Browsing by Author "Brás, Susana"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
- Electroencephalogram-based indices applied to dogs' depth of anaesthesia monitoringPublication . Brás, Susana; Georgakis, A.; Ribeiro, L.; Ferreira, David A.; Silva, A.; Antunes, L.; Nunes, Catarina S.Hypnotic drug administration causes alterations in the electroencephalogram (EEG) in a dose-dependent manner. These changes cannot be identified easily in the raw EEG, therefore EEG based indices were adopted for assessing depth of anaesthesia (DoA). This study examines several indices for estimating dogs' DoA. Data (EEG, clinical end-points) were collected from 8 dogs anaesthetized with propofol. EEG was initially collected without propofol. Then, 100 ml h−1 (1000 mg h−1) of propofol 1% infusion rate was administered until a deep anaesthetic stage was reached. The infusion rate was temporarily increased to 200 ml h−1 (2000 mg h−1) to achieve 80% of burst suppression. The index performance was accessed by correlation coefficient with the propofol concentrations, and prediction probability with the anaesthetic clinical end-points. The temporal entropy and the averaged instantaneous frequency were the best indices because they exhibit: (a) strong correlations with propofol concentrations, (b) high probabilities of predicting anaesthesia clinical end-points.
- Garch models for drug effects on patient heart rate, during general anaesthesiaPublication . Brás, Susana; Nunes, Catarina S.; Amorim, PedroA model that can describe the effect of anaesthetic drugs on patient’s heart rate (HR) is of great importance when considering haemodynamic stability under surgery. A Generalized Autoregressive Conditional Heteroscedasticity (GARCH) model was used to model HR considering the effect concentrations of the anaesthetic propofol and the analgesic remifentanil, using the clinical data of 16 patients. The model was able to capture the HR trend in all 16 patients with very small errors throughout the surgical time. A correlation was found between the GARCH parameters and patient baseline characteristics, leading to the possibility a patient adjusted adaptive model.