Genetic fuzzy modelling and control of bispectral index (BIS) for general intravenous anaesthesia

Artificial intelligence-assisted interventions for perioperative anesthetic management: a systematic review and meta-analysis

BACKGROUND

Integration of artificial intelligence (AI) into medical practice has increased recently. Numerous AI models have been developed in the field of anesthesiology; however, their use in clinical settings remains limited. This study aimed to identify the gap between AI research and its implementation in anesthesiology via a systematic review of randomized controlled trials with meta-analysis (CRD42022353727).

METHODS

We searched the databases of Medical Literature Analysis and Retrieval System Online (MEDLINE), Excerpta Medica Database (Embase), Web of Science, Cochrane Central Register of Controlled Trials (CENTRAL), Institute of Electrical and Electronics Engineers Xplore (IEEE), and Google Scholar and retrieved randomized controlled trials comparing conventional and AI-assisted anesthetic management published between the date of inception of the database and August 31, 2023.

RESULTS

Eight randomized controlled trials were included in this systematic review (n = 568 patients), including 286 and 282 patients who underwent anesthetic management with and without AI-assisted interventions, respectively. AI-assisted interventions used in the studies included fuzzy logic control for gas concentrations (one study) and the Hypotension Prediction Index (seven studies; adding only one indicator). Seven studies had small sample sizes (n = 30 to 68, except for the largest), and meta-analysis including the study with the largest sample size (n = 213) showed no difference in a hypotension-related outcome (mean difference of the time-weighted average of the area under the threshold 0.22, 95% confidence interval -0.03 to 0.48, P = 0.215, I2 93.8%).

CONCLUSIONS

This systematic review and meta-analysis revealed that randomized controlled trials on AI-assisted interventions in anesthesiology are in their infancy, and approaches that take into account complex clinical practice should be investigated in the future.

TRIAL REGISTRATION

This study was registered with the International Prospective Register of Systematic Reviews (PROSPERO ID: CRD42022353727).

Machine Learning, Deep Learning, and Closed Loop Devices-Anesthesia Delivery

With the tremendous volume of data captured during surgeries and procedures, critical care, and pain management, the field of anesthesiology is uniquely suited for the application of machine learning, neural networks, and closed loop technologies. In the past several years, this area has expanded immensely in both interest and clinical applications. This article provides an overview of the basic tenets of machine learning, neural networks, and closed loop devices, with emphasis on the clinical applications of these technologies.

Artificial Intelligence in Anesthesiology: Current Techniques, Clinical Applications, and Limitations

Artificial intelligence has been advancing in fields including anesthesiology. This scoping review of the intersection of artificial intelligence and anesthesia research identified and summarized six themes of applications of artificial intelligence in anesthesiology: (1) depth of anesthesia monitoring, (2) control of anesthesia, (3) event and risk prediction, (4) ultrasound guidance, (5) pain management, and (6) operating room logistics. Based on papers identified in the review, several topics within artificial intelligence were described and summarized: (1) machine learning (including supervised, unsupervised, and reinforcement learning), (2) techniques in artificial intelligence (e.g., classical machine learning, neural networks and deep learning, Bayesian methods), and (3) major applied fields in artificial intelligence.The implications of artificial intelligence for the practicing anesthesiologist are discussed as are its limitations and the role of clinicians in further developing artificial intelligence for use in clinical care. Artificial intelligence has the potential to impact the practice of anesthesiology in aspects ranging from perioperative support to critical care delivery to outpatient pain management.

Adaptive fuzzy modeling of the hypnotic process in anesthesia

This paper addresses the problem of patient model synthesis in anesthesia. Recent advanced drug infusion mechanisms use a patient model to establish the proper drug dose. However, due to the inherent complexity and variability of the patient dynamics, difficulty obtaining a good model is high. In this paper, a method based on fuzzy logic and genetic algorithms is proposed as an alternative to standard compartmental models. The model uses a Mamdani type fuzzy inference system developed in a two-step procedure. First, an offline model is obtained using information from real patients. Then, an adaptive strategy that uses genetic algorithms is implemented. The validation of the modeling technique was done using real data obtained from real patients in the operating room. Results show that the proposed method based on artificial intelligence appears to be an improved alternative to existing compartmental methodologies.

Reinforcement learning: a novel method for optimal control of propofol-induced hypnosis

Reinforcement learning (RL) is an intelligent systems technique with a history of success in difficult robotic control problems. Similar machine learning techniques, such as artificial neural networks and fuzzy logic, have been successfully applied to clinical control problems. Although RL presents a mathematically robust method of achieving optimal control in systems challenged with noise, nonlinearity, time delay, and uncertainty, no application of RL in clinical anesthesia has been reported.

Reinforcement learning versus proportional-integral-derivative control of hypnosis in a simulated intraoperative patient

BACKGROUND

Research has demonstrated the efficacy of closed-loop control of anesthesia using bispectral index (BIS) as the controlled variable. Model-based and proportional-integral-derivative (PID) controllers outperform manual control. We investigated the application of reinforcement learning (RL), an intelligent systems control method, to closed-loop BIS-guided, propofol-induced hypnosis in simulated intraoperative patients. We also compared the performance of the RL agent against that of a conventional PID controller.

METHODS

The RL and PID controllers were evaluated during propofol induction and maintenance of hypnosis. The patient-hypnotic episodes were designed to challenge both controllers with varying degrees of interindividual variation and noxious surgical stimulation. Each controller was tested in 1000 simulated patients, and control performance was assessed by calculating the median performance error (MDPE), median absolute performance error (MDAPE), Wobble, and Divergence for each controller group. A separate analysis was performed for the induction and maintenance phases of hypnosis.

RESULTS

During maintenance, RL control demonstrated an MDPE of -1% and an MDAPE of 3.75%, with 80% of the time at BIS(target) ± 5. The PID controller yielded a MDPE of -8.5% and an MDAPE of 8.6%, with 57% of the time at BIS(target) ± 5. In comparison, the MDAPE in the worst-controlled patient of the RL group was observed to be almost half that of the worst-controlled patient in the PID group.

CONCLUSIONS

When compared with the PID controller, RL control resulted in slower induction but less overshoot and faster attainment of steady state. No difference in interindividual patient variation and noxious destabilizing challenge on control performance was observed between the 2 patient groups.

Human heart beat analysis using a modified algorithm of detrended fluctuation analysis based on empirical mode decomposition

How to quantify the complexity of a physiological signal is a crucial issue for verifying the underlying mechanism of a physiological system. The original algorithm of detrended fluctuation analysis (DFA) quantifies the complexity of signals using the DFA scaling exponent. However, the DFA scaling exponent is suitable only for an integrated time series but not the original signal. Moreover, the method of least squares line is a simple detrending operation. Thus, the analysis results of the original DFA are not sufficient to verify the underlying mechanism of physiological signals. In this study, we apply an innovative timescale-adaptive algorithm of empirical mode decomposition (EMD) as the detrending operation for the modified DFA algorithm. We also propose a two-parameter scale of randomness for DFA to replace the DFA scaling exponent. Finally, we apply this modified algorithm to the database of human heartbeat interval from Physiobank, and it performs well in identifying characteristics of heartbeat interval caused by the effects of aging and of illness.