Revisão sistemática quanto ao uso do EEG para diagnostico de Morte encefálica

Abstract

This study presents a systematic review on the use of electroencephalogram (EEG) as a diagnostic tool for brain death. The main objective was to quantitatively analyze the results found in the literature on this topic. The research was conducted through the ScienceDirect and Google Scholar platforms, covering studies published between 2015 and 2025. Articles that used EEG in patients with a confirmed clinical diagnosis of brain death were included, provided they contained a quantitative analysis of the EEG signal. The selection process followed the PRISMA guidelines, resulting in the inclusion of six studies that met the eligibility criteria. The extracted data were organized into two categories: clinical aspects and technical aspects. The results showed that all studies used a sampling frequency of 1000 Hz, with electrodes focused on frontal regions of the skull and the application of band-pass filters combined with advanced preprocessing techniques such as wavelet decomposition, ICA, VMD, and Dynamic 2T-EMD. An average reduction of 87% in EEG spectral energy was observed in patients with brain death when compared to coma. Multiscale entropy also showed an average decrease of 54%, indicating lower signal complexity in brain death cases. Computational models such as Support Vector Machine and Random Forest achieved accuracy rates of up to 99.6% in the classification of neurological states. These findings demonstrate the potential of EEG, especially when combined with quantitative techniques and machine learning, as a robust auxiliary tool in the diagnosis of brain death, promoting greater safety and precision in clinical practice.