Back to search

National Technical University of Athens (NTUA)

Μοντελοποίηση και έλεγχος βιολογικών και φυσιολογικών συστημάτων

Abstract

dc:description

The problem of drug administration is purely a control problem where the target is to keep the drug concentration at certain organs in the body close to the desired therapeutic set-points while the concentration in other organs and tissues does not exceed certain safety limits. The manipulated variable is the administered dosage and the controlled variable is the concentration of the drug in some tissue or organ of the body. However, unlike most engineering systems, the human body is a system which we have limited monitoring and intervention capabilities for. In order to design meaningful administration policies one needs to limit the frequency of (oral) administration and feedback sampling (i.e. how often the patient visits the medical centre to check on the progress of their therapy). All these restrictions call for novel control methods and motivated our work.Standard methods stemming from Digital Systems Theory, such as the Finite Impulse Response (FIR) filter are introduced in Pharmacokinetic Modelling and are used for the first time to formulate and to solve an optimal control problem for the oral administration of drugs. An FIR model derived from experimental pharmacokinetic data is employed to correlate the administered dose with the concentration-time profiles of the drug in the compartments (organs) of the body. The introduction of integer variables and the formulation of an Integer Programming mathematical optimisation problem yields solutions (administration sequences) which satisfy the administration frequency and dose availability constraints. The proposed method is extended to account for the stochastic variant of the optimal control problem sothat it can be used in practical applications where large populations of patients are to be treated. We use the proposed method to calculate the optimal administration course of L-dopa to a population of patients suffering from the Parkinson’s disease.Next, we investigate how Fractional Dynamical Systems help us fathom the interplay among the main factors of drug distribution. They allow us to draw individualised concentration-time profiles and implement feedback settings to control the concentration of the administered drug. Fractional Dynamics are used by many authors to describe processes with memory such as drug distribution when anomalous diffusion and deep tissue trapping phenomena are involved. A Fractional-PID controller is designed for the i.v. administration of Amiodarone – an anti-arrhythmic agent – and the dynamical characteristics of the closed-loop system are studied in the frequency domain.We identified an inefficiency that arises when designing controllers for constrained systems in the discrete-time domain. It happens that when such a controller is applied to the actual (continuous-time) system, the constraints are in for violation. Designing controllers that satisfy the constraints is of high importance when it comes to drug administration in order to avoid toxic side-effects. Using methods of polytopic over-approximation of matrix-exponential functions we reformulate the sampled-data MPC problem so that the resulting closed-loop system satisfied the constraints.Linear Impulsive Control Systems have been extensively studied with respect to their equilibrium points which, in most cases, are no other than the origin. As a result the trajectory of the system cannot be stabilised to arbitrary desired points. This imposes a significant restriction towards their utilisation in drug administration. In this dissertation, we study the equilibrium of Linear Impulsive Systems in light of target-sets instead of the standard equilibrium point approach. We properly extend the notion of invariant sets which is crucial in designing asymptotically stable Model Predictive Controllers (MPC).Finally, a distributed computational network for building (Q)SAR models to predict biological endpoints was developed within the FP7-funded EU research project “OpenTox”. OpenTox enables us to predict pharmacokinetic parameters and contributes to establishing an integrated computer-aided drug design framework. Overall, the achievements of this dissertation generously contribute to this direction.At the meeting point of Medicine, Biology and Engineering there are unprecedented opportunities for research and innovation as Systems and Control Theory can help towards making the best out of a pharmaceutical compound.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2012

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Sopasakis, Pantelis
  • Σωπασάκης, Παντελής

Subjects

dc:subject × 19

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/38677

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Sopasakis, Pantelis; Σωπασάκης, Παντελής. Μοντελοποίηση και έλεγχος βιολογικών και φυσιολογικών συστημάτων. National Technical University of Athens (NTUA), 2012. http://hdl.handle.net/10442/hedi/38677