Multiple-drug hemodynamic control using fuzzy decision theory

The significance of artificial intelligence in drug delivery system design

Over the last decade, increasing interest has been attracted towards the application of artificial intelligence (AI) technology for analyzing and interpreting the biological or genetic information, accelerated drug discovery, and identification of the selective small-molecule modulators or rare molecules and prediction of their behavior. Application of the automated workflows and databases for rapid analysis of the huge amounts of data and artificial neural networks (ANNs) for development of the novel hypotheses and treatment strategies, prediction of disease progression, and evaluation of the pharmacological profiles of drug candidates may significantly improve treatment outcomes. Target fishing (TF) by rapid prediction or identification of the biological targets might be of great help for linking targets to the novel compounds. AI and TF methods in association with human expertise may indeed revolutionize the current theranostic strategies, meanwhile, validation approaches are necessary to overcome the potential challenges and ensure higher accuracy. In this review, the significance of AI and TF in the development of drugs and delivery systems and the potential challenging issues have been highlighted.

Credibility Evidence for Computational Patient Models Used in the Development of Physiological Closed-Loop Controlled Devices for Critical Care Medicine

Physiological closed-loop controlled medical devices automatically adjust therapy delivered to a patient to adjust a measured physiological variable. In critical care scenarios, these types of devices could automate, for example, fluid resuscitation, drug delivery, mechanical ventilation, and/or anesthesia and sedation. Evidence from simulations using computational models of physiological systems can play a crucial role in the development of physiological closed-loop controlled devices; but the utility of this evidence will depend on the credibility of the computational model used. Computational models of physiological systems can be complex with numerous non-linearities, time-varying properties, and unknown parameters, which leads to challenges in model assessment. Given the wide range of potential uses of computational patient models in the design and evaluation of physiological closed-loop controlled systems, and the varying risks associated with the diverse uses, the specific model as well as the necessary evidence to make a model credible for a use case may vary. In this review, we examine the various uses of computational patient models in the design and evaluation of critical care physiological closed-loop controlled systems (e.g., hemodynamic stability, mechanical ventilation, anesthetic delivery) as well as the types of evidence (e.g., verification, validation, and uncertainty quantification activities) presented to support the model for that use. We then examine and discuss how a credibility assessment framework (American Society of Mechanical Engineers Verification and Validation Subcommittee, V&V 40 Verification and Validation in Computational Modeling of Medical Devices) for medical devices can be applied to computational patient models used to test physiological closed-loop controlled systems.

Automated Cardiac Drug Infusion System Using Adaptive Fuzzy Neural Networks Controller

This paper presents a fuzzy neural network (FNN) control system to automatically manage the hemodynamic variables of patients with hypertension and congestive heart failure (CHF) via simultaneous infusion of cardiac drugs such as vasodilators and inotropic agents. The developed system includes two FNN sub-controllers for regulating cardiac output (CO) and mean arterial pressure (MAP) by cardiac drugs, considering interactive pharmacological effects. The adaptive FNN controller was tested and evaluated on a cardiovascular model. Six short-term therapy conditions of hypertension and CHF are presented under different sensitivities of a vasodilator drug. The results of the automated system showed that root mean square errors were ≤ 5.56 mmHg and ≤ 0.22 L min-1 for regulating MAP and CO, respectively, providing short settling time responses of MAP (≤ 10.9 min) and CO (≤ 8.22 min) in all therapy conditions. The proposed FNN control scheme can significantly improve the performance of cardiac drug infusion System.

Accuracy enhancement in a fuzzy expert decision making system through appropriate determination of membership functions and its application in a medical diagnostic decision making system

The paper attempts to improve the accuracy of a fuzzy expert decision making system by tuning the parameters of type-2 sigmoid membership functions of fuzzy input variables and hence determining the most appropriate type-1 membership function. The current work mathematically models the variability of human decision making process using type-2 fuzzy sets. Moreover, an index of accuracy of a fuzzy expert system has been proposed and determined analytically. It has also been ascertained that there exists only one rule in the rule base whose associated mapping for the ith linguistic variable maps to the same value as the maximum value of the membership function for the ith linguistic variable. The improvement in decision making accuracy was successfully verified in a medical diagnostic decision making system for renal diagnostic applications. Based on the accuracy estimations applied over a set of pathophysiological parameters, viz. body mass index, glucose, urea, creatinine, systolic and diastolic blood pressure, appropriate type-1 fuzzy sets of these parameters have been determined assuming normal distribution of type-1 membership function values in type-2 fuzzy sets. The type-1 fuzzy sets so determined have been used to develop an FPGA based smart processor. Using the processor, renal diagnosis of patients has been performed with an accuracy of 98.75%.

A tool predicting future mean arterial blood pressure values improves the titration of vasoactive drugs

BACKGROUND

Vasoactive drug infusion rates are titrated to achieve a desired effect, e.g., mean arterial blood pressure (MAP), rather than using infusion rates based on body weight. The purpose of this study is to evaluate a method to automatically identify a patient's sensitivity to sodium-nitroprusside, dobutamine or dopamine and to evaluate, whether an advisory system that predicts MAP 5 min in the future enhances a clinician's ability to titrate sodium-nitroprusside infusions.

METHODS

We used published models implemented in MATLAB to simulate the response of 100 individual patients to infusions of sodium-nitroprusside, dopamine and dobutamine. The simulated patient's sensitivity to the three drugs was identified using an adaptive filter approach, where MAP was altered in a binary stepwise fashion. Next, 9 nurses were asked to control the MAP of 6 of the simulated patients. For half of the patients, we used the identified sensitivity to predict and display MAP 5 min into the future.

RESULTS

Identifying each individual patient's sensitivity improved the accuracy of the MAP prediction by 75% for sodium-nitroprusside, 82% for dopamine and 52% for dobutamine over the MAP prediction based on an "average" patient's sensitivity. The advisory system shortened the median time to reach the desired MAP from 10.2 to 4.1 min, decreased the median number of infusion rate changes from 6 to 4, and resulted in a significant reduction of mental workload and effort.

DISCUSSION

Patient-specific drug sensitivity identifi- cation significantly improved the prediction of future MAP. By predicting and displaying the expected MAP 5 min in the future, the advisory system helped nurses titrate faster, reduced their perceived workload and might improve patient safety.

ASIC design of a digital fuzzy system on chip for medical diagnostic applications

The paper presents the ASIC design of a digital fuzzy logic circuit for medical diagnostic applications. The system on chip under consideration uses fuzzifier, memory and defuzzifier for fuzzifying the patient data, storing the membership function values and defuzzifying the membership function values to get the output decision. The proposed circuit uses triangular trapezoidal membership functions for fuzzification patients' data. For minimizing the transistor count, the proposed circuit uses 3T XOR gates and 8T adders for its design. The entire work has been carried out using TSMC 0.35 µm CMOS process. Post layout TSPICE simulation of the whole circuit indicates a delay of 31.27 ns and the average power dissipation of the system on chip is 123.49 mW which indicates a less delay and less power dissipation than the comparable embedded systems reported earlier.

Automatic regulation of hemodynamic variables in acute heart failure by a multiple adaptive predictive controller based on neural networks

Automated drug-delivery systems that can tolerate various responses to therapeutic agents have been required to control hemodynamic variables with heart failure. This study is intended to evaluate the control performance of a multiple adaptive predictive control based on neural networks (MAPC_NN) to regulate the unexpected responses to therapeutic agents of cardiac output (CO) and mean arterial pressure (MAP) in cases of heart failure. The NN components in the MAPC_NN learned nonlinear responses of CO and MAP determined by hemodynamics of dogs with heart failure. The MAPC_NN performed ideal control against unexpected (1) drug interactions, (2) acute disturbances, and (3) time-variant responses of hemodynamics [average errors between setpoints (+35 ml kg⁻¹ min⁻¹ in CO and ±0 mmHg in MAP) and observed responses; 6.4, 3.7, and 4.2 ml kg⁻¹ min⁻¹ in CO and 1.6, 1.4, and 2.7 mmHg (10.5, 20.8, and 15.3 mmHg without a vasodilator) in MAP] during 120-min closed-loop control. The MAPC_NN could also regulate the hemodynamics in actual heart failure of a dog. Robust regulation of hemodynamics by the MAPC_NN was attributable to the ability of on-line adaptation to adopt various responses and predictive control using the NN. Results demonstrate the feasibility of applying the MAPC_NN using a simple NN to clinical situations.

Supervisory Control of Fuzzy Discrete Event Systems: A Formal Approach

Fuzzy discrete event systems (DESs) were proposed recently by Lin and Ying [19], which may better cope with the real-world problems of fuzziness, impreciseness, and subjectivity such as those in biomedicine. As a continuation of [19], in this paper, we further develop fuzzy DESs by dealing with supervisory control of fuzzy DESs. More specifically: 1) we reformulate the parallel composition of crisp DESs, and then define the parallel composition of fuzzy DESs that is equivalent to that in [19]. Max-product and max-min automata for modeling fuzzy DESs are considered, 2) we deal with a number of fundamental problems regarding supervisory control of fuzzy DESs, particularly demonstrate controllability theorem and nonblocking controllability theorem of fuzzy DESs, and thus, present the conditions for the existence of supervisors in fuzzy DESs; 3) we analyze the complexity for presenting a uniform criterion to test the fuzzy controllability condition of fuzzy DESs modeled by max-product automata; in particular, we present in detail a general computing method for checking whether or not the fuzzy controllability condition holds, if max-min automata are used to model fuzzy DESs, and by means of this method we can search for all possible fuzzy states reachable from initial fuzzy state in max-min automata. Also, we introduce the fuzzy n-controllability condition for some practical problems, and 4) a number of examples serving to illustrate the applications of the derived results and methods are described; some basic properties related to supervisory control of fuzzy DESs are investigated. To conclude, some related issues are raised for further consideration.

Reduction of interference in oscillometric arterial blood pressure measurement using fuzzy logic

In oscillometry, oscillation amplitudes (OAs) embedded in the cuff pressure are drastically affected by a variety of artifacts and cardiovascular diseases, leading to inaccurate arterial blood pressure (ABP) measurement. The purpose of this paper is to improve the accuracy in the arterial pressure measurement by reducing interference in the OAs using a recursive weighted regression algorithm (RWRA). This method includes a fuzzy logic discriminator (FLD) and a recursive regression algorithm. The FLD is used to reduce the effect of artifacts caused by measurement motion disturbance or cardiovascular diseases, and to determine the truthfulness of the oscillation pulse. According to the truth degree, the relationship between the cuff pressure and OA is reconstructed using the regression algorithm. Because the regression method must utilize inverse matrix operation, which will be difficult to implement in an automatic or ambulatory monitor, the recursive regression method is proposed to solve this problem. To test the performance of this RWRA, 47 subjects underwent the ABP measurement using both the auscultation and the oscillometry combined with the RWRA. It was found that the average difference between the pooled blood pressures measured by the auscultation and those by the oscillometry combined with the RWRA was found to be only 4.9 mmHg. Clinical results demonstrated that the proposed RWRA is more robust than the traditional curve fitting algorithm (TCFA). We conclude that the proposed RWRA can be applied to effectively improve the accuracy of the oscillometric blood pressure measurement.

Experimental studies on multiple-model predictive control for automated regulation of hemodynamic variables

A model-based control methodology was developed for automated regulation of mean arterial pressure and cardiac output in critical care subjects using inotropic and vasoactive drugs. The control algorithm used a multiple-model adaptive approach in a model predictive control framework to account for variability and explicitly handle drug rate constraints. The controller was experimentally evaluated on canines that were pharmacologically altered to exhibit symptoms of hypertension and depressed cardiac output. The controller performed better as compared to experiments on manual regulation of the hemodynamic variables. After the model bank was determined, mean arterial pressure was held within +/- 5 mm Hg 88.9% of the time with a standard deviation of 3.9 mm Hg. The cardiac output was held within +/- 1 l/min 96.1% of the time with a standard deviation of 0.5 l/min. The manual runs maintain mean arterial pressure only 82.3% of the time with a standard deviation of 5 mm Hg, and cardiac output 92.2% of the time with a standard deviation of 0.6 l/min.

Neuro-fuzzy closed-loop control of depth of anaesthesia

The utility of the auditory evoked potential (AEP) is under investigation as a feedback signal for the automatic closed-loop control of general anaesthesia using neural networks and fuzzy logic. The AEP is a signal derived from the electroencephalogram (EEG) in response to auditory stimulation, which may be useful as an index of the 'depth' of anaesthesia. A simple back-propagation neural network can learn the AEP and provides a satisfactory input to a fuzzy logic infusion controller for the administration of anaesthetic drugs, but the problem remains that of reliable signal acquisition.

Hemodynamic management of congestive heart failure by means of a multiple mode rule-based control system using fuzzy logic

A rule-based system was designed to control the mean arterial pressure (MAP) and the cardiac output (CO) of a patient with congestive heart failure (CHF), using two vasoactive drugs: sodium nitroprusside (SNP) and dopamine (DPM). The controller has three different modes, that engage according to the hemodynamic state. The critical conditions control mode (CCC) determines the initial infusion rates, and continues active if the MAP or the CO fall outside of the defined criticality thresholds: an upper and a lower boundary for the MAP and a lower boundary for the CO. Inside the boundaries the control is performed by noncritical conditions control modes (NCC's), which are fuzzy logic controllers. If the CO is within normal range and the MAP is close to the goal range, then the MAP is driven using only SNP, in a single-input-single-output mode (NCC-SISO). Otherwise the NCC multiple-input-multiple-output is active (NCC-MIMO). The goal values for the controlled variables are defined as a band of 5 mmHg for the MAP and 5 mL/kg/min for the CO, but there is little concern for this application if the CO is too high (i.e., in practical terms the CO only needs to achieve a necessary minimum rate). The NCC-MIMO includes a gain adaptation algorithm to cope with the wide variety in sensitivities to SNP. Supervisory capabilities to ensure adequate drug delivery complete the controller scheme. After extensive testing and tuning on a CHF-hemodynamics nonlinear model, the control system was applied in dog experiments, which led to further enhancements. The results show an adequate control, presenting a fast response to setpoint changes with an acceptable overshoot.