RICERCA ITALIANA     

Bioengineering of the Respiratory System

Università degli Studi di Bologna

Abstract

The quantitative evaluation of pulmonary mechanics, in patients undergoing mechanical ventilation, has a well recognised importance. In fact, in order to know the state of the patient and to estimate the therapeutic treatment adequacy, it is required to determine the value of parameters characterising the total respiratory functionality.
In this regard, the present research program aims at the development of theories, methods, and algorithms able to improve the knowledge of the physiopathological mechanisms of the respiratory functions and to help monitoring of the main viscoelastic properties of the respiratory mechanics in artificially ventilated patients. In particular, models previously developed will be improved so that they explicitly include physiopathological changes connected with COPD (Chronic Obstructive Pulmonary Disease), a pathology with a relevant socio-economical impact, since it represents the fourth leading cause of death in the Western countries. Therefore we intend to realize a simulation tool able to help the understanding of mechanisms producing expiratory flow limitation and the studying of its progression during the evolution of an obstructive disease. Moreover we intend to develop parameter estimation techniques for monitoring the patient status and optimizing the respiratory treatment.
The project is the natural continuation of a research activity already financed by MIUR in two previous periods 2000-2002 and 2002-2004, and it is characterized by a strong international impact. The project will comprise two strictly connected parts: one essentially methodological and one with a mainly applied character. The main aspects of the project will include:
1) The accomplishment of a nonlinear dynamic morphometric model of the tracheobronchial tree incorporating the main mechanisms responsible for expiratory flow limitation (EFL) and able to describe how the system behaves under mechanical ventilation in healthy subjects and COPD patients. This model will be used to address the role played by these mechanisms in COPD subjects and to establish the position of flow-limiting airway segments in the tracheobronchial tree. The model will be validated both in normal and COPD patients during respiratory treatment in the intensive care unit at the University of Siena. To this aim, a system based on the Negative Expiratory Pressure (NEP) technique, acquired in a previous financed period, will be used.
2) Development of parametric estimation techniques to monitor the patient status and to optimize the respiratory treatment. To this aim, starting from the simulation model previously developed ("direct model", see point 1), we intend to realize a new model ("inverse model") that allows physically significant parameters to be estimated from measurements available in the clinical practice.
The credibility of the program is guaranteed by the participation of investigators who have a large international relevance, and who exhibit a great experience on the subjects under study and on the techniques employed, as documented by several international publications produced in recent years (see references [10-14,16,29,31,38-40]). The necessary clinical competences are provided by professors Mario Chiavarelli and PierPaolo Giomarelli, members of Siena staff. <<<

Principal Investigator

Guido AVANZOLINI Università degli Studi di BOLOGNA

Research Objectives

The advance in basic knowledge on the respiratory system is generally recognized to be fundamental in order to improve the processes of diagnosis, surveillance and therapy. In particular, respiratory mechanics parameters in intensive care patients under assisted ventilation are components of so-called risk indices on which therapy and prognosis are based.
The present research program is the natural continuation of a research activity characterized by a strong international impact and already financed in the period 2002-2004 in the PRIN context. The main goal is the development of theories, methods, and algorithms suitable to improve the quantitative knowledge of the lung mechanics in patients affected by COPD (Chronic Obstructive Pulmonary Disease), with two precise purposes:
(A) to elucidate the mechanisms producing expiratory flow limitation and its progression during the evolution of an obstructive disease;
(B) to develop parameter estimation techniques for monitoring the patient status and optimizing the respiratory treatment.

More details on the intermediate objectives of part (A) and (B) are given at the end of this section.

COPD patients progressively develop pulmonary airflow limitation, which is not completely reversible. Expiratory flow limitation (EFL) leads to dynamic hyperinflation of the lungs and positive intrinsic end-expiratory pressure (PEEPi) which may require respiratory assistance. In this situation, mechanical ventilation is not without problems, since patients with EFL may have grave difficulty during weaning from the ventilator. Aiding interpretation of the non linear mechanisms producing EFL during assisted ventilation is a starting point for optimising ventilator strategy. Experimental studies are difficult to conduct in COPD patients during mechanical ventilation, due to the risks associated with critical patient condition. A modelling approach could therefore be convenient, especially if dynamic models of the tracheobronchial tree that take into account the main morphometric and mechanical characteristics of the system, can be used to simulate respiratory mechanics under physiopathological conditions.

As intermediate objectives, Section (A) includes:
A.1) The analysis of the non linear mechanisms determining expiratory flow limitation (EFL) and its evolution as well as the generation of positive end-expiratory pressure (PEEPi). We will make use of a non linear dynamic morphometric model of the tracheobronchial tree. The model will include both wave speed and viscous mechanisms determining flow limitation and will describe the system behaviour in artificial ventilation both in healthy and in COPD subjects. This activity will be carried out by the Unit of Bologna in cooperation with the Unit of Siena.
A.2) Identification of the tracheobronchial tree segments mainly involved in expiratory flow-limitation, when varying physiopathological conditions, and analysis of the possible correlation with the pathology location. This activity will be carried out by the Unit of Siena in cooperation with the Unit of Bologna.
A.3) Analysis of the effects of lung nonhomogeneity on the flow limitation. In fact, experimental and model studies demonstrated that nonhomogeneous constriction of airways accompany several respiratory diseases [15,32]. Therefore we intend to study what happens in a mechanically ventilated patient when only a certain part of lungs is affected by an obstructive pulmonary condition. In order to reach this objective, we will draw inspiration from both the symmetrical model proposed by Weibel and the asymmetrical one proposed by Horsfield and we will modify the model described in the point A.1 to allow the tracheobronchial tree to be nonhomogeneously characterized. This activity will be carried out by the Unit of Bologna in cooperation with the Unit of Siena.
A.4) Model validation, on the basis of clinical data attained by techniques routinely used in intensive care unit to ascertain the EFL. This activity will be carried out by the Unit of Siena in cooperation with the Unit of Bologna.

Section (B), developed in close collaboration by the Units of Bologna and Siena, includes the following intermediate objectives:
B.1) Development of a simplified non linear inverse model that could be usefully employed in expiratory flow-limited COPD patients for estimating their main parameters. For this purpose we will start from the encouraging results already described in literature, also by other research groups, which were obtained by using functional models incorporating non linear mechanisms capable to reproduce the flow-limitation phenomenon (such as for example the possibility of the airways to collapse) [8,10,14]. In particular, we will try to improve the structure of such functional model taking advantage of the availability of a detailed dynamic nonlinear morphometric forward simulator, which has been developed in the previous research activity and mentioned in the point A.1. Moreover, we will study the possibility of using the knowledge of other respiratory quantities available in the intensive care unit in addition to pressure and flow at the mouth (for example, the endopleural pressure), with the purpose of estimating directly single parameters of clinical interest, such as the resistance of lower airways in presence of flow limitation.
B.2) Test of the estimation algorithms and of the plausibility of the estimates. For this purpose we will utilize the dynamic morphometric simulator of the entire tracheo-bronchial tree to make a complex model to simple model comparison. This will be particularly convenient, since the values of the complex model parameters can be changed easily to simulate different respiratory conditions and therefore, since such a complex model is known in every detail, the physical and physiopathological interpretation of the parameters of the simple model used in the parameter estimation can also be studied.
B.3) Definition of guidelines to select from a set of models - the classic two-parameter model (total resistance and elastance), the three-parameter one (with different inspiratory and expiratory resistances), nonlinear functional models (capable to reproduce the flow limitation) - the most fit one in relation to the severity of the respiratory disease. For this purpose too, we will adopt a model to model approach.

The credibility of the program is guaranteed by the participation of investigators who, among the first in the world, developed on-line methods to estimate the respiratory mechanics parameters (see references [18-22,24-26,29] of the following scientific background), and it is amply supported by the several international publications produced during the two previous financed periods (see references [10-14,16,29,31,38-40] of the following scientific background). <<<

Timescale

24 months

National and international background

The two Units involved in the program have a consolidated experience in the proposed research themes, as illustrated in Models B. In particular, the above experience is documented by a number of international publications, reporting results that constitute a solid scientific background for the present program. Moreover, the same units have already worked in close cooperation in research projects concerning the "Bioengineering of the respiratory system", also in the PRIN, showing the ability to effectively coordinate their work and to take advantage of the different competencies and available resources to pursue a common goal.
The know-how possessed by the two Units is largely complementary, therefore we stress that the scientific collaboration of the two research units taking part in the project is indispensable for achieving the final objective. In particular, the Unit of Bologna will be the main reference for developing inverse mathematical models of respiratory mechanics, aimed at parameter identification. The Unit of Siena will constitute the main clinical reference, being a mixed group of Clinicians and Bioengineers working in a University Medical Department. Moreover, this Unit will work on the development of the simulation models of respiratory mechanics, in close cooperation with the Unit of Bologna.
In the last forty years, researches in the field of respiratory mechanics have led to significant improvements for comprehension, diagnosis and treatment of various respiratory diseases. Nowadays, thanks to methodological and technological advances as well as to the close collaboration between physiologists and engineers, several aspects of respiratory mechanics have been significantly clarified contributing to understanding the physiological bases of important pathologies. The Units involved in the Project have amply contributed to the study of respiratory physiopathology, and in particular, the work previously developed by the researchers of the two Units is summarized below:
- linear and nonlinear models of breathing mechanics able to reproduce different physiopathological conditions [3,7,10-11,14,16];
- on-line and off-line techniques for estimating respiratory parameters [17-25,29]; for example the utilisation of time-varying models [26] for the first time allowed the time course of respiratory resistance and elastance to be monitored within the same respiratory cycle;
- physico-mathematical models of respiratory gas transport and of ventilation control [30-31], later applied to the analysis of physiological quantities measured during intensive care [33-37], to study acid-base status in patients undergoing hemodialysis therapy [38] and to study the response to changes in oxygen (O2) and carbon dioxide (CO2) concentrations in arterial blood (hypoxia, hypercapnia, etc.) [39-40].

However, several problems requiring further investigation still exist. In particular, studying chronic obstructive pulmonary disease (COPD) is a matter of great interest. Indeed, today this pathology is probably the fourth cause of death in the Western countries, as well as cause of frequent hospitalisation and chronic invalidity. COPD is characterised by obstruction of the airways, which progressively produces a limitation of the expiratory flow (EFL). Moreover, a deeper understanding of the processes causing dynamic hyperinflation (DH) appears necessary. Indeed, DH produces a significant increase in intrinsic end-expiratory pressure (PEEPi) which may induce barotrauma, in patients subjected to mechanical ventilation, and may have negative effects on hemodynamics [5-6]. Because of the relevant socio-economical impact of COPD, studying mechanisms causing its onset and evolution has not only a knowledge value, but also it may have important clinical implications, for example in personalizing and optimizing respiratory assistance to critical patients [9].
A further problem of great clinical value concerns the formulation of mathematical models of reduced order, which may allow the main parameters of respiratory mechanics to be estimated, also in the presence of nonlinear phenomena, such as those responsible for flow-limitation.
The present research program represents an attempt to overcome, at least in part, such limitations and constitutes the natural continuation of programs financed in previous periods (2000-2002 and 2002-2004). <<<