Results
- Showing results for:
- Reset all filters
Search results
-
Journal articleHasan MK, Yang G, Yap CH, 2026,
An efficient, scalable, and adaptable plug-and-play temporal attention module for motion-guided cardiac segmentation with sparse temporal labels
, Medical Image Analysis, Vol: 110, ISSN: 1361-8415Cardiac anatomy segmentation is essential for clinical assessment of cardiac function and disease diagnosis to inform treatment and intervention. Deep learning (DL) has improved cardiac anatomy segmentation accuracy, especially when information on cardiac motion dynamics is integrated into the networks. Several methods for incorporating motion information have been proposed; however, existing methods are not yet optimal: adding the time dimension to input data causes high computational costs, and incorporating registration into the segmentation network remains computationally costly and can be affected by errors of registration, especially with non-DL registration. While attention-based motion modeling is promising, suboptimal design constrains its capacity to learn the complex and coherent temporal interactions inherent in cardiac image sequences. Here, we propose a novel approach to incorporating motion information in the DL segmentation networks: a computationally efficient yet robust Temporal Attention Module (TAM), modeled as a small, multi-headed, cross-temporal attention module, which can be plug-and-play inserted into a broad range of segmentation networks (CNN, transformer, or hybrid) without a drastic architecture modification. Extensive experiments on multiple cardiac imaging datasets, such as 2D echocardiography (CAMUS and EchoNet-Dynamic), 3D echocardiography (MITEA), and 3D cardiac MRI (ACDC), confirm that TAM consistently improves segmentation performance across datasets when added to a range of networks, including UNet, FCN8s, UNetR, SwinUNetR, and the recent I<sup>2</sup>UNet and DT-VNet. Integrating TAM into SAM yields a temporal SAM that reduces Hausdorff distance (HD) from 3.99 mm to 3.51 mm on the CAMUS dataset, while integrating TAM into a pre-trained MedSAM reduces HD from 3.04 to 2.06 pixels after fine-tuning on the EchoNet-Dynamic dataset. On the ACDC 3D dataset, our TAM-UNet and TAM-DT-VNet achieve substantial reductions in HD
-
Journal articleZhou T, Li M, Ruan S, et al., 2026,
A reliable framework for brain tumor segmentation via multi-modal fusion and uncertainty modeling
, Information Fusion, Vol: 129, ISSN: 1566-2535Accurate brain tumor segmentation from MRI scans is critical for effective diagnosis and treatment planning. Recent advances in deep learning have significantly improved brain tumor segmentation performance. However, these models still face challenges in clinical adoption due to their inherent uncertainties and potential for errors. In this paper, we propose a novel MR brain tumor segmentation approach that integrates multi-modal data fusion and uncertainty quantification to improve the accuracy and reliability of brain tumor segmentation. Recognizing that each MR modality contributes unique insights into the tumor’s characteristics, we propose a novel modality-aware guidance by explicitly categorizing the modalities into ”teacher” (FLAIR and T1c) and ”student” (T2 and T1) groups. Since the teacher modalities are the most informative modalities for identifying brain tumors, we propose a multi-modal teacher-student fusion strategy. This strategy leverages the teacher modalities to guide the student modalities in both spatial and channel feature representation aspects. To address prediction reliability, we employ Monte Carlo dropout during training to generate multiple uncertainty estimates. Additionally, we develop a novel uncertainty-aware loss function that optimizes segmentation accuracy while quantifying the uncertainty in predictions. Experimental results conducted on three BraTS datasets demonstrate the effectiveness of the proposed components and the superior performance compared to the state-of-the-art methods, highlighting their potential for clinical application.
-
Journal articleKonstantinou K, Koumelli A, Apostolos A, et al., 2026,
Short-term and mid-term blood pressure variability in acute myocardial infarction: a prospective cohort study on in-hospital and long-term prognostic impact.
, J Hypertens, Vol: 44, Pages: 662-672INTRODUCTION: Blood pressure variability (BPV) is a prognostic marker in hypertension and coronary artery disease (CAD), but its role in acute myocardial infarction (AMI) remains unknown. This study assessed the association of short-term (24-h ambulatory BP monitoring, ABPM) and mid-term BPV with adverse in-hospital and long-term outcomes in AMI patients. METHODS: Mid-term BPV was calculated as the standard deviation (SD) of daily in-hospital BP readings; short-term BPV was measured by average real variability (ARV) from ABPM. Patients were evaluated as continuous variables and by quartiles (Q1-Q4). Logistic regression and Cox models assessed in-hospital and 3-year outcomes. RESULTS: In this prospective, single-center cohort, 441 of 677 AMI patients were included. Each 1 mmHg rise in day-to-day systolic BPV (SBP-SD) increased in-hospital MACE risk by 24% [odds ratio (OR): 1.24, 95% confidence interval (CI): 1.17-1.31], with Q4 showing the highest risk (OR: 28.89, 95% CI: 8.58-97.28). ABPM-derived SBP-ARV predicted in-hospital mortality (OR: 1.58, 95% CI: 1.21-2.07) and MACE (OR: 1.35, 95% CI: 1.23-1.48). Diastolic ARV was linked to in-hospital myocardial infarction (MI), arrhythmias, and shock. At 3-year follow up, Q4 of SBP-SD showed higher risk of composite outcomes (hazard ratio: 29.88, 95% CI: 10.93-81.66) and all-cause mortality (hazard ratio: 11.85, 95% CI: 2.81-49.91). SBP-ARV independently predicted both all-cause mortality (hazard ratio: 1.37, 95% CI: 1.25-1.51) and adverse events (hazard ratio: 1.29, 95% CI: 1.22-1.36), while diastolic BPV was primarily associated with arrhythmias and heart failure hospitalization. CONCLUSION: Systolic BPV independently predicts in-hospital and long-term outcomes in AMI. BPV assessment may aid post-MI risk stratification and guide novel therapeutic strategies in this high-risk population.
-
Journal articleLiao Y, Zheng Y, Zhu J, et al., 2026,
Self-attention-based mixture-of-experts framework for non-invasive prediction of MGMT promoter methylation in glioblastoma using multi-modal MRI
, Displays, Vol: 92, ISSN: 0141-9382Glioblastoma (GBM) is an aggressive brain tumor associated with poor prognosis and limited treatment options. The methylation status of the O6-methylguanine-DNA methyltransferase (MGMT) promoter is a critical biomarker for predicting the efficacy of temozolomide chemotherapy in GBM patients. However, current methods for determining MGMT promoter methylation, including invasive and costly techniques, hinder their widespread clinical application. In this study, we propose a novel non-invasive deep learning framework based on a Mixture-of-Experts (MoE) architecture for predicting MGMT promoter methylation status using multi-modal magnetic resonance imaging (MRI) data. Our MoE model incorporates modality-specific expert networks built on the ResNet18 architecture, with a self-attention-based gating mechanism that dynamically selects and integrates the most relevant features across MRI modalities (T1-weighted, contrast-enhanced T1, T2-weighted, and fluid-attenuated inversion recovery). We evaluate the proposed framework on the BraTS2021 and TCGA-GBM datasets, showing superior performance compared to conventional deep learning models in terms of accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC). Furthermore, Grad-CAM visualizations provide enhanced interpretability by highlighting biologically relevant regions in the tumor and peritumoral areas that influence model predictions. The proposed framework represents a promising tool for integrating imaging biomarkers into precision oncology workflows, offering a scalable, cost-effective, and interpretable solution for non-invasive MGMT methylation prediction in GBM.
-
Journal articleCheng CW, Huang J, Zhang Y, et al., 2026,
Mamba neural operator: Who wins? transformers vs. state-space models for PDEs
, Journal of Computational Physics, Vol: 548, ISSN: 0021-9991Partial differential equations (PDEs) are widely used to model complex physical systems, but solving them efficiently remains a significant challenge. Recently, Transformers have emerged as the preferred architecture for PDEs due to their ability to capture intricate dependencies. However, they struggle with representing continuous dynamics and long-range interactions. To overcome these limitations, we introduce the Mamba Neural Operator (MNO), a novel framework that enhances neural operator-based techniques for solving PDEs. MNO establishes a formal theoretical connection between structured state-space models (SSMs) and neural operators, offering a unified structure that can adapt to diverse architectures, including Transformer-based models. By leveraging the structured design of SSMs, MNO captures long-range dependencies and continuous dynamics more effectively than traditional Transformers. Through extensive analysis, we show that MNO significantly boosts the expressive power and accuracy of neural operators, making it not just a complement but a superior framework for PDE-related tasks, bridging the gap between efficient representation and accurate solution approximation.
-
Journal articleJing P, Lee K, Zhang Z, et al., 2026,
Reason like a radiologist: Chain-of-thought and reinforcement learning for verifiable report
, MEDICAL IMAGE ANALYSIS, Vol: 109, ISSN: 1361-8415 -
Journal articleMa X, Tao Y, Zhang Z, et al., 2026,
Test-time generative augmentation for medical image segmentation
, MEDICAL IMAGE ANALYSIS, Vol: 109, ISSN: 1361-8415 -
Journal articleHatipoglu S, Voges I, Pushparajah K, et al., 2026,
Stress imaging in paediatric and congenital heart disease patients.
, Eur Heart J Cardiovasc Imaging, Vol: 27, Pages: 567-581Stress imaging in paediatric cardiology and congenital heart disease patients has an increasing role for functional assessment. Indications include coronary artery anomalies and disease in association with anomalous aortic origin of coronary arteries, Kawasaki disease or surgical manipulation of the coronary ostia, as well as assessment of elevated filling pressures, dynamic left ventricular outflow obstruction or significance of valvular heart disease. This review provides practical guidance focused on commonly used stress echocardiography and stress cardiovascular magnetic resonance in context of their clinical indications for this age group.
-
Journal articleLuo Y, Sesia D, Wang F, et al., 2026,
Explicit differentiable slicing and global deformation for cardiac mesh reconstruction.
, Med Image Anal, Vol: 111Three-dimensional (3D) mesh reconstruction of the cardiac anatomy from medical images is useful for shape and motion measurements and biophysics simulations. However, 3D medical images are often acquired as 2D slices that are sparsely sampled (e.g., large slice spacing) and noisy, and 3D mesh reconstruction on such data is a challenging task. Traditional voxel-based approaches utilize non-differentiable pre- and post-processing that compromises fidelity to images, while mesh-level deep learning approaches require large 3D mesh annotations that are difficult to obtain. Differentiable cross-domain supervision from 2D images to 3D meshes is therefore crucial for enabling end-to-end optimization in medical imaging. While there have been attempts to approximate the voxelization and slicing of meshes that are being optimized, there has not yet been a method for directly using 2D slices to supervise 3D mesh reconstruction in a differentiable manner. Here, we propose a novel explicit differentiable voxelization and slicing (DVS) algorithm allowing gradient backpropagation to a 3D mesh from its slices, which facilitates refined mesh optimization directly supervised by the losses defined on 2D images. Further, we propose an innovative framework for extracting patient-specific left ventricle (LV) meshes from medical images by coupling DVS with a graph harmonic deformation (GHD) mesh morphing descriptor of cardiac shape that naturally preserves mesh quality and smoothness during optimization. The proposed framework achieves state-of-the-art performance in cardiac mesh reconstruction tasks from densely sampled (CT) as well as sparsely sampled (MRI stack with few slices) images, outperforming alternatives, including Marching Cubes, statistical shape models, algorithms with vertex-based mesh morphing algorithms and alternative methods for image-supervision of mesh reconstruction. Experimental results demonstrate that our method achieves an overall Dice score of 90% during a sparse
-
Journal articleBenfield N, Thami PK, Ware J, et al., 2026,
GWAS reveals common SLX4 variants associated with telomere length and hypertension in individuals of African ancestry.
, Genes Genomics -
Journal articleLi K, Xiao X, Zhong Z, et al., 2026,
Accurate and generalizable protein-ligand binding affinity prediction with geometric deep learning
, IEEE Open Journal of Engineering in Medicine and Biology, ISSN: 2644-1276Goal: Protein-ligand binding complexes are ubiquitous and essential to life. Protein-ligand binding affinity prediction (PLA) quantifies the binding strength between ligands and proteins, providing crucial insights for discovering and designing potential candidate ligands. While recent advances have been made in predicting protein-ligand complex structures, existing algorithms for interaction and affinity prediction suffer from a sharp decline in performance when handling ligands bound with novel unseen proteins. Methods: We propose IPBind, a geometric deep learning-based computational method, enabling robust predictions by leveraging interatomic potential between complex’s bound and unbound status. Results: Experimental results on widely used binding affinity prediction benchmarks demonstrate the effectiveness and universality of IPBind. Meanwhile, it provids atom-level insights into prediction. Conclusions: This work highlight the advantage of leveraging machine learning interatomic potential for predicting protein-ligand binding affinity. Index Terms—Deep learning, drug discovery, physics-informed neural networks, protein-ligand binding affinity prediction. Impact Statement–This study extends state-of-the-art deep learning algorithms to applications in protein-ligand binding affinity prediction. This study has implications for enhancing the generalization capability of protein-ligand interactions prediction methods by interatomic potential modeling.
-
Journal articleStroeks SLVM, Oko-Osi S, Arasu A, et al., 2026,
Sex differences in dilated cardiomyopathy: evidence gaps and future directions
, JACC, Vol: 87, Pages: 723-735, ISSN: 0735-1097Dilated cardiomyopathy (DCM), which affects 1 in 250 people, is a leading global cause of heart failure and the most common indication for heart transplantation. Evidence suggests that DCM is more prevalent in men, but whether this reflects biological differences or underdiagnosis in women remains uncertain. This review explores the impact of sex on DCM, examining differences in epidemiology, etiology, clinical presentation, treatment response, and outcomes. Women often present with less severe cardiac phenotypes, including lower levels of fibrosis and better left ventricular function, yet the long-term prognosis of DCM in women is less clear. Through a systematic review and meta-analysis, we found that male DCM patients with variants in PLN, DSP, and LMNA had higher arrhythmic event rates compared with TTNtv and BAG3 carriers. In female patients with DCM, those with RBM20, DSP, and PLN variants faced the highest arrhythmic risk, and TTNtv carriers the lowest. PLN and LMNA variants had the highest heart failure risk in both sexes, whereas BAG3, RBM20, and TTN variants had lower heart failure rates in female compared with male carriers. These findings highlight the influence of sex and genotype on clinical outcomes. Current risk-stratification tools, such as those used for implantable cardioverter-defibrillators, may undertreat women owing to reliance on sex-neutral thresholds. We highlight the role of genetic, environmental, and reproductive factors in shaping these disparities, including the influence of pregnancy, pregnancy complications, and menopause. This review identifies key gaps in knowledge and calls for expanded representation of women in DCM studies and the development of sex-specific risk models. Addressing these gaps is essential to improving outcomes and advancing equitable personalized care for all DCM patients.
-
Journal articleda Rocha GL, Feiner J, Lazarte J, et al., 2026,
Cardiomyopathy Gene Variants and Polygenic Risk Scores in Atrial Fibrillation: Evidence for an Atrial-First Phenotype.
, J Am Coll CardiolBACKGROUND: Atrial fibrillation (AF) is heritable and its complex underlying genetic substrate is gradually being unraveled. OBJECTIVES: We sought to explore the impact of disease-causing cardiomyopathy variants on the risk of AF after adjustment for incident ventricular cardiomyopathy and clinical heart failure in 2 cohort studies (UK Biobank [UKB] and All of Us [AoU]) and evaluate the utility of polygenic risk scores (PRS) to further discern the risk of atrial and ventricular phenotypes in carriers. METHODS: Cox regression was used to evaluate for associations between disease-causing variants within genes for 3 cardiomyopathies (dilated cardiomyopathy [DCM], hypertrophic cardiomyopathy [HCM], and arrhythmogenic right ventricular cardiomyopathy) and AF. Disease-specific PRSs for AF, DCM, and HCM stratified study participants into quintiles. A HR random-effects meta-analysis was performed using the DerSimonian-Laird method. The Kaplan-Meier method was used to ascertain cumulative incidence from birth to 75 years of age. RESULTS: Among 655,796 individuals from UKB and AoU, presence of a disease-causing variant was associated with an increased AF hazard (HR: 1.73; 95% CI: 1.59-1.89; P < 0.001), including after adjustment for incident ventricular cardiomyopathy or clinical heart failure (adjusted to HR: 1.55; 95% CI: 1.46-1.64, P < 0.001). The cumulative AF risk for study participants with a putative disease-causing rare variant and a PRSAF within the top-risk quintile ranged from 32.5% (UKB) to 32.4% (AoU) relative to 9.8% (UKB) and 11.0% (AoU) for individuals without a putative disease-causing variant and a PRSAF within the lowest-risk quintile. The absolute cumulative cardiomyopathy risk among study participants with both a putative disease-causing variant and a disease-specific PRS within the top-risk quintile ranged from 5.9% (UKB) to 15.2% (AoU) for DCM and from 11.7% (UKB) to 19.1% (AoU) for HCM. CONCLUSIONS: Genetic variants that cause cardiomyopathy also
-
Journal articlede Villiers C, Ormondroyd E, Thomson K, et al., 2026,
Hypertrophic cardiomyopathy caused by filamin-C variants has restrictive and extracardiac features and a distinctive ECG.
, Heart RhythmBACKGROUND: Filamin-C (FLNC) gene variants are associated with cardiac and skeletal muscle diseases including a clear role of loss-of-function variants in dilated cardiomyopathy. OBJECTIVE: This study aimed to assess the contribution of rare FLNC variants to hypertrophic cardiomyopathy (HCM)/restrictive cardiomyopathy (RCM). METHODS: Family-based studies in 2 specialist services and statistical modeling of rare FLNC missense variants were conducted, using a cohort of 3289 sarcomere-negative HCM cases and 122,348 genome aggregation database controls. RESULTS: Clinical evaluation of patients with HCM/RCM and a rare FLNC variant identified a distinct electrocardiographic (ECG) repolarization phenotype in 37% (19 of 51 individuals, from 12 families), which was observed in only 1.0% of a control HCM cohort (2 of 197). FLNC variant carriers with the characteristic ECG had smaller left ventricular cavity size, lower contractility, and more severe diastolic dysfunction and were more likely to have a restrictive phenotype. Heart failure death, transplant, or cardiac arrest occurred in at least 1 individual in 7 of the 12 families (58%) in the "ECG-positive" group, and musculoskeletal abnormalities were present in 4 families (33%). 5 of 12 variants (41.7%) in the "ECG-positive" group cosegregated, and 2 were apparently de novo. 11 variants were missense, and 1 splice site. Rare FLNC missense variant burden indicated a low case excess among all HCM cases (etiologic fraction, 0.45; 95% confidence interval, 0.36-0.54), but in "ECG-positive" cases the etiologic fraction was substantially higher (0.98; 95% confidence interval, 0.97-0.99). CONCLUSION: Pathogenic FLNC variants in patients with HCM/RCM are nontruncating and cause a discrete phenotype comprising a characteristic ECG, hypertrophic and restrictive features without hypercontractility, and extracardiac abnormalities.
-
Journal articleKhalique Z, Scott AD, Ferreira PF, et al., 2026,
Diffusion Tensor CMR Assessment of the Microstructural Response to Dobutamine Stress in Health and Comparison With Patients With Recovered Dilated Cardiomyopathy.
, Circ Cardiovasc Imaging, Vol: 19BACKGROUND: Contractile reserve assessment assesses myocardial performance and prognosis. The microstructural mechanisms that facilitate increased cardiac function have not been described, but can be studied using diffusion tensor cardiovascular magnetic resonance. Resting microstructural contractile function is characterized by reorientation of aggregated cardiomyocytes (sheetlets) from wall-parallel in diastole to a more wall-perpendicular configuration in systole, with the diffusion tensor cardiovascular magnetic resonance parameter E2A defining their orientation, and sheetlet mobility defining the angle through which they rotate. We used diffusion tensor cardiovascular magnetic resonance to identify the microstructural response to dobutamine stress in healthy volunteers and then compared with patients with recovered dilated cardiomyopathy (rDCM). METHODS: In this first-of-its-kind prospective observational study, 20 healthy volunteers and 32 patients with rDCM underwent diffusion tensor cardiovascular magnetic resonance at rest, during dobutamine, and on recovery. RESULTS: In healthy volunteers, both diastolic and systolic E2A increased with dobutamine stress (13±3° to 17±5°; P<0.001 and 59±11° to 65±7°; P=0.002). Sheetlet mobility remained unchanged (45±11° to 49±10°; P=0.19), but biphasic mean E2A increased (36±6° to 41±4°; P<0.001). In rDCM, diastolic E2A at rest was higher than in healthy volunteers (20±8° versus 13±3°, P<0.001), and sheetlet mobility was reduced (34±12° versus 45±11°; P<0.001). During dobutamine stress, rDCM diastolic and systolic E2A increased compared with rest (20±8° to 24±10°; P=0.001 and 54±13° to 63±11°; P=0.005). However, sheetlet mobility in patients with rDCM failed to increase with dobutamine to healthy levels (39±13° versus 49±
-
Journal articleZhang S, Nan Y, Fang Y, et al., 2026,
Dynamical multi-order responses and global semantic-infused adversarial learning: A robust airway segmentation method
, MEDICAL IMAGE ANALYSIS, Vol: 108, ISSN: 1361-8415 -
Journal articleMa J, Jiang M, Fang X, et al., 2026,
Hybrid aggregation strategy with double inverted residual blocks for lightweight salient object detection
, NEURAL NETWORKS, Vol: 194, ISSN: 0893-6080 -
Journal articleArdissino M, Morley AP, Truong B, et al., 2026,
Genetic Association of Circulating Proteins and Gene Transcripts With Spontaneous Coronary Artery Dissection.
, Circ Genom Precis Med, Vol: 19BACKGROUND: Spontaneous coronary artery dissection (SCAD) is an uncommon cause of myocardial infarction that disproportionately affects women, particularly during pregnancy and the peripartum period. Limited understanding of its underlying pathophysiology hinders the development of effective preventive and therapeutic strategies. METHODS: This study investigated associations between genetically predicted circulating proteins and tissue-specific RNA levels with genetically predicted SCAD risk using Mendelian randomization and Bayesian colocalization. Genetic scores for >1500 circulating proteins were derived from the UK Biobank (N=34 557) and deCODE (N=35 559). Scores for 13 848 gene transcripts in arterial and fibroblast tissues were generated from Genotype-Tissue Expression data. Associations between these scores and SCAD were assessed in a genome-wide association study meta-analysis of 1917 individuals with SCAD and 9292 controls. Findings were validated in vitro using mass spectrometry-based proteomic analysis of extracellular vesicles from 50 patients with SCAD and 50 healthy controls. RESULTS: Genetic associations of 4 circulating proteins with SCAD (AFAP1 [actin filament-associated protein 1], ECM1 [extracellular matrix protein 1], SPON1 [spondin 1], and STAT6 [signal transducer and activator of transcription 6]) were identified. Two were supported by gene expression data (AFAP1 and ECM1), and one by tissue-specific Bayesian colocalization analyses (ECM1). Protein interaction mapping identified potential shared pathways through the JAK-STAT (Janus kinases and signal transducers and activators of transcription) signaling pathway and inflammatory regulation. Mass spectrometry-based proteomic analysis demonstrated that ECM1 was significantly upregulated in SCAD cases versus controls. CONCLUSIONS: Integrative analysis of proteomic, transcriptomic, and experimental data revealed 4 circulating proteins genetically associated with SCAD risk, w
-
Journal articleKramarenko DR, Haydarlou P, Powell GJ, et al., 2026,
Leveraging the shared and opposing genetic mechanisms in the heritable cardiomyopathies.
, Res SqDilated cardiomyopathy (DCM) and hypertrophic cardiomyopathy (HCM) are heart muscle diseases with largely opposing structural and functional phenotypes. Yet, both may lead to the same devastating outcomes of advanced heart failure and life-threatening arrhythmias. Using genome-wide association data from 9,365 DCM cases, 5,900 HCM cases, and over 1.2 million controls, we show that DCM and HCM are largely inversely associated across multiple genomic levels. Modeling both disorders as opposing genetic entities, in case-case GWAS approaches, we identify 100 loci (17 novel) underlying the cardiomyopathy spectrum. Several loci map to potential therapeutic targets (e.g., ADM, CACNA2D2), and polygenic risk scores derived from these data show strong discrimination between DCM and HCM patients in external datasets (AUC 0.78-0.84; AUPRC ~ 0.85). The pervasive opposing associations suggest that cardiomyocyte-directed therapies may often have opposite effects in DCM versus HCM. Nevertheless, a shared-effect analysis reveals a single locus - near the calcium-buffering gene CASQ2 - and also identifies a concordant genomic component associated with cardiometabolic health and extracardiac risk factors. By leveraging the shared and opposing genetic mechanisms of DCM and HCM, our work defines the genomic architecture of major cardiomyopathy subtypes and suggests new directions for therapeutics and precision medicine in heart failure.
-
Journal articleRadia UK, Ahluwalia V, Rosen SD, 2026,
Cardio-Oncology: At the Nexus of Two 21st Century Epidemics.
, Br J Hosp Med (Lond), Vol: 87, ISSN: 1750-8460 -
Journal articleMohal JS, Whinnett ZI, Mohiddin SA, et al., 2026,
Electromechanically Optimized Right Ventricular Pacing for Obstructive Hypertrophic Cardiomyopathy: The EMORI-HCM Trial.
, J Am Coll Cardiol, Vol: 87, Pages: 124-139BACKGROUND: Many patients with symptomatic obstructive hypertrophic cardiomyopathy (oHCM) have devices capable of right ventricular pacing (RVP). Although pacing can reduce left ventricular outflow tract gradient (LVOTg), it can also reduce cardiac output, so its net effect is variable. OBJECTIVES: We tested whether electromechanical optimization of the programmed atrio-ventricular delay (AVD) allows RVP to achieve a net benefit on symptoms. METHODS: EMORI-HCM (Electromechanically Optimized Right Ventricular Pacing in Obstructive Hypertrophic Cardiomyopathy) is a multicenter, blinded, randomized, crossover trial of AVD-optimized RVP in patients with symptomatic oHCM with resting or provoked gradient of at least 30 mm Hg. Patients with existing dual-chamber devices were randomized to either 3 months of continuous AVD-optimized RVP (intervention) followed by 3 months of backup-only RVP (control), or vice versa. AVD was optimized using a high-precision multiple-alternation protocol assessing acute change in beat-by-beat blood pressure while varying AVD. The primary outcome was symptoms measured by the Kansas City Cardiomyopathy Questionnaire Clinical Summary Score. Secondary outcomes include patient-reported daily symptom data collected using a dedicated smartphone application (ORBITA-app), dichotomous patient preference, EQ-5D, exercise capacity, and LVOTg. Patients were blinded to treatment allocation. Symptom assessments were self-administered. Outcome measures were recorded at baseline, crossover, and completion. Analysis was by Bayesian ordinal mixed modeling. RESULTS: Between October 2021 and October 2024, 117 screened patients met the inclusion criteria, of whom 60 were randomized. AVD-optimized RVP improved Kansas City Cardiomyopathy Questionnaire Clinical Summary Score (+4.5; 95% credible interval [CrI]: 1.3-8.1; probability of benefit [Prbenefit] = 0.997) and daily symptom scores (OR: 1.29; 95% CrI: 0.98-1.68; Prbenefit: 0.969) compared with backup-only pacin
-
Conference paperHasan MK, Yang G, Yap CH, 2026,
Motion-Enhanced Cardiac Anatomy Segmentation via an Insertable Temporal Attention Module
, Pages: 143-153, ISSN: 0302-9743Cardiac anatomy segmentation is useful for clinical assessment of cardiac morphology to inform diagnosis and intervention. Deep learning (DL), especially with motion information, has improved segmentation accuracy. However, existing techniques for motion enhancement are not yet optimal, and they have high computational costs due to increased dimensionality or reduced robustness due to suboptimal approaches that use non-DL motion registration, non-attention models, or single-headed attention. They further have limited adaptability and are inconvenient for incorporation into existing networks where motion awareness is desired. Here, we propose a novel, computationally efficient Temporal Attention Module (TAM) that offers robust motion enhancement, modeled as a small, multi-headed, cross-temporal attention module. TAM’s uniqueness is that it is a lightweight, plug-and-play module that can be inserted into a broad range of segmentation networks (CNN-based, Transformer-based, or hybrid) for motion enhancement without requiring substantial changes in the network’s backbone. This feature enables high adaptability and ease of integration for enhancing both existing and future networks. Extensive experiments on multiple 2D and 3D cardiac ultrasound and MRI datasets confirm that TAM consistently improves segmentation across a range of networks while maintaining computational efficiency and improving on currently reported performance. The evidence demonstrates that it is a robust, generalizable solution for motion-awareness enhancement that is scalable (such as from 2D to 3D). The code is available at https://github.com/kamruleee51/TAM.
-
Conference paperZhang H, Huang J, Wu Y, et al., 2026,
Lightweight Hypercomplex MRI Reconstruction: A Generalized Kronecker-Parameterized Approach
, Pages: 95-105, ISSN: 0302-9743Magnetic Resonance Imaging (MRI) is crucial for clinical diagnostics but is hindered by prolonged scan times. Current deep learning models enhance MRI reconstruction but are often memory-intensive and unsuitable for resource-limited systems. This paper introduces a lightweight MRI reconstruction model leveraging Kronecker-Parameterized Hypercomplex Neural Networks to achieve high performance with reduced parameters. By integrating Kronecker-based modules, including Kronecker MLP, Kronecker Window Attention, and Kronecker Convolution, the proposed model efficiently extracts spatial features while preserving representational power. We introduce Kronecker U-Net and Kronecker SwinMR, which maintain high reconstruction quality with approximately 50% fewer parameters compared to existing models. Experimental evaluation on the FastMRI dataset demonstrates competitive PSNR, SSIM, and LPIPS metrics, even at high acceleration factors (8× and 16×), with no significant performance drop. Additionally, Kronecker variants exhibit superior generalization and reduced overfitting on limited datasets, facilitating efficient MRI reconstruction on hardware-constrained systems. This approach sets a new benchmark for parameter-efficient medical imaging models. Code is available at:https://github.com/Whethe/HyperKron-MRI-Recon.
-
Journal articleYu Q, Zhang C, Jin G, et al., 2026,
StealthMark: Harmless and Stealthy Ownership Verification for Medical Segmentation via Uncertainty-Guided Backdoors.
, IEEE Trans Image Process, Vol: 35, Pages: 1290-1304Annotating medical data for training AI models is often costly and limited due to the shortage of specialists with relevant clinical expertise. This challenge is further compounded by privacy and ethical concerns associated with sensitive patient information. As a result, well-trained medical segmentation models on private datasets constitute valuable intellectual property requiring robust protection mechanisms. Existing model protection techniques primarily focus on classification and generative tasks, while segmentation models-crucial to medical image analysis-remain largely underexplored. In this paper, we propose a novel, stealthy, and harmless method, StealthMark, for verifying the ownership of medical segmentation models under closed-box conditions. Our approach subtly modulates model uncertainty without altering the final segmentation outputs, thereby preserving the model's performance. To enable ownership verification, we incorporate model-agnostic explanation methods, e.g. LIME, to extract feature attributions from the model outputs. Under specific triggering conditions, these explanations reveal a distinct and verifiable watermark. We further design the watermark as a QR code to facilitate robust and recognizable ownership claims. We conducted extensive experiments across four medical imaging datasets (CMR dataset from UK Biobank, the SEG fundus dataset, the EchoNet echocardiography dataset, and the PraNet colonoscopy dataset) and five mainstream segmentation models. The results demonstrate the effectiveness, stealthiness, and harmlessness of our method on the original model's segmentation performance. For example, when applied to the SAM model, StealthMark consistently achieved attack success rates (ASR) above 95% across various datasets while maintaining less than a 1% drop in Dice and AUC scores-significantly outperforming backdoor-based watermarking methods and highlighting its strong potential for practical deployment. Our implementation code is made
-
Journal articleRjoob K, McGurk K, Zheng S, et al., 2025,
A multi-modal vision knowledge graph of cardiovascular disease
, Nature Cardiovascular Research, ISSN: 2731-0590Understanding gene-disease associations is important for uncovering pathological mechanisms and identifying potential therapeutic targets. Knowledge graphs can represent and integrate data from multiplebiomedical sources, but lack individual-level information on target organ structure and function. Here wedevelop CardioKG, a knowledge graph that integrates over 200,000 computer vision-derived cardiovascular phenotypes from biomedical images with data extracted from 18 biological databases to model overa million relationships. We used a variational graph auto-encoder to generate node embeddings from theknowledge graph to predict gene-disease associations, assess druggability and identify drug repurposing strategies. The model predicted genetic associations and therapeutic opportunities for leading causesof cardiovascular disease, which were associated with improved survival. Candidate therapies includedmethotrexate for heart failure and gliptins for atrial fibrillation, and the addition of imaging data enhancedpathway discovery. These capabilities support the use of biomedical imaging to enhance graph-structuredmodels for identifying treatable disease mechanisms.
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.