Untitled

Funder

Organizational Unit

Publications

A novel multivariate STeady-state index during general ANesthesia (STAN)

Publication . Castro, Ana; Almeida, Fernando Gomes de; Amorim, Pedro; Nunes, Catarina S.

The assessment of the adequacy of general anesthesia for surgery, namely the nociception/anti-nociception balance, has received wide attention from the scientific community. Monitoring systems based on the frontal EEG/EMG, or autonomic state reactions (e.g. heart rate and blood pressure) have been developed aiming to objectively assess this balance. In this study a new multivariate indicator of patients' steady-state during anesthesia (STAN) is proposed, based on wavelet analysis of signals linked to noxious activation. A clinical protocol was designed to analyze precise noxious stimuli (laryngoscopy/intubation, tetanic, and incision), under three different analgesic doses; patients were randomized to receive either remifentanil 2.0, 3.0 or 4.0 ng/ml. ECG, PPG, BP, BIS, EMG and [Formula: see text] were continuously recorded. ECG, PPG and BP were processed to extract beat-to-beat information, and [Formula: see text] curve used to estimate the respiration rate. A combined steady-state index based on wavelet analysis of these variables, was applied and compared between the three study groups and stimuli (Wilcoxon signed ranks, Kruskal-Wallis and Mann-Whitney tests). Following institutional approval and signing the informed consent thirty four patients were enrolled in this study (3 excluded due to signal loss during data collection). The BIS index of the EEG, frontal EMG, heart rate, BP, and PPG wave amplitude changed in response to different noxious stimuli. Laryngoscopy/intubation was the stimulus with the more pronounced response [Formula: see text]. These variables were used in the construction of the combined index STAN; STAN responded adequately to noxious stimuli, with a more pronounced response to laryngoscopy/intubation (18.5-43.1 %, [Formula: see text]), and the attenuation provided by the analgesic, detecting steady-state periods in the different physiological signals analyzed (approximately 50 % of the total study time). A new multivariate approach for the assessment of the patient steady-state during general anesthesia was developed. The proposed wavelet based multivariate index responds adequately to different noxious stimuli, and attenuation provided by the analgesic in a dose-dependent manner for each stimulus analyzed in this study.

2017-08Journal article

Open access

Controlling the hypnotic drug (propofol) to maintain a stable depth of anesthesia, in dogs

Publication . Bras, Susana; Ribeiro, Lenio; Ferreira, David A.; Antunes, Luis; Nunes, Catarina S.

The development of control and automatic systems is important to guarantee a stable anesthesia, with no under or over dosage, and no awareness episodes. In this study a controller for the Cerebral State Index (CSI — an electroencephalogram derived signal) was developed. This study was a simulation study, the CSI was modeled using a fuzzy logic model with two inputs the effect-site concentration of propofol and the electromyography (EMG). The controller was tested using constant and variable references in an exhaustive set of simulations. The controller developed presents a good performance in all simulations and the controlled variable seems to be influenced by electromyography level. A controller for propofol anesthesia for veterinary use is an important step towards the improvement of animal welfare. The overall aim is to improve animal safety and comfort.

2014-07-24Conference object

Open access

Towards automation in anaesthesia: a review

Publication . Nunes, Catarina S.

Surgeries represent a risk for patients and a big cost for the hospital. Anaesthesia represents a complex part of surgery also carries risks for patients. The most known are awareness (with deep psychological consequences); increased risk of morbidity and mortality; adverse reactions and long post-op recovery. The complexity of anaesthesia management can be reduced by studying the patients' responses and developing indicators of the patient state. To assess the level of depth of anaesthesia, the anaesthetist needs to be aware of the patient physiological responses to the drugs and to surgical stimuli. A system that could advise on the patient state considering all clinical signs being measured, the patient individual response and the amount of drugs, will have a big impact on patient overall safety and future health, post-op recovery and hospital resources. This paper does a review of different systems and methods applied to several aspects of the anaesthesia field. All with the goal of working towards automation in this very complex area, that involves high risks for patients. This paper covers advisor systems; signal processing; new monitors and devices; mathematical modelling; and control algorithms; all focused on practical clinical implementation. The objective is to have an overview of the work done so far and the steps taken towards automation in anaesthesia.

2014-07-24Conference object

Open access