Epicardial adipose tissue (EAT) is increasingly recognized as an important contributor to cardiovascular disease (CVD), but its spatial distribution across the heart remains insufficiently characterized due to limitations in existing segmentation and analysis methods. In this study, we developed a deep learning-based framework that integrates automated cardiac MRI segmentation with spatial statistical modeling to enable quantitative, chamber-resolved characterization of EAT distribution. An asymmetric multi-modal CNN-Transformer network was developed to segment the four cardiac chambers and EAT from cardiac MRI. Based on the resulting whole-heart segmentations, EAT was automatically partitioned into four chamber-specific subregions, voxel-wise thickness maps were reconstructed, and spatial statistics were applied to identify localized clustering patterns of EAT. Chamber-resolved and region-specific quantitative features were extracted to characterize the spatial distribution of EAT across the heart. The proposed approach was evaluated on a cohort including individuals with type 2 diabetes (T2D) and matched controls, revealing distinct T2D-associated remodeling, including increased chamber-specific burden, localized thickening, and spatial hotspot clustering in metabolically vulnerable regions. This work introduces a scalable and automated method for regional EAT phenotyping and provides spatially resolved imaging biomarkers that may support CVD research and risk stratification.

Atrial fibrillation (AF), the most common cardiac arrhythmia, affects one in three adults over 45 years of age. Improving its treatment requires a better understanding of bi-atrial anatomy. Existing benchmarks have focused on the left atrial (LA) cavity, overlooking the fundamental challenges posed by bi-atrial anatomy, most notably the thin atrial walls, which are critical for substrate-guided ablation planning in patients with atrial fibrillation. To address these limitations, the Multi-class Bi-Atrial Segmentation 2024 Challenge (MBAS2024) introduced the first large-scale, multi-class benchmark for simultaneous segmentation of the LA cavity, right atrial (RA) cavity, and bi-atrial walls from late gadolinium-enhanced (LGE) MRI. We systematically evaluated 13 state-of-the-art methods on the world's largest curated bi-atrial dataset, comprising 175 3D multi-center scans with expert-validated annotations, providing a comprehensive assessment of current methodological capabilities and limitations. Key findings include: segmentation of the LA and RA cavities is generally robust to image quality, whereas atrial wall delineation is highly sensitive to image degradation. Performance varies across centers, indicating limited generalization of atrial wall segmentation across different acquisition protocols. Model architecture, rather than hyperparameter tuning, is the primary driver of performance, with U-Net-based models and emerging state-space models (e.g., UMambaBot) achieving higher accuracy at modest computational cost. Segmentation accuracy also varies along the slice dimension, with central slices segmented more reliably. Finally, hybrid labeling strategies-separating LA and RA cavities while merging bi-atrial walls into a single class-consistently improve performance. The MBAS2024 challenge establishes a foundational benchmark for bi-atrial segmentation, providing validated baselines and actionable insights to guide the development of clinically relevant, efficient, and anatomically aware segmentation algorithms to improve targeted ablation in patients with AF.

Atrial fibrillation (AF) is associated with progressive structural remodeling of the atria, including chamber dilation, fibrosis, and variations in atrial wall thickness (AWT). Late gadolinium-enhanced (LGE) magnetic resonance imaging (MRI) has been used to quantify left atrium (LA) fibrosis for guiding adjunctive ablation beyond pulmonary vein isolation, though results have varied. A major limitation is the lack of a robust segmentation method for accurately assessing both atrial anatomy and fibrosis, coupled with the exclusion of the right atrium (RA) in the analysis. This study introduces biAtriaNet, a deep learning pipeline developed to automate segmentation of both LA and RA and to evaluate atrial fibrosis, AWT, and chamber diameter and volume from LGE-MRIs to support targeted AF ablation. biAtriaNet was trained and validated on 2D cine-MRIs from 4860 UK Biobank participants and 3D LGE-MRIs from 60 AF patients from the University of Utah, with independent testing on 11 3D LGE-MRIs at Waikato Hospital, New Zealand. The biAtriaNet consists of two CNNs based on a modified U-Net architecture with residual connections and batch normalization, optimized based on prior global benchmark study. This approach achieved accurate, consistent segmentation and biomarker extraction in UK Biobank and Utah datasets, validated against expert annotations. Additionally, biAtriaNet showed high transferability to independent datasets, achieving Dice scores of 91.1% for LA and 88.6% for RA. Chamber volume estimates closely matched ground truth values (LA: 89.8 ± 33.0 ml versus 91.1 ± 41.2 ml; RA: 70.8 ± 16.9 ml versus 72.3 ± 20.5 ml) with >90% accuracy in chamber measurements. AWT accuracies were 95.9% for LA and 94.6% for RA, while fibrosis estimates showed Kolmogorov-Smirnov correlations of 86.3% (LA) and 90.6% (RA) (p < 0.05). By enabling robust bi-atrial segmentation and biomarker extraction from LGE-MRIs, biAtriaNet has the potential to enhance patient-specific AF treatment strategies.

Atrial fibrillation (AF) significantly affects morbidity and mortality rates. Class III antiarrhythmic drugs (AADs) play a crucial role in managing AF but often exhibit gender-specific complications. Our study aims to identify gender-specific Class III AADs by integrating in vitro measurements, in silico models, and machine learning (ML). By simulating drug effects on a diverse cardiomyocyte model population (5,663 males and 6,184 females), we classified drugs based on changes in action potentials and calcium transients. Using sex-dependent Support Vector Machine (SVM) algorithms, we achieved high prediction accuracy (>89%) and F1 score (>87%). Key features included changes in resting membrane potential and action potential amplitude, duration and area. Gender differences in drug responses were attributed to lower IK1, INa, and Ito in females.

The right atrium (RA) is critical for cardiac hemodynamics but is often overlooked in clinical diagnostics. This study presents a benchmark framework for RA cavity segmentation from late gadolinium-enhanced magnetic resonance imaging (LGE-MRIs), leveraging a two-stage strategy and a novel 3D deep learning network, RASnet. The architecture addresses challenges in class imbalance and anatomical variability by incorporating multi-path input, multi-scale feature fusion modules, Vision Transformers, context interaction mechanisms, and deep supervision. Evaluated on datasets comprising 354 LGE-MRIs, RASnet achieves SOTA performance with a Dice score of 92.19% on a primary dataset and demonstrates robust generalizability on an independent dataset. The proposed framework establishes a benchmark for RA cavity segmentation, enabling accurate and efficient analysis for cardiac imaging applications. Open-source code (https://github.com/zjinw/RAS) and data (https://zenodo.org/records/15524472) are provided to facilitate further research and clinical adoption.

The sinoatrial node (SAN), the primary pacemaker of the heart, is responsible for the initiation and robust regulation of sinus rhythm. 3D mapping studies of the ex-vivo human heart suggested that the robust regulation of sinus rhythm relies on specialized fibrotically-insulated pacemaker compartments (head, center and tail) with heterogeneous expressions of key ion channels and receptors. They also revealed up to five sinoatrial conduction pathways (SACPs), which electrically connect the SAN with neighboring right atrium (RA). To elucidate the role of these structural-molecular factors in the functional robustness of human SAN, we developed comprehensive biophysical computer models of the SAN based on 3D structural, functional and molecular mapping of ex-vivo human hearts. Our key finding is that the electrical insulation of the SAN except SACPs, the heterogeneous expression of If, INa currents and adenosine A1 receptors (A1R) across SAN pacemaker-conduction compartments are required to experimentally reproduce observed SAN activation patterns and important phenomena such as shifts of the leading pacemaker and preferential SACP. In particular, we found that the insulating border between the SAN and RA, is required for robust SAN function and protection from SAN arrest during adenosine challenge. The heterogeneity in the expression of A1R within the human SAN compartments underlies the direction of pacemaker shift and preferential SACPs in the presence of adenosine. Alterations of INa current and fibrotic remodelling in SACPs can significantly modulate SAN conduction and shift the preferential SACP/exit from SAN. Finally, we show that disease-induced fibrotic remodeling, INa suppression or increased adenosine make the human SAN vulnerable to pacing-induced exit blocks and reentrant arrhythmia. In summary, our computer model recapitulates the structural and functional features of the human SAN and can be a valuable tool for investigating mechanisms of SAN automaticity and conduction as well as SAN arrhythmia mechanisms under different pathophysiological conditions.

A building digital twin (BDT) can maintain an up-to-date digital model reflecting physical world conditions and has become necessary for building applications. Recent studies on the BDT employed the Internet of Things to sense physical-world conditions. Although cameras are one of the most widely used facilities in buildings, their adoption in the BDT remains unexplored. This study proposes a novel computer-vision (CV)-enabled BDT scheme using building information modeling (BIM) taking camera videos as input, which addresses the dimension, coordinate system, and object inconsistencies between BIM and camera videos. First, the proposed BDT scheme detects objects’ locations and rotations jointly using a 2-D object detection network and a 3-D object estimation network. Then, theorem and lemmas are presented to compute the 3-D locations in BCS using detected 2-D locations. Thirdly, both cold-start object matching and run-time object matching schemes are proposed to address the object inconsistency between camera videos and BIM. Finally, experiments were conducted in the real-world environment. The experiment results showed that the proposed BDT scheme maintained average location errors of 0.181 m with distortions preserved and 0.165 m with distortions removed in the manual calibration scenario, 0.166 m with distortions preserved, and 0.195 m with distortions removed in automatic calibration scenario. This finding proved the effectiveness of the proposed BDT scheme. This study is the first to explore a BDT scheme on top of BIM using CV. It is anticipated that this study will inspire more intelligent studies in smart buildings jointly employing both CV and BIM.

Crowdsensing is an emerging paradigm of ubiquitous sensing, through which a crowd of workers are recruited to perform sensing tasks collaboratively. Although it has stimulated many applications, an open fundamental problem is how to select among a massive number of workers to perform a given sensing task under a limited budget. Nevertheless, due to the proliferation of smart devices equipped with various sensors, it is very difficult to profile the workers in terms of sensing ability. Although the uncertainties of the workers can be addressed by conventional Combinatorial Multi-Armed Bandit (CMAB) framework through a trade-off between exploration and exploitation, we do not have sufficient allowance to directly explore and exploit the workers under the limited budget. Furthermore, since the sensor devices usually have quite limited resources, the workers may have bounded capabilities to perform the sensing task only few times, which further restricts our opportunities to learn the uncertainty. To address the above issues, we propose a Context-Aware Worker Selection (CAWS) algorithm in this paper. By leveraging the correlation between the context information of the workers and their sensing abilities, CAWS aims at maximizing the expected cumulative sensing revenue efficiently with both budget constraint and capacity constraints respected, even when the number of the uncertain workers is massive. The efficacy of CAWS can be verified by rigorous theoretical analysis and extensive experiments.

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia and is a major cause of stroke and morbidity. Recent genome-wide association studies have shown that paired-like homeodomain transcription factor 2 (Pitx2) to be strongly associated with AF. However, the mechanisms underlying Pitx2 modulated arrhythmogenesis and variable effectiveness of antiarrhythmic drugs (AADs) in patients in the presence or absence of impaired Pitx2 expression remain unclear. We have developed multi-scale computer models, ranging from a single cell to tissue level, to mimic control and Pitx2-knockout atria by incorporating recent experimental data on Pitx2-induced electrical and structural remodeling in humans, as well as the effects of AADs. The key findings of this study are twofold. We have demonstrated that shortened action potential duration, slow conduction and triggered activity occur due to electrical and structural remodelling under Pitx2 deficiency conditions. Notably, the elevated function of calcium transport ATPase increases sarcoplasmic reticulum Ca2+ concentration, thereby enhancing susceptibility to triggered activity. Furthermore, heterogeneity is further elevated due to Pitx2 deficiency: 1) Electrical heterogeneity between left and right atria increases; and 2) Increased fibrosis and decreased cell-cell coupling due to structural remodelling slow electrical propagation and provide obstacles to attract re-entry, facilitating the initiation of re-entrant circuits. Secondly, our study suggests that flecainide has antiarrhythmic effects on AF due to impaired Pitx2 by preventing spontaneous calcium release and increasing wavelength. Furthermore, our study suggests that Na+ channel effects alone are insufficient to explain the efficacy of flecainide. Our study may provide the mechanisms underlying Pitx2-induced AF and possible explanation behind the AAD effects of flecainide in patients with Pitx2 deficiency.

Pathfinding is a fundamental problem for many areas, e.g., robotics, automation, computer-aided design, and computer graphics. Although outdoor pathfinding is fledged, indoor pathfinding remains a challenge due to the lack of indoor maps. Currently, some efforts have utilized building information modeling (BIM) to generate either the grid-based map or the topological map. However, either the grid-based map or the topological map is not sufficient to provide accurate and efficient pathfinding service. This article proposes a novel grid-topological map and develops an accurate and efficient indoor pathfinding scheme based on BIM. The grid-topological map is modeled jointly adopting the advantages of both the grid-based map and the topological map. First, the grid-based map is generated using the BIM data by extracting and mapping geometric and semantic data into planar grids. Second, a grid thinning algorithm is proposed to produce the topological map directly using the grid-based map. Third, a grid-topological map is presented by combining both the grid-based map and topological map. On top of the grid-topological map, an accurate and efficient pathfinding algorithm is developed. Empirical studies proved the effectiveness of the grid-topological map, as well as the accuracy and efficiency of the proposed indoor pathfinding algorithm.

The integration of Building Information Modelling (BIM) and geographic information systems (GIS) is a promising but challenging topic to solve problems in construction industry. However, loading and rendering rich BIM geometric data and large-scale GIS spatial information in a unified system is still technologically challenging. Current efforts mainly simplify the geometry in BIM models, or convert BIM geometric data to a lower level of detail (LOD). By noticing that only exterior features of BIM models are visible from outdoor observation points, culling BIM interior facilities can dramatically reduce the computational burden when visualizing BIM models in GIS. This study explores the outline detection problem and presents the OutDet algorithm, which selects representative observation points, transforms & projects the BIM geometric data into the same coordinate system, and detects the visible facilities. Empirical study results show that OutDet can cull a large portion of unnecessary features when rendering BIM models in GIS. The use of outlines of BIM models is not an alternative but rather a supplementary approach for current solutions. Jointly using LOD and outer surface can help improve the efficiency of integrated BIM-GIS visualization.Because OutDet retains BIM geometry and semantics, it can be applied to more BIM-GIS applications..

Atrial fibrillation (AF) is the most prevalent form of cardiac arrhythmia. Current treatments for AF remain suboptimal due to a lack of understanding of the underlying atrial structures that directly sustain AF. Existing approaches for analyzing atrial structures in 3-D, especially from late gadolinium-enhanced (LGE) magnetic resonance imaging, rely heavily on manual segmentation methods that are extremely labor-intensive and prone to errors. As a result, a robust and automated method for analyzing atrial structures in 3-D is of high interest. We have, therefore, developed AtriaNet, a 16-layer convolutional neural network (CNN), on 154 3-D LGE-MRIs with a spatial resolution of 0.625 mm × 0.625 mm × 1.25 mm from patients with AF, to automatically segment the left atrial (LA) epicardium and endocardium. AtriaNet consists of a multi-scaled, dualpathway architecture that captures both the local atrial tissue geometry and the global positional information of LA using 13 successive convolutions and three further convolutions for merging. By utilizing computationally efficient batch prediction, AtriaNet was able to successfully process each 3-D LGE-MRI within 1 min. Furthermore, benchmarking experiments have shown that AtriaNet has outperformed the state-of-the-art CNNs, with a DICE score of 0.940 and 0.942 for the LA epicardium and endocardium, respectively, and an inter-patient variance of3of the ground truths. Our proposed CNN was tested on the largest known data set for LA segmentation, and to the best of our knowledge, it is the most robust approach that has ever been developed for segmenting LGE-MRIs. The increased accuracy of atrial reconstruction and analysis could potentially improve the understanding and treatment of AF.

Identification of anonymous identical users of cross-platforms refers to the recognition of the accounts belonging to the same individual among multiple Social Network (SN) platforms. Evidently, cross-platform exploration may help solve many problems in social computing, in both theory and practice. However, it is still an intractable problem due to the fragmentation, inconsistency, and disruption of the accessible information among SNs. Different from the efforts implemented on user profiles and users' content, many studies have noticed the accessibility and reliability of network structure in most of the SNs for addressing this issue. Although substantial achievements have been made, most of the current network structure-based solutions, requiring prior knowledge of some given identified users, are supervised or semi-supervised. It is laborious to label the prior knowledge manually in some scenarios where prior knowledge is hard to obtain. Noticing that friend relationships are reliable and consistent in different SNs, we proposed an unsupervised scheme, termed Friend Relationship-based User Identification algorithm without Prior knowledge (FRUI-P). The FRUI-P first extracts the friend feature of each user in an SN into friend feature vector, and then calculates the similarities of all the candidate identical users between two SNs. Finally, a one-to-one map scheme is developed to identify the users based on the similarities. Moreover, FRUI-P is proved to be efficient theoretically. Results of extensive experiments demonstrated that FRUI-P performs much better than current state-of-art network structure-based algorithm without prior knowledge. Due to its high precision, FRUI-P can additionally be utilized to generate prior knowledge for supervised and semi-supervised schemes. In applications, the unsupervised anonymous identical user identification method accommodates more scenarios where the seed users are unobtainable.

Micro-computed tomography (micro-CT) has been widely used to generate high-resolution 3-D tissue images from small animals nondestructively, especially for mineralized skeletal tissues. However, its application to the analysis of soft cardiovascular tissues has been limited by poor inter-tissue contrast. Recent ex vivo studies have shown that contrast between muscular and connective tissue in micro-CT images can be enhanced by staining with iodine. In the present study, we apply this novel technique for imaging of cardiovascular structures in canine hearts. We optimize the method to obtain high-resolution X-ray micro-CT images of the canine atria and its distinctive regions-including the Bachmann's bundle, atrioventricular node, pulmonary arteries and veins-with clear inter-tissue contrast. The imaging results are used to reconstruct and segment the detailed 3-D geometry of the atria. Structure tensor analysis shows that the arrangement of atrial fibers can also be characterized using the enhanced micro-CT images, as iodine preferentially accumulates within the muscular fibers rather than in connective tissues. This novel technique can be particularly useful in nondestructive imaging of 3-D cardiac architectures from large animals and humans, due to the combination of relatively high speed ( ~ 1 h/per scan of the large canine heart) and high voxel resolution (36 μm) provided. In summary, contrast micro-CT facilitates fast and nondestructive imaging and segmenting of detailed 3-D cardiovascular geometries, as well as measuring fiber orientation, which are crucial in constructing biophysically detailed computational cardiac models.

Computer models provide a powerful platform for investigating mechanisms that underlie atrial rhythm disturbances. We have used novel techniques to build a structurally-detailed, image-based model of 3-D atrial anatomy. A volume image of the atria from a normal sheep heart was acquired using serial surface macroscopy, then smoothed and down-sampled to 50 μm(3) resolution. Atrial surface geometry was identified and myofiber orientations were estimated throughout by eigen-analysis of the 3-D image structure tensor. Sinus node, crista terminalis, pectinate muscle, Bachman's bundle, and pulmonary veins were segmented on the basis of anatomic characteristics. Heterogeneous electrical properties were assigned to this structure and electrical activation was simulated on it at 100 μm(3) resolution, using both biophysically-detailed and reduced-order cell activation models with spatially-varying membrane kinetics. We confirmed that the model reproduced key features of the normal spread of atrial activation. Furthermore, we demonstrate that vulnerability to rhythm disturbance caused by structural heterogeneity in the posterior left atrium is exacerbated by spatial variation of repolarization kinetics across this region. These results provide insight into mechanisms that may sustain paroxysmal atrial fibrillation. We conclude that image-based computer models that incorporate realistic descriptions of atrial myofiber architecture and electrophysiologic properties have the potential to analyse and identify complex substrates for atrial fibrillation.