**Lung Ventilation Modeling for Assessment of Lung Status: Detection of Lung Disease and Indication for Extubation of Mechanically-Ventilated COPD Patients**

Dhanjoo N. Ghista1, Kah Meng Koh2, Rohit Pasam3 and Yi Su4 *1Department of Graduate and Continuing Education, Framingham State University, Framingham, Massachusetts, 2VicWell BioMedical, 3Quodient, Inc, 4Institute of High Performance Computing, Agency for Science, Technology and Research, 1,3USA 2,4Singapore* 

### **1. Introduction**

830 Biomedical Science, Engineering and Technology

[6] Loh K M, Ng David, Ghista D N, Ridolph H. Quantitation of renal function based on two-compartmental modeling of renal pelvis. IEEE Conference 2005. [7] Mazumdar J. An Introduction to Mathematical Physiology & Biology. Cambridge

University Press. Second edition 1999.

In pulmonary medicine, it is important to detect lung diseases, such as chronic obstructive pulmonary disease (COPD), emphysema, lung fibrosis and asthma. These diseases are characterized in terms of lung compliance and resistance-to-airflow. Another important endeavour of pulmonary medicine is mechanical ventilation of COPD patients and determining when to wean off these patients from the mechanical ventilator. In both these medical domains, lung ventilation dynamics plays a key role.

So in this chapter, we develop the lung ventilation dynamics model in terms of monitored lung volume (*V*) and driving pressure (*PL*), in the form of a differential equation with parameters of lung compliance (*C*) and resistance-to-airflow (*R*). We obtain the solution of this equation in the forms of lung volume (*V*) function of *PL*, *C* and *R*. For the monitored lung volume *V* and pressure *PL* data, we can evaluate *C* and *R* by matching the model solution expression with the monitored lung volume *V* and driving pressure *PN* data. So what we have done here is to develop the method for determining lung compliance (*C*) and resistance-toairflow (*R*) as average values of *C* (= *Ca*) and *R* (= *Ra*) during the ventilation cycle.

Now in some lung diseases such as in emphysema, the lung compliance (*C*) is high. In other lung diseases such as in asthma, the airway resistance (*R*) is high. So we need to determine the ranges of *C* and *R* for normal lung status as well as for lung disease states. Then, we can develop a 2-parameter *R*-*C* diagnostic coordinate plane, on which we can demarcate the zones for different diseases. Then, for any patient's (*R*, *C*) value, we can plot the (*R*, *C*) point in the *R*-*C* diagnostic coordinate plane, and based on its location designate the lung disease state of the patient. A more convenient way for detecting lung disease is to combine *R* and *C* along with some ventilator data (such as tidal volume and breathing rate) into a nondimensional lung ventilator index (*LVI*). Then, we can determine the ranges of *LVI* for normal and disease states, and thereby employ the patient's computed values of *LVI* to designate a specific lung disease for the patient. The *LVI* concept for detecting lung disease is more convenient to adopt in clinical practice, because it enables detection of lung disease states in the form of just one lung-ventilation number.

Now, in this methodology, we need to monitor (i) lung volume, by means of a spirometer, and (ii) lung pressure (*PL*) equal to *Pmo* (pressure at mouth) minus pleural pressure (*Pp*). The pleural pressure measurement involves placing a balloon catheter transducer through the nose into the esophagus, whereby the esophageal tube pressure is assumed to be equal to the pressure in the pleural space surrounding it. Now this procedure cannot be carried out non-traumatically and routinely in patients. Hence, for routineand noninvasive assessment of lung ventilation for detection of lung disease states, it is necessary to have a method for determining *R* and *C* from only lung volume data. So, then, we have shown how we can compute *R*, *C* and lung pressure values non-invasively from just lung volume measurement.

Finally, we have presented how the lung ventilation modeling can be applied to study the lung ventilation dynamics of COPD patients on mechanical ventilation. We have shown how a COPD patient's lung *C* and *R* can be evaluated in terms of the monitored lung volume and applied ventilatory pressure. We have also formulated another lung ventilator index to study and assess the lung status improvement of COPD patients on mechanical ventilation, and to decide when they can be weaned off mechanical ventilation.
