E-Nose system for anesthetic dose level detection using artificial neural network

Area of Science:

• Biomedical Engineering
• Anesthesiology
• Sensor Technology

Background:

• Accurate anesthetic dose monitoring is crucial for patient safety during surgery.
• Existing methods for anesthetic monitoring may have limitations in real-time precision.
• Sevoflurane is a commonly used inhalation anesthetic agent.

Purpose of the Study:

• To develop and validate an electronic nose (E-Nose) system for predicting anesthetic dose levels.
• To assess the efficacy of quartz crystal microbalances (QCM) sensors in measuring sevoflurane concentrations.
• To utilize artificial neural networks (ANNs) for precise anesthetic dose level prediction.

Main Methods:

• An E-Nose system equipped with a QCM sensor array was employed to measure sevoflurane.
• Frequency changes from QCM sensors corresponding to eight sevoflurane dose levels were recorded.
• A multilayer feedforward artificial neural network (MLNN) was trained using the Levenberg-Marquardt algorithm.

Main Results:

• The MLNN model successfully established a relationship between QCM sensor frequency changes and sevoflurane dose levels.
• The trained MLNN demonstrated acceptable accuracy in predicting anesthetic dose levels when tested with random data.
• The E-Nose system showed promising results for real-time anesthetic monitoring.

Conclusions:

• The developed E-Nose system integrated with MLNNs provides a viable approach for accurate sevoflurane anesthetic dose prediction.
• QCM sensor technology is effective for detecting and quantifying anesthetic agents.
• This system has the potential to enhance anesthetic management and patient safety in surgical settings.