837326","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":"Ultrasound Thermometry Using Echo Stretching for Microwave Hyperthermia.","authors":"Muthu Rattina Subash Ramu, Kavitha Arunachalam, Arun Kumar Thittai","doi":"10.1109/TBME.2026.3690468","DOIUrl":"https://doi.org/10.1109/TBME.2026.3690468","url":null,"abstract":"<p><p>This study presents an echo-stretching-based ultrasound technique for estimating temperature rise during hyperthermia therapy, eliminating the need for low-pass filtering and gradient computation used in conventional methods. The algorithm was evaluated using multiphysics simulations for thermal gradient of 0-0.1 °C using consecutive frames, and 0-6 °C using initial frame as reference. Inadequate up-sampling of radio frequency (RF) beamformed data resulted in spikes in the temperature estimates on echo stretching, which was reduced using adaptive up-sampling and median filtering. The average temperature estimation error relative to the peak temperature rise was less than 5% for 6 °C temperature gradient using 10λ sliding window, where, λ is the wavelength of ultrasound excitation in soft tissue. Window length of 40λ could resolve temperature gradient < 0.1 °C at the cost of spatial resolution. Axial resolution of 2.5 to 5 mm was achieved in hyperthermia temperature rise of 6 °C. Temperature estimation deteriorated with decline in signal to noise ratio (SNR) and depth. Experimental verification of echo stretching algorithm is tissue mimicking phantom and ex-vivo bovine tissues subjected to microwave hyperthermia at 434 MHz for 6 minutes using water loaded microstrip patch antenna indicated estimation error < 5 and 20% in phantoms and heterogeneous ex-vivo tissues, respectively. It is concluded that temperature rise estimated using the echo stretching of ultrasound RF data could be used for microwave hyperthermia treatment monitoring.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147837347","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":"Stable Wireless Power Transfer Using a Novel Omnidirectional Receiver and a Flexible Transmitter for Capsule Robots.","authors":"Heng Zhang, Zong-Ying Lin, Ching-Ming Lai, Chi-Kwan Lee","doi":"10.1109/TBME.2026.3690621","DOIUrl":"https://doi.org/10.1109/TBME.2026.3690621","url":null,"abstract":"<p><p>Non-invasive capsule robots offer significant advantages for painless gastrointestinal examination. However, with the advancement of capsule robot technology, its energy demand has significantly increased. Due to its compact size, integrating a high-capacity battery remains challenging, often leading to power insufficiency issues. In this paper, we propose a flexible transmitter independent of human body size and a 3-dimensional receiving coil (3DRC) based on flexible PCB. A mathematical model is developed to analyze the relationship between the transmitter and receiver, while a bending model is established to characterize the flexible transmitter coil. Furthermore, two control strategies, the dual-loop control strategy and the single-loop control strategy, are proposed to regulate the output voltage for stable wireless power transfer. Dynamic step response experiments are conducted to compare the performance of the two control algorithms. The stability of wireless charging is first evaluated under static conditions at various positions, followed by dynamic stability tests incorporating different translational velocities and angular rotation speeds. Experimental results demonstrate that at a target voltage of 3300 mV, the mean absolute error is limited to 20.2 mV, corresponding to less than 1% of the nominal voltage, thereby confirming the high regulation accuracy and stable performance of the proposed wireless charging system for capsule robots.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147837271","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 framework for quantifying and leveraging uncertainty in pre-trained CT denoising model.","authors":"Hao Gong, Nathan R Huber, Shravani A Kharat, Joel G Fletcher, Chi Wan Koo, Lifeng Yu, Shuai Leng, Scott S Hsieh, Cynthia H McCollough","doi":"10.1109/TBME.2026.3690022","DOIUrl":"https://doi.org/10.1109/TBME.2026.3690022","url":null,"abstract":"<p><strong>Objective: </strong>To develop an architecture-agnostic framework that estimates, calibrates, and leverages total uncertainty (aleatoric + epistemic) in pre-trained, deep-learning denoising models for low-dose computed tomography (CT).</p><p><strong>Methods: </strong>Aleatoric and epistemic uncertainties were estimated using physics-based inference-time augmentation and training-free, post-hoc Monte Carlo dropout, respectively, followed by non-parametric re-calibration for improved uncertainty calibration. To leverage uncertainty, we explored adaptive local fusion (ALF) guided by local mean-to-uncertainty ratio. For proof-of-concept, this framework was assessed using pre-trained U-net and ResNet-based models across datasets varying in CT tasks, radiation dose, and lesion characteristics. Uncertainty estimation and calibration were assessed in cadaver scans, using normalized-root-mean-square-error (NRMSE) and normalized-calibration-error (NCE), respectively. ALF was evaluated with chest and liver exams, using noise, structural similarity index (SSIM), and lesion detectability. Lesion detectability was quantified using clinically validated deep-learning model observer, with Wilcoxon signed-rank test to assess significance.</p><p><strong>Results: </strong>This framework provided accurate uncertainty quantification and calibration: NRMSE range [1.2%, 2.4%], NCE [0.9%, 2.2%]. Compared to original pre-trained models, ALF yielded comparable or lower noise, improved lesion structural fidelity and detectability (p<0.05): For lung nodules - noise reduction up to 69.7%, SSIM range (ALF vs pre-trained) [0.92, 0.96] vs [0.78, 0.87], detectability improvement up to 12.9%; for liver metastases - noise reduction up to 35.0%, SSIM range (ALF vs pre-trained) [0.82, 0.99] vs [0.80, 0.98], detectability improvement up to 13.2%.</p><p><strong>Conclusion: </strong>Our framework effectively benchmarked and utilized total uncertainty to enhance diagnostic image quality with pre-trained CT denoising models.</p><p><strong>Significance: </strong>This framework can facilitate performance monitoring, deployment optimization, and trustworthiness establishment.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147837255","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":"How Statistical Methods, Hemispheric Data and Masking Approaches Shape Probabilistic Sweet Spots in Deep Brain Stimulation.","authors":"Vittoria Bucciarelli, Dorian Vogel, Teresa Nordin, Jerome Coste, Jean-Jacques Lemaire, Karin Wardell, Raphael Guzman, Simone Hemm","doi":"10.1109/TBME.2026.3690018","DOIUrl":"https://doi.org/10.1109/TBME.2026.3690018","url":null,"abstract":"<p><strong>Objective: </strong>Probabilistic mapping is increasingly used to identify optimal stimulation regions (Probabilistic Sweet Spots, PSS) in Deep Brain Stimulation (DBS). Outcomes, however, depend on workflow parameters. This study examined how methodological and data-handling choices affect PSS stability and spatial consistency across varying sample sizes.</p><p><strong>Methods: </strong>Intraoperative stimulation test data from 36 Parkinson's Disease patients were analyzed. PSS were computed across increasing sample sizes using four statistical approaches: Bayesian t-test (BAYES), Logistic Regression Model (LRM), Wilcoxon test with FDR correction (WFDR), and Wilcoxon test with permutation correction (WPERM). We assessed the effects of statistical tests, hemispheric data handling, and masking parameters (i.e., minimum number of patients and stimulations per voxel) on PSS stability and consistency, evaluated in terms of size and spatial location.</p><p><strong>Results: </strong>BAYES was more robust at small to intermediate sample sizes, while WFDR and LRM stabilized only in larger cohorts (∼25-30 patients). WPERM consistently underperformed. Stability was higher in the left hemisphere. Combining hemispheres did not improve stability, suggesting asymmetries in stimulation effects. Masking parameters mainly affected PSS volume, with stricter thresholds reducing absolute size, but did not alter stability patterns.</p><p><strong>Conclusion: </strong>Statistical test choice, hemispheric analysis, and masking parameters strongly influence PSS outcomes. The Bayesian t-test is recommended for small to intermediate cohorts, and hemispheres should be analyzed separately to avoid masking clinically relevant asymmetries.</p><p><strong>Significance: </strong>By highlighting the interplay between sample size, statistical methods, hemispheric data, and masking strategies, this work contributes to standardizing probabilistic mapping practices and improving their reliability for clinical translation.</p>","PeriodicalId":13245,"journal":{"name":"IEEE Transactions on Biomedical Engineering","volume":"PP ","pages":""},"PeriodicalIF":4.5,"publicationDate":"2026-05-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147837253","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}