The paper describes a method for automatic assessment of the risk of delayed cerebral ischemia (DCI) of the brain after subarachnoid hemorrhage (SAH) based on continuous electroencephalographic (EEG) monitoring. The method is based on the study of the quantity of wave trains (local maxima) in wavelet spectrograms of background EEG signals. Using the AUC diagram, the boundaries of the wave train parameter ranges were determined, in which differences between patients with DCI and patients without DCI were observed. The data of continuous EEG monitoring of patients in the intensive care unit of the N.V. Sklifosovsky Research Institute of Emergency Care were analyzed. The sample of reference patients included 6 patients with DCI (a total of 76 six-hour measurements) and 8 patients without DCI (a total of 92 six-hour measurements). The following wave train parameters were used to assess the risk of DCI: central frequency, maximum spectral power density, and wave train bandwidth. The method for assessing the risk of developing DCI includes: calculating the number of wave trains per hour, the parameters of which lie in the ranges determined using AUC diagrams, calculating the medians of the number of wave trains detected in the EEG channels F3, F4, F7, F8, C3, C4, P3, P4, O1, and O2, and comparing the calculated medians with threshold values. The risk of developing DCI is assessed on a scale of “a very high risk”, “a high risk”, and “a low risk”. The advantages of the method include the ability to apply it to patients who have no EEG alpha rhythm for reasons not related to the development of DCI, as well as a high accuracy of assessing the risk of developing DCI. The accuracy of the method is 87.1% in assessing a very high risk of developing DCI and 90.0% in assessing a high risk of developing DCI. In some patients, the method reliably identifies the risk of developing DCI as early as 48 hours after the surgery for non-traumatic subarachnoid hemorrhage.
Sushkova et al. (2025) studied this question.