Cardiovascular Disease, Sleep-Disordered Breathing, and Artificial Intelligence: From Neutral Trials to Precision Sleep Cardiology

Fig. 1.. Central illustration: overview of AI applications in SDB and cardiovascular disease. While SDB contributes to cardiovascular injury through complex mechanisms, including intermittent hypoxemia and inflammation, conventional treatments often yield inconclusive results. This figure highlights how AI could help integrate detection, risk stratification, and personalized management to bridge the gap and improve clinical outcomes. PSG, polysomnography; ML, machine learning; MACE, major adverse cardiovascular events; AI, artificial intelligence; SDB, sleep-disordered breathing.

Fig. 2.. PRISMA flow diagram of the study selection process. This flowchart depicts the search and screening process of selecting studies of this review. A thorough search was performed across PubMed, Web of Science, Embase, and Scopus. Following the removal of duplicates and the exclusion of studies that did not meet the eligibility criteria, relevant articles were chosen for inclusion in the final analysis.

Fig. 3.. Pathophysiological mechanisms linking sleep apnea to cardiovascular outcomes. Both obstructive and central sleep apnea trigger a cascade of physiological changes, including intermittent hypoxemia, fluctuations in intrathoracic pressure, and sleep fragmentation. These alterations result in oxidative stress, sympathetic activation, inflammation, endothelial dysfunction, and metabolic dysregulation. These mechanisms contribute to cardiovascular outcomes like heart failure, myocardial ischemia, arrhythmias, hypertension, and stroke. CPAP may help mitigate these effects in sleep apnea by addressing upstream pathophysiological disturbances. CPAP, continuous positive airway pressure; NF-κB, nuclear factor-κB.

Fig. 4.. The evolving landscape of cardiology–sleep medicine: from traditional approaches to AI-driven precision care. AI could enable a shift from traditional, generalized sleep medicine to preventive and personalized cardiology–sleep care. Through automated detection, individualized risk stratification, and tailored therapy, AI can address key limitations such as mechanistic uncertainty and uniform treatment. The goal is timely diagnosis and risk-based management of SDB, supported—not replaced—by technology. Overcoming clinical, technical, and ethical hurdles will be essential, but AI holds promise for reducing CVDs by targeting modifiable contributors like sleep apnea.