Teresa Mendonça - Academia.edu (original) (raw)
Papers by Teresa Mendonça
On-line controller autocalibration based on parameter predictors:
2009 European Control Conference (ECC), 2009
The paper considers the control of neuromuscular blockade (NMB) induced by atracurium on patients... more The paper considers the control of neuromuscular blockade (NMB) induced by atracurium on patients subject to general surgery. The high levels of uncertainty in the process model are tackled with supervised multi-model adaptive control, for which an innovative supervisor based on least squares estimation is presented to deal with constant unknown offsets in output measurement. In order to avoid excessive switching among the controllers, a model clustering is made, with a common controller in each cluster. Furthermore, an antiwindup algorithm that may be applied to an integrator that is common to a bank of controllers is proposed. The control structure is tested on simulation using realistic nonlinear models for NMB with parameters estimated from actual patient data.
2007 European Control Conference (ECC), 2007
This paper presents a contribution to the use of Hopfield neural networks (HNNs) for parameter es... more This paper presents a contribution to the use of Hopfield neural networks (HNNs) for parameter estimation. Our focus is on time-invariant systems that are linear in the parameters. We introduce a suitable HNN and present a weaker condition than the currently existing ones that guarantees the convergence of the parameterization estimated by the network to the actual parameterization. The application of our results is illustrated in a parameter estimation problem for a two carts system.
IFAC-PapersOnLine, 2018
In this paper a new simplified control scheme for the neuromuscular blockade level that only requ... more In this paper a new simplified control scheme for the neuromuscular blockade level that only requires the knowledge of one model parameter is proposed. The control law is designed to track a desired target neuromuscular blockade level. Furthermore, the performance of this approach is compared with the results of a PI controller. The results were validated by simulations based on real data collected during surgeries.
IFAC-PapersOnLine, 2015
This paper provides an up-to-date review of intersecting know-how in engineering, medical and com... more This paper provides an up-to-date review of intersecting know-how in engineering, medical and computer science frameworks. The text interweaves available measures, models, controls and technology for drug delivery systems in general anesthesia. The three main actors, equally important, along with their role in overall performance are presented and analyzed from a global perspective. The availability of patient information for personalized healthcare is systematically analyzed and the possibility to integrate this hybrid information sources into personalized patient models are critically sought. Finally, an important role is given to the accurate, crisp and real-time information availability in medical devices. Modular integration of new items is of paramount importance in terms of space, computational burden and number of training hours.
IFAC-PapersOnLine, 2015
In this paper a state-feedback law for the control of the neuromuscular blockade level is present... more In this paper a state-feedback law for the control of the neuromuscular blockade level is presented. The control law is designed based on an optimal problem that is relaxed into a semidefinite program using a change of variable. For that purpose a parsimoniously parameterized model is used to describe the patient's response to a muscle relaxant. Due to clinical restrictions the controller action begins when the patient recovers after an initial drug bolus. The results obtained encourage the implementation of this controller in the clinical environment even in the presence of noise.
2015 23rd Mediterranean Conference on Control and Automation (MED), 2015
In this paper the neuromuscular blockade level and the bispectral index level tracking problems b... more In this paper the neuromuscular blockade level and the bispectral index level tracking problems by means of automatic control are considered in the context of general anesthesia. These tracking problems are formulated as optimal control problems that are numerically solved using direct methods. The obtained results are encouraging when compared with another strategy recently proposed in the literature.
Proceedings of the 19th IFAC World Congress, 2014
In this paper, a retuning strategy for a controller is proposed in order to improve the reference... more In this paper, a retuning strategy for a controller is proposed in order to improve the reference BIS tracking in patients, by means of simultaneous administration of propofol and of remifentanil, in the presence of model uncertainties. This strategy proves to be useful as is shown by simulations.
Asian Journal of Control, 2014
An adaptive controller based on a minimally parameterized parsimonious Wiener model for the effec... more An adaptive controller based on a minimally parameterized parsimonious Wiener model for the effect of the muscle relaxant rocuronium in the neuromuscular blockade is presented. The controller structure combines inversion of the recursively identified static nonlinearity of the Wiener model with a positive compartmental control law for the linearized system. The overall strategy exploits the fact that the model has only two parameters, which are estimated by an extended Kalman filter. Due to the fact that the positive control law for total mass conservation of compartmental systems is only proven to be convergent for time-invariant systems, the identification of the parameter in the linear block of the minimally parameterized parsimonious Wiener model is stopped when the controller is turned on. The controller was implemented in the platform Galeno and tested in simulation and in thirteen real cases of patients under general anesthesia. The good reference tracking results and robustness to noise show the reliability of the proposed strategy.
2014 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2014
In this paper, the study of the robustness of the positive control law introduced in [1] in the p... more In this paper, the study of the robustness of the positive control law introduced in [1] in the presence of parameter uncertainties is made. This controller was developed to track a desired reference level for the BIS of a patient by means of the simultaneous administration of propofol and of remifentanil. Here, it is proven that in the presence of uncertainties in the BIS model, the controller still has a good performance and the BIS of the patient converges to clinically acceptable values. These results are illustrated by simulations.
ECAI 2014 - 21st European Conference on Artificial Intelligence, 2014
The emergence of real temporal applications under nonstationary scenarios has drastically altered... more The emergence of real temporal applications under nonstationary scenarios has drastically altered the ability to generate and gather information. Nowadays, under dynamic scenarios, potentially unbounded and massive amounts of information are generated at high-speed rate, known as data streams. Dealing with evolving data streams imposes the online monitoring of data in order to detect changes. The contribution of this paper is to present the advantage of using fading histograms to compare data distribution for change detection purposes. In an windowing scheme, data distributions provided by the fading histograms are compared using the Kullback-Leibler divergence. The experimental results support that the detection delay time is smaller when using fading histograms to represent data instead of standard histograms.
Progress in Artificial Intelligence, 2014
The ability to collect data is changing drastically. Nowadays, data is gathered in the form of tr... more The ability to collect data is changing drastically. Nowadays, data is gathered in the form of transient and finite data streams. Memory restrictions preclude keeping all received data in memory. When dealing with massive data streams, it is mandatory to create compact representations of data, also known as synopses structures or summaries. Reducing memory occupancy is of utmost importance when handling a huge amount of data. This paper addresses the problem of constructing histograms from data streams under error constraints. When constructing online histograms from data streams there are two main characteristics to embrace: the updating facility and the error of the histogram. Moreover, in dynamic environments, besides the need of compact summaries to capture the most important properties of data, it is also essential to forget old data. Therefore, this paper presents sliding histograms and fading histograms, an abrupt and a smooth strategies to forget outdated data.
IFAC Proceedings Volumes (IFAC-PapersOnline), 2011
This paper presents a compartmental model-based control strategy to drive the NeuroMuscular Block... more This paper presents a compartmental model-based control strategy to drive the NeuroMuscular Blockade level of patients undergoing general anesthesia to a predefined target. For that purpose, a compartmental realization of a minimally parameterized nonlinear Wiener model was derived and used on an adapted version of the standard compartmental control law for linear systems. The identification of the model parameters was recursively performed by one Extended Kalman Filter during the initial bolus induction period and stopped afterwards. To overcome the fact that this identification is stopped during the closed-loop control period, uncertainties in the parameters are assumed to be present and included in the control law. Information taken from the identification of real collected cases was used to tune the parameter uncertainties. The feasibility of the whole strategy was evaluated in a bank of simulated models, giving rise to good reference tracking results even in the presence of noise.
Proceedings of the IEEE Conference on Decision and Control, 2011
This paper presents a control strategy for exact reference tracking in the presence of parameter ... more This paper presents a control strategy for exact reference tracking in the presence of parameter uncertainty for the NeuroMuscular Blockade (NMB) level of patients undergoing general anesthesia. For this control application, a compartmental realization of a minimally parameterized Wiener model was used in [1] in which the parameters were estimated from a bank of collected real data using the Extended Kalman Filter (EKF). Due to the parameter uncertainty this procedure did not achieve the desired tracking goal. Here a modified control strategy is presented to overcome this drawback.
IFAC Proceedings Volumes (IFAC-PapersOnline), 2012
I am thankful to my friends, specially to Pedro Ferreira, Carlos Faria and Tiago Esteves for thei... more I am thankful to my friends, specially to Pedro Ferreira, Carlos Faria and Tiago Esteves for their fellowship, support, knowledge and for the entertainment they provided me. Last but not least, I would like to thank to Ana Bela, for being a complementation of myself, for being the last voice I heard during these last months when writing my MSc thesis, being always a great support and encouragement.
In this paper a new estimation procedure for the parameters of the PK/PD model for the simultaneo... more In this paper a new estimation procedure for the parameters of the PK/PD model for the simultaneous administration of propofol and remifentanil introduced in [1] is presented. This model has the advantage of being parsimoniously parameterized, which allows a simple parameter estimation procedure, based on the patient's step response. It is shown that the parameter estimates obtained in this way provide good results when used to tune the positive control law introduced in [2].
This paper addresses the problem of joint estimation of the state and parameters for a determinis... more This paper addresses the problem of joint estimation of the state and parameters for a deterministic continuous time system, with discrete time observations, in which the parameter vector is constant but its value is not known, being a random variable with a known distribution. Along time, the uncertainty in the parameter induces uncertainty in the plant state. The joint probability density function (pdf) satisfies the Liouville partial differential equation that is a limit case of the Fokker-Planck equation for vanishing diffusion. The continuous-discrete filter proposed operates as follows: Between two consecutive output sampling time instants, the pdf is propagated by solving the Liouville equation for an augmented state and is then corrected by using the last observation and Bayes law. An application to state estimation of the neuromuscular blockade of patients subject to general anesthesia, where parameter uncertainty is due to inter-patient variability, is described.
AdvancesinSystems Science, SpringerInternationalPublishing, 2014
This work addresses the problem of computing the time evolution of the probability density functi... more This work addresses the problem of computing the time evolution of the probability density function (pdf) of the state in a nonlinear neuromuscular blockade (NMB) model, assuming that the source of uncertainty is the knowledge about one parameter. The NMB state is enlarged with the parameter, that verifies an equation given by its derivative being zero and has an initial condition described by a known pdf. By treating the resulting enlarged state-space model as a stochastic differential equation, the pdf of the state verifies a special case of the Fokker-Planck equation in which the second derivative terms vanish. This partial differential equation is solved with a numerical method based on Trotter's formula for semigroup decomposition. The method is illustrated with results for a reduced complexity NMB model. A comparison of the predicted state pdf with clinical data for real patients is provided.
2007 Annual International Conference of the Ieee Engineering in Medicine and Biology Society, Vols 1-16, 2007
![Research paper thumbnail of Robust Control of Maintenance-Phase Anesthesia [Applications of Control]](https://attachments.academia-assets.com/110157327/thumbnails/1.jpg)
](IEEE Control Systems, 2014
On-line controller autocalibration based on parameter predictors:
2009 European Control Conference (ECC), 2009
The paper considers the control of neuromuscular blockade (NMB) induced by atracurium on patients... more The paper considers the control of neuromuscular blockade (NMB) induced by atracurium on patients subject to general surgery. The high levels of uncertainty in the process model are tackled with supervised multi-model adaptive control, for which an innovative supervisor based on least squares estimation is presented to deal with constant unknown offsets in output measurement. In order to avoid excessive switching among the controllers, a model clustering is made, with a common controller in each cluster. Furthermore, an antiwindup algorithm that may be applied to an integrator that is common to a bank of controllers is proposed. The control structure is tested on simulation using realistic nonlinear models for NMB with parameters estimated from actual patient data.
2007 European Control Conference (ECC), 2007
This paper presents a contribution to the use of Hopfield neural networks (HNNs) for parameter es... more This paper presents a contribution to the use of Hopfield neural networks (HNNs) for parameter estimation. Our focus is on time-invariant systems that are linear in the parameters. We introduce a suitable HNN and present a weaker condition than the currently existing ones that guarantees the convergence of the parameterization estimated by the network to the actual parameterization. The application of our results is illustrated in a parameter estimation problem for a two carts system.
IFAC-PapersOnLine, 2018
In this paper a new simplified control scheme for the neuromuscular blockade level that only requ... more In this paper a new simplified control scheme for the neuromuscular blockade level that only requires the knowledge of one model parameter is proposed. The control law is designed to track a desired target neuromuscular blockade level. Furthermore, the performance of this approach is compared with the results of a PI controller. The results were validated by simulations based on real data collected during surgeries.
IFAC-PapersOnLine, 2015
This paper provides an up-to-date review of intersecting know-how in engineering, medical and com... more This paper provides an up-to-date review of intersecting know-how in engineering, medical and computer science frameworks. The text interweaves available measures, models, controls and technology for drug delivery systems in general anesthesia. The three main actors, equally important, along with their role in overall performance are presented and analyzed from a global perspective. The availability of patient information for personalized healthcare is systematically analyzed and the possibility to integrate this hybrid information sources into personalized patient models are critically sought. Finally, an important role is given to the accurate, crisp and real-time information availability in medical devices. Modular integration of new items is of paramount importance in terms of space, computational burden and number of training hours.
IFAC-PapersOnLine, 2015
In this paper a state-feedback law for the control of the neuromuscular blockade level is present... more In this paper a state-feedback law for the control of the neuromuscular blockade level is presented. The control law is designed based on an optimal problem that is relaxed into a semidefinite program using a change of variable. For that purpose a parsimoniously parameterized model is used to describe the patient's response to a muscle relaxant. Due to clinical restrictions the controller action begins when the patient recovers after an initial drug bolus. The results obtained encourage the implementation of this controller in the clinical environment even in the presence of noise.
2015 23rd Mediterranean Conference on Control and Automation (MED), 2015
In this paper the neuromuscular blockade level and the bispectral index level tracking problems b... more In this paper the neuromuscular blockade level and the bispectral index level tracking problems by means of automatic control are considered in the context of general anesthesia. These tracking problems are formulated as optimal control problems that are numerically solved using direct methods. The obtained results are encouraging when compared with another strategy recently proposed in the literature.
Proceedings of the 19th IFAC World Congress, 2014
In this paper, a retuning strategy for a controller is proposed in order to improve the reference... more In this paper, a retuning strategy for a controller is proposed in order to improve the reference BIS tracking in patients, by means of simultaneous administration of propofol and of remifentanil, in the presence of model uncertainties. This strategy proves to be useful as is shown by simulations.
Asian Journal of Control, 2014
An adaptive controller based on a minimally parameterized parsimonious Wiener model for the effec... more An adaptive controller based on a minimally parameterized parsimonious Wiener model for the effect of the muscle relaxant rocuronium in the neuromuscular blockade is presented. The controller structure combines inversion of the recursively identified static nonlinearity of the Wiener model with a positive compartmental control law for the linearized system. The overall strategy exploits the fact that the model has only two parameters, which are estimated by an extended Kalman filter. Due to the fact that the positive control law for total mass conservation of compartmental systems is only proven to be convergent for time-invariant systems, the identification of the parameter in the linear block of the minimally parameterized parsimonious Wiener model is stopped when the controller is turned on. The controller was implemented in the platform Galeno and tested in simulation and in thirteen real cases of patients under general anesthesia. The good reference tracking results and robustness to noise show the reliability of the proposed strategy.
2014 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2014
In this paper, the study of the robustness of the positive control law introduced in [1] in the p... more In this paper, the study of the robustness of the positive control law introduced in [1] in the presence of parameter uncertainties is made. This controller was developed to track a desired reference level for the BIS of a patient by means of the simultaneous administration of propofol and of remifentanil. Here, it is proven that in the presence of uncertainties in the BIS model, the controller still has a good performance and the BIS of the patient converges to clinically acceptable values. These results are illustrated by simulations.
ECAI 2014 - 21st European Conference on Artificial Intelligence, 2014
The emergence of real temporal applications under nonstationary scenarios has drastically altered... more The emergence of real temporal applications under nonstationary scenarios has drastically altered the ability to generate and gather information. Nowadays, under dynamic scenarios, potentially unbounded and massive amounts of information are generated at high-speed rate, known as data streams. Dealing with evolving data streams imposes the online monitoring of data in order to detect changes. The contribution of this paper is to present the advantage of using fading histograms to compare data distribution for change detection purposes. In an windowing scheme, data distributions provided by the fading histograms are compared using the Kullback-Leibler divergence. The experimental results support that the detection delay time is smaller when using fading histograms to represent data instead of standard histograms.
Progress in Artificial Intelligence, 2014
The ability to collect data is changing drastically. Nowadays, data is gathered in the form of tr... more The ability to collect data is changing drastically. Nowadays, data is gathered in the form of transient and finite data streams. Memory restrictions preclude keeping all received data in memory. When dealing with massive data streams, it is mandatory to create compact representations of data, also known as synopses structures or summaries. Reducing memory occupancy is of utmost importance when handling a huge amount of data. This paper addresses the problem of constructing histograms from data streams under error constraints. When constructing online histograms from data streams there are two main characteristics to embrace: the updating facility and the error of the histogram. Moreover, in dynamic environments, besides the need of compact summaries to capture the most important properties of data, it is also essential to forget old data. Therefore, this paper presents sliding histograms and fading histograms, an abrupt and a smooth strategies to forget outdated data.
IFAC Proceedings Volumes (IFAC-PapersOnline), 2011
This paper presents a compartmental model-based control strategy to drive the NeuroMuscular Block... more This paper presents a compartmental model-based control strategy to drive the NeuroMuscular Blockade level of patients undergoing general anesthesia to a predefined target. For that purpose, a compartmental realization of a minimally parameterized nonlinear Wiener model was derived and used on an adapted version of the standard compartmental control law for linear systems. The identification of the model parameters was recursively performed by one Extended Kalman Filter during the initial bolus induction period and stopped afterwards. To overcome the fact that this identification is stopped during the closed-loop control period, uncertainties in the parameters are assumed to be present and included in the control law. Information taken from the identification of real collected cases was used to tune the parameter uncertainties. The feasibility of the whole strategy was evaluated in a bank of simulated models, giving rise to good reference tracking results even in the presence of noise.
Proceedings of the IEEE Conference on Decision and Control, 2011
This paper presents a control strategy for exact reference tracking in the presence of parameter ... more This paper presents a control strategy for exact reference tracking in the presence of parameter uncertainty for the NeuroMuscular Blockade (NMB) level of patients undergoing general anesthesia. For this control application, a compartmental realization of a minimally parameterized Wiener model was used in [1] in which the parameters were estimated from a bank of collected real data using the Extended Kalman Filter (EKF). Due to the parameter uncertainty this procedure did not achieve the desired tracking goal. Here a modified control strategy is presented to overcome this drawback.
IFAC Proceedings Volumes (IFAC-PapersOnline), 2012
I am thankful to my friends, specially to Pedro Ferreira, Carlos Faria and Tiago Esteves for thei... more I am thankful to my friends, specially to Pedro Ferreira, Carlos Faria and Tiago Esteves for their fellowship, support, knowledge and for the entertainment they provided me. Last but not least, I would like to thank to Ana Bela, for being a complementation of myself, for being the last voice I heard during these last months when writing my MSc thesis, being always a great support and encouragement.
In this paper a new estimation procedure for the parameters of the PK/PD model for the simultaneo... more In this paper a new estimation procedure for the parameters of the PK/PD model for the simultaneous administration of propofol and remifentanil introduced in [1] is presented. This model has the advantage of being parsimoniously parameterized, which allows a simple parameter estimation procedure, based on the patient's step response. It is shown that the parameter estimates obtained in this way provide good results when used to tune the positive control law introduced in [2].
This paper addresses the problem of joint estimation of the state and parameters for a determinis... more This paper addresses the problem of joint estimation of the state and parameters for a deterministic continuous time system, with discrete time observations, in which the parameter vector is constant but its value is not known, being a random variable with a known distribution. Along time, the uncertainty in the parameter induces uncertainty in the plant state. The joint probability density function (pdf) satisfies the Liouville partial differential equation that is a limit case of the Fokker-Planck equation for vanishing diffusion. The continuous-discrete filter proposed operates as follows: Between two consecutive output sampling time instants, the pdf is propagated by solving the Liouville equation for an augmented state and is then corrected by using the last observation and Bayes law. An application to state estimation of the neuromuscular blockade of patients subject to general anesthesia, where parameter uncertainty is due to inter-patient variability, is described.
AdvancesinSystems Science, SpringerInternationalPublishing, 2014
This work addresses the problem of computing the time evolution of the probability density functi... more This work addresses the problem of computing the time evolution of the probability density function (pdf) of the state in a nonlinear neuromuscular blockade (NMB) model, assuming that the source of uncertainty is the knowledge about one parameter. The NMB state is enlarged with the parameter, that verifies an equation given by its derivative being zero and has an initial condition described by a known pdf. By treating the resulting enlarged state-space model as a stochastic differential equation, the pdf of the state verifies a special case of the Fokker-Planck equation in which the second derivative terms vanish. This partial differential equation is solved with a numerical method based on Trotter's formula for semigroup decomposition. The method is illustrated with results for a reduced complexity NMB model. A comparison of the predicted state pdf with clinical data for real patients is provided.
2007 Annual International Conference of the Ieee Engineering in Medicine and Biology Society, Vols 1-16, 2007
IEEE Control Systems, 2014