Atenção multidimensional no diagnóstico do acidente vascular cerebral: evidências do uso da inteligência artificial na prática em saúde

Abstract

Introduction: Artificial Intelligence (AI) has been emerging as an important tool in diagnostic medicine, especially in the medical fields of cardiology and neurology. Its application in the diagnosis of cardiovascular diseases and the management of strokes represents a significant advancement in clinical accuracy, speed, and efficiency. From this perspective, it becomes important to critically analyze the current contributions of Artificial Intelligence in medical practice. Objective: Therefore, the objective of this study was to investigate the application of AI in the diagnosis of stroke cases and its benefits, limitations, and impacts in the clinical context. Methodology: This is a narrative review of articles published between 2021 and 2025, selected from the PubMed, Scopus, and Embase databases. Studies that directly addressed the use of AI in echocardiography, electrocardiogram, and neuroimaging related to stroke were included. The following descriptors were used: “Stroke,” “Artificial Intelligence,” “Diagnosis,” “Computed Tomography,” and “Magnetic Resonance Imaging.” Results: Learning algorithms such as convolutional neural networks have shown high performance in the early detection of strokes and left ventricular systolic dysfunction. AI demonstrates high accuracy and predictive value, especially in tests such as ECG and echocardiogram, and enables rapid and reliable analysis of brain images in the acute setting. Conclusion: AI plays a promising and growing role in medicine, contributing to faster and more accurate diagnoses, but its implementation requires rigorous criteria and professional training. The technology should be seen as a support for clinical decision-making, not as a replacement for medical reasoning.