Isabela Birs

TE56/2025

Event-based surgical protocol digitalization for system identification and robust control in general anesthesia

This project is funded by a grant of the Ministry of Research, Innovation and Digitization, CNCS -UEFISCDI, project number PN-IV-P2-2.1-TE-2023-0831, within PNCDI IV.

Project results are reported here:

• Etapa I – 2025
• Etapa II – 2026

In this project, novel concepts of fractional order control are designed and tested on a full anesthesia-hemodynamic patient simulator. In order to implement such validation and demonstration tools, additional features must be added to the existing patient simulator, such as patient model variability, blood loss effects in time constant and steady-state values, and disturbance profiles tailored to different types of surgery. The transition between the induction phase – typically a reference tracking control issue – and the maintenance phase – a perturbation rejection mode – is also proposed to be resolved using funnel-type output limit interval guidance. A 5×5 multivariable simulator will be utilised to validate the proposed control strategies and investigate the inherent closed loops’ stability. Existing patient databases and knowledge of surgical protocols are utilised to create a realistic simulator environment. The deliverable represents a state-of-the-art breakthrough for the regulatory paradigm of anesthesia depth and a cross-disciplinary value addition to both clinical practice and control methodologies.

PD15/2021

Event-based fractional order control of general anaesthesia

State of the art biomedical practice uses computer-based control of anesthesia for patient stabilization during surgical operations, intensive care units and rehabilitation periods.  The main objective as well as the major difficulty of general anesthesia are to obtain a perfect balance between hypnosis, analgesia and areflexia during surgery with a minimum impact upon the patient. Current limitations of existing approaches include inter- and intra- patient variability and focus of the control strategy on isolated parts of the anesthesia process, with a complete disregard of the high coupling between anesthesia and hemodynamics. The project proposes a novel approach to overcome these limitations by controlling the combined anesthesia and hemodynamic system as a multivariable process. The main methodology consists of the event-based fractional order control, a novel strategy that combines well known advantages of fractional calculus with event-based control benefits. Surgical stimulus and bolus are considered events that trigger the computation of new control variables based on predefined paradigms of real-life anesthesia conditions, in order to improve the closed loop system response during both induction and maintenance phases and overcome intra-patient variability (various drug dosages leading to different responses in the same patient) and inter-patient variability (data available for 24 patients). The methodology is validated on a highly customizable patient simulator.