Publications by Year: 2026

Artificial intelligence (AI) is rapidly reshaping cardiovascular (CV) medicine, driving a paradigm shift toward truly personalized and data-driven care. This comprehensive review examines the conceptual foundations, clinical applications, and future implications of AI across the CV continuum, spanning prevention, diagnosis, risk stratification, and therapy. Core AI methodologies (including machine learning, deep learning, natural language processing, and computer vision) are discussed in the context of cardiology's uniquely data-rich environment, encompassing imaging, electrocardiography, electronic health records, wearable devices, and multi-omics data. This systematic review highlights major clinical domains where AI has demonstrated a substantial impact, including CV imaging, ECG interpretation, hypertension and heart failure management, coronary artery disease, acute coronary syndromes, interventional cardiology, and cardiac surgery. AI-driven predictive analytics enable early detection of subclinical disease, improved prognostication, and individualized prevention strategies, while wearable technologies and remote monitoring platforms facilitate continuous, real-world patient surveillance. Emerging applications in pharmacotherapy, drug repurposing, and genomics further reinforce AI's role in advancing precision cardiology. Equally emphasized are the ethical, legal, and social challenges accompanying AI adoption, such as algorithmic bias, data privacy, cybersecurity, interpretability, and regulatory oversight. Our review underscores the necessity of rigorous clinical validation, transparent model design, and seamless integration into clinical workflows to ensure safety, equity, and physician trust. Ultimately, AI is best positioned as an augmentative tool that complements (but does not replace!) clinical expertise. By fostering hybrid intelligence that integrates human judgment with computational power, AI has the potential to redefine CV care delivery, improve outcomes, and support a more proactive, patient-centered healthcare model.

The intricate interplay between the brain and heart underpins both physiological regulation and pathophysiological processes, yet decoding these interactions remains a formidable challenge. Recent advances in artificial intelligence (AI) offer transformative opportunities to map, model, and predict brain-heart network dynamics with unprecedented precision. This review synthesizes current knowledge on AI approaches applied to neurocardiology, encompassing multimodal data integration from neuroimaging, electrophysiology, autonomic signals, and cardiovascular monitoring. We examine machine learning and deep learning strategies for identifying biomarkers, forecasting adverse cardiac events, and elucidating mechanisms linking neurological, psychiatric, and cardiovascular disorders. Clinical applications are explored across heart failure, arrhythmias, stroke-induced cardiac dysfunction, epilepsy, and stress-related conditions, highlighting AI's potential for personalized risk stratification. The role of wearable devices, digital phenotyping, and real-world data collection in continuous brain-heart monitoring is discussed, alongside AI-enabled early warning systems. Critical considerations regarding data quality, bias, interpretability, privacy, and ethical governance are emphasized to guide responsible deployment. Finally, we outline emerging directions, including integrative digital twins, federated AI, and closed-loop neuromodulation. By bridging computational innovation and clinical neuroscience, AI-driven approaches promise to redefine neurocardiology, offering predictive, mechanistic, and therapeutic insights into the brain-heart axis.

UNLABELLED: Phosphoglucomutase 1 (PGM1) is a type 1 diabetes susceptibility gene that potentially plays a key role in regulating central carbon metabolism in β-cells. Previous work suggested that β-cell PGM1 transcription is lowered after coxsackievirus B4 infection. Thus, we hypothesized that decreased PGM1 levels disrupt β-cell metabolic homeostasis and result in β-cell fragility and type 1 diabetes. First, we showed that the synthetic double-stranded RNA polyinosinic:polycytidylic acid, or Poly(I:C) attenuated PGM1 transcription both in human islets and EndoC-βH1 cell line. At 5.5 mmol/L glucose, PGM1 deficiency enhanced the rate of glycolysis, tricarboxylic acid cycle, hexosamine, and pentose phosphate pathway. However, at 20 mmol/L glucose, PGM1-deficient cells showed impaired mitochondrial respiration. Moreover, truncated N-glycans were enriched in PGM1-deficient cells, suggesting aberrant protein glycosylation. Autophagic flux, which was dependent on the lysosomal glycosylated protein function, was impaired in PGM1-deficient cells. Increased endoplasmic reticulum stress was evident in PGM1-deficient cells. Our results suggest that PGM1 is a metabolic regulator of pancreatic β-cells. Its deficiency leads to metabolic imbalance and cellular stress, potentially augmenting type 1 diabetes development.

ARTICLE HIGHLIGHTS: In the β-cell, the expression of phosphoglucomutase 1 (PGM1), a type 1 diabetes risk gene, is reduced by double-stranded RNA exposure, modeled by polyinosinic:polycytidylic acid transfection. Deficient PGM1 expression disrupts central carbon metabolism, protein glycosylation, and autophagic flux. These changes precipitate endoplasmic reticulum stress and mitochondrial dysfunction, potentially augmenting type 1 diabetes development.

Hypertension remains the most prevalent modifiable risk factor for cardiovascular morbidity and mortality worldwide, yet rates of effective blood pressure control remain persistently suboptimal despite the availability of multiple therapeutic options. This gap reflects fundamental limitations of current care models, which rely on episodic measurements, population-based treatment algorithms, and incomplete representation of the biological, behavioral, and social complexity underlying blood pressure regulation. Artificial intelligence (AI) offers a transformative framework to address these challenges by enabling the integration of longitudinal, multimodal data and modeling nonlinear, dynamic relationships that are difficult to capture with conventional approaches. This systematic review synthesizes emerging evidence on the application of AI across the hypertension care continuum, including risk prediction, phenotyping, blood pressure measurement, wearable-based monitoring, clinical trial analysis, population health modeling, detection of secondary hypertension, behavioral and adherence interventions, and multi-omics-driven precision medicine. We highlight the methodological foundations required for clinically meaningful AI, emphasizing robust ground-truth definitions, external and temporal validation, interpretability, workflow integration, and equity-aware design. The review also examines the promise and limitations of natural language processing, cuffless blood pressure technologies, and AI-guided decision support systems, alongside ethical, regulatory, and implementation challenges. Collectively, current evidence suggests that AI has the potential to shift hypertension management from a reactive, threshold-based paradigm toward a more predictive, personalized, and patient-centered model. Realizing this potential will depend on rigorous validation, thoughtful implementation, and sustained alignment with clinical, ethical, and equity principles.

Cellular phenotypes are shaped not only by current molecular states but by transient transcriptional programs that encode prior experiences and influence future behavior. Conventional transcriptomic approaches, including bulk and single-cell RNA sequencing, provide high-resolution snapshots of gene expression but are intrinsically destructive, precluding direct linkage between past transcriptional states and downstream cellular fate. In this context, "TimeVault" introduces a fundamentally new paradigm by enabling intracellular storage of endogenous transcriptomes within living cells. By repurposing vault ribonucleoprotein particles to sequester and stabilize polyadenylated mRNA, TimeVault preserves unbiased, transcriptome-wide records of transcriptional states over timescales far exceeding native mRNA half-lives. This capability allows retrospective reconstruction of molecular histories that would otherwise be lost, bridging a critical gap between transient gene expression and long-term phenotypic outcomes. Application of TimeVault to canonical stress responses demonstrates precise temporal gating and durable transcript preservation, while its use in cancer models reveals preexisting transcriptional programs that predict drug-tolerant persister cell formation prior to therapy. These findings highlight the power of molecular memory devices to uncover causal relationships that remain invisible to conventional endpoint analyses. TimeVault establishes intracellular transcriptome archiving as a versatile tool with broad implications for developmental biology, stress adaptation, and therapeutic resistance.

Mitochondrial dysfunction devastates the heart in major cardiovascular diseases, yet the mechanisms governing mitochondrial quality control remain elusive. We discovered that TIGAR (TP53-induced glycolysis and apoptosis regulator) deficiency established profound cardiac protection through developmental epigenetic programming of Parkin expression. Using whole-body and cardiomyocyte-specific TIGAR knockout mice, we demonstrated remarkable cardioprotection following myocardial infarction with maintained ejection fraction, and complete resistance to diet-induced cardiac hypertrophy despite comparable weight gain. TIGAR deficiency triggered dramatic increases in Parkin expression across all somatic tissues except testes, where Parkin levels remained extraordinarily high (100-fold greater than cardiac levels) regardless of TIGAR status, revealing tissue-specific regulatory mechanisms. This protection was entirely Parkin-dependent, as double knockout mice lost all cardioprotective benefits. Crucially, adult TIGAR manipulation failed to alter Parkin levels, demonstrating that this pathway operated exclusively during critical developmental windows to program lifelong cardiac resilience. Whole-genome bisulfite sequencing identified reduced DNA methylation in Prkn intron 10 as the key regulatory mechanism, with CRISPR deletion dramatically increased Parkin expression in multiple cell lines. Our findings reveiled how early cardiac metabolism programmed lifelong cardiac function through epigenetic mechanisms, and identifyied developmental metabolic programming as a potential therapeutic target for preventing both ischemic heart disease and metabolic cardiomyopathy.