Enhancement of Protective Ventilation Strategies Using Electrical Impedance Tomography

The individualization of protective ventilation strategies for patients suffering from acute respiratory failure requires permanent bedside monitoring of pulmonary parameters and an immediate response to changing lung conditions. This can be achieved using the electrical impedance tomography (EIT), a fast, non-invasive technique to observe global and regional lung properties corresponding to changes in the conductivity distribution of the thorax. In this work, EIT has been evaluated in terms of its potential to assess different conditions of the lung during lung recruitment maneuvers (RM) which are frequently used for the treatment of respiratory failure. Therefore, an EIT system was integrated into a fuzzy-controller based medical expert system which is capable of automatically performing RM while recording physiological parameters. The setup additionally triggers a CT scanner to obtain reference images. Six anesthetized, artificially ventilated pigs received repetitive lung lavages with NaCl-solution to induce respiratory failure. In a first trial, PEEP was step-wise increased and subsequently decreased to examine the influence of PEEP on the patient. In a second trial, the expert system performed an automatic RM. Differential and functional EIT images were continuously recorded and compared to CT images made at the end of inspiration and expiration. Global impedance and mean lung density as well as tidal volume and tidal variation demonstrated a high linear correlation. However, the proportionality factor between the respective quantities changed with the state of lung recruitment. This was probably due to the grossly changing geometry of the conductivity distribution. During the automated RM, a clear shift of ventilation from dorsal to ventral regions of the lung could be observed in both, EIT and CT images. The experiments showed that EIT is capable of identifying dynamic changes in the lung such as global and regional recruitment and may provide important control variables for closed-loop protective ventilation controllers.