In management of cerebrovascular diseases, artificial intelligence (AI) emerges and plays a role through advancements in prediction, imaging, diagnosis, and treatment. Revolutionized AI-based tools allow us to catch a glimpse of possibility of rapid clinical workflows in detection, reducing time to the treatment while improving diagnostic accuracy compared to conventional methods. AI has the potential to facilitate primary prevention, improve acute detection, and provide additional decision-making beyond acute care. This review synthesizes applications of AI in major cerebrovascular diseases, introducing an overview of its efficiency and limitations.
- Artificial intelligence
- Cerebrovascular diseases
- Stroke
- Aneurysms
- +4
PerspectiveElectrophysiology is undergoing rapid transformation, with pulsed field ablation, high‑density mapping, conduction system pacing, and leadless devices reshaping both practice and training. The 2026 ACC/AHA/HRS Advanced Training Statement reflects this shift by emphasizing competency‑based assessment and expanding core procedural exposure. Yet as automation streamlines workflows, it risks diminishing the mechanism‑based reasoning that has long defined the field, echoing broader concerns about physician deskilling. Ensuring that technology enhances rather than replaces interpretive skill is essential. The future electrophysiologist must pair technological fluency with the ability to independently explain arrhythmia mechanisms and interpret data skills that remain central to safe, thoughtful, and effective practice.
Metabolic dysfunction-associated steatotic liver disease (MASLD), affecting nearly 40% of the global population, represents a paradigm shift that views liver fat as a primary hepatic manifestation of systemic metabolic dysregulation rather than an isolated liver condition. MASLD and its progressive form, metabolic dysfunction-associated steatohepatitis (MASH), drive substantial extrahepatic mortality through cardiovascular disease, chronic kidney disease, and extrahepatic malignancies. However, current care models remain fragmented, as non-hepatology specialties frequently overlook hepatic manifestations and primary care often underestimates fibrosis risk in patients with normal transaminase levels. Addressing this systemic metabolic crisis requires health systems to implement an integrated, interdisciplinary framework combining standardized primary care screening protocols using non-invasive tools (e.g., FIB-4 index), multi-target pharmacotherapies (such as GLP-1 receptor agonists, SGLT2 inhibitors, and resmetirom), and dedicated co-managed clinics across primary care, cardiology, endocrinology, nephrology, oncology, and hepatology to detect fibrosis earlier and prevent irreversible cardiorenal and hepatic outcomes.
- MASLD
- Fatty liver disease
- NAFLD
- MASH
- +4
EditorialLipoprotein(a) [Lp(a)] is a genetically determined cardiovascular risk factor associated with atherosclerotic cardiovascular disease through proatherogenic, proinflammatory, and prothrombotic mechanisms. The neutral primary outcome of the Lp(a)HORIZON trial, despite pharmacologic Lp(a) lowering with pelacarsen, challenges assumptions about the clinical benefits of targeting Lp(a) in patients with established cardiovascular disease. However, these findings do not necessarily invalidate Lp(a) as a causal risk factor or the clinical value of measuring Lp(a). Lp(a)HORIZON primarily tested whether lowering Lp(a) after established disease reduces recurrent events, which is a different question from whether lifelong exposure to elevated Lp(a) contributes to disease development or whether earlier intervention can prevent first cardiovascular events. Ongoing primary- and mixed-prevention trials may help address this distinction. The Lp(a) experience highlights the importance of separating risk assessment from therapeutic efficacy and raises broader questions about the timing, magnitude, and duration of biomarker-targeted interventions. Adaptive guideline frameworks and AI-assisted evidence synthesis may help integrate emerging trial data more rapidly while preserving rigorous human clinical judgment.
- Lipoprotein(a)
- Lp(a)
- cardiovascular risk
- primary prevention
- +3
EditorialArtificial intelligence (AI) is rapidly transforming scientific publishing, yet current disclosure practices provide limited information about the extent of AI involvement. The proposed Clinical AI Score (CAIS) is a standardized framework for quantifying AI assistance across manuscript preparation, analysis, and visualization. CAIS is intended to promote transparency and human accountability rather than restrict AI use. A preliminary threshold of <10 is proposed to represent limited-to-moderate AI involvement, pending future validation. Transparent AI disclosure, combined with digital provenance and human oversight, may provide a practical framework for maintaining scientific integrity in the evolving era of AI-assisted medical publishing.
- Artificial intelligence
- Clinical AI Score
- CAIS
- Scientific publishing
- +2
EditorialThe rapid growth of cardiovascular evidence creates significant challenges in accessing timely, accurate, and clinically relevant information. CardioOracle is a free, AI-powered clinical intelligence platform developed specifically for cardiovascular medicine. It integrates curated guidelines, consensus documents, landmark clinical trials, meta-analyses, and evidence-based reviews to provide evidence-grounded responses for healthcare professionals and patients. The platform also supports clinical documentation and patient education. CardioOracle is designed to reduce information-retrieval burden while supporting, rather than replacing, clinical judgment.
- Artificial intelligence
- Cardiology
- Clinical decision support
- Evidence-based medicine
- +3
EditorialCardiovascular medicine has evolved rapidly, while physician certification remains largely centered on periodic knowledge-based examinations. The emergence of increasingly capable artificial intelligence (AI) systems challenges this traditional model and raises a fundamental question: should certification continue to primarily measure what physicians know, or should it measure how physicians apply knowledge in clinical practice? I propose a modern certification framework focused on demonstrated clinical competence, continuous learning, individualized assessment, practice improvement, and the ability to appropriately use, supervise, and override AI. Such a model could better reflect the evolving competencies required of contemporary cardiologists while maintaining accountability and professional standards. The future of cardiovascular certification should be centered not simply on knowledge acquisition, but on clinical judgment, adaptability, and competence in an AI enabled healthcare environment.
- Cardiovascular medicine
- Board certification
- Artificial Intelligence
- Medical education
- +2
