In the last decades, different algorithms for anesthesia monitoring have been developed and several devices are available in the operating theaters for that purpose. Even though they have improved the general anesthesia procedures by monitoring the effects of hypnotics in the brain and reducing the incidence of awareness (among other positive outcomes), the management of analgesics to avoid pain is still not fully covered by them.

Patients under anesthesia are unconscious and often paralyzed. When painful stimulation is applied, such as surgical incision or intubation, physiological reactions to the aggression take place, provoking alterations in the patient condition and/or problems in the post-operative period. It is therefore key to monitor the level of pain that an anesthetized patient is suffering in order to administer enough analgesics to avoid pain reactions, while not exceeding the appropriate dose, since this would have undesired effects such as bradycardia and/or hypotension.

Several algorithms exist aiming at reflecting the nociceptive/anti-nociceptive balance of a patient, but most of them are known to be unspecific and cannot be used as a guidance to deliver analgesics during anesthesia. This project aims at developing a new nociception index that continuously measures the nociception/anti-nociception balance of the patient and that proves to be useful guiding analgesia titration, by abolishing patient reactions to pain without causing major hemodynamic alterations due to analgesics overdose.

The main tasks for this research project are:

State of the art – Literature review of available technologies to monitor analgesia during general anesthesia and their advantages and limitations.

Data recording and processing: Planning and executing clinical studies to collect clinical data such as raw EEG data, drug titration and other relevant clinical recordings. Signal processing of the recorded signals, mainly the EEG and other physiological signals, using both linear and non-linear techniques, for: noise removal, feature extraction and other signal conditioning strategies to obtain a clean dataset for further analysis.

Algorithm development phase 1 – Data Analysis: Identification of the physiological signals from the recorded database showing (1) correlation with painful episodes and (2) correlation with analgesic dosages.

Algorithm development phase 2 – Software Implementation: Development of new algorithms reflecting the nociceptive/anti-nociceptive balance of the patient, based on EEG and other physiological variables and their combination, using non-linear techniques, such as fuzzy logic and machine learning. Algorithm verification and evaluation of the impact of its use in current clinical practice.

Thesis writing.