Aplicações de Inteligência Artificial na Neurologia: A Influência das Tecnologias no Diagnóstico e Monitoramento

Abstract

Artificial intelligence (AI), introduced by John McCarthy in 1956, represents the fourth industrial revolution, transforming areas of everyday life. Machine learning (ML), a branch of AI created by Arthur Samuel in 1959, uses algorithms and neural networks to recognize patterns in data. Deep learning (DL), a specialization of ML, mimics brain processing with convolutional neural networks, especially in image and video analysis. In healthcare, AI helps analyze medical data and diagnose neurological disorders, with ML techniques showing advances in identifying conditions such as Alzheimer's, schizophrenia and others. The study aims to investigate how AI technologies impact the diagnosis and monitoring of neurological disorders. It focuses on the effectiveness of wearable devices such as smartwatches in detecting tremors and neurological abnormalities and personalizing treatment. It also explores the role of mobile applications in medication adherence and disease screening, seeking to improve diagnostic efficiency and clinical outcomes. This research is a qualitative, exploratory and descriptive literature review that investigates the applications of artificial intelligence in the clinical practice of neurology. The search was carried out in the SciELO and PubMed databases, using the search terms “Artificial Intelligence (AI)” AND “Neurology”. Original, free articles published in Portuguese in the last ten years were included. The study excluded incomplete, repeated works that did not meet the proposed criteria. The analysis was completed in August 2024. The advancement of wearable technologies and AI has revolutionized the diagnosis and monitoring of neurological disorders, highlighted by the use of smartwatches and tablets. Devices such as Apple's SDS and machine learning algorithms have enabled the detection and analysis of tremors with high precision, benefiting the diagnosis of diseases such as Parkinson's and epilepsy. Mobile applications and electronic questionnaires complement these devices, providing a more complete assessment of patients. The integration of AI into neuroimaging and the use of sensors for continuous monitoring have also shown significant improvements in diagnostic accuracy and treatment personalization. However, challenges such as algorithm complexity and data protection need to be addressed to optimize the effectiveness of these technologies.