| dc.description.abstract | Protective ventilation strategies have been proposed to minimize the occurrence of lung overdistension (OD), as well as the presence of tidal recruitment (TR) of airways and alveolar units. In this context, Electrical Impedance Tomography (EIT) emerges as a noninvasive, radiation-free imaging technique, for bedside monitoring the regional distribution of ventilation in the lungs. Currently, different EIT methods exist to quantify TR and OD, among them the PVShape quantifies the curvature of the pressure-volume curve during inspiration (PVShapeinsp) and expiration (PVShapeexp); the algorithm RACH (recruitable alveolar colapse and hyperdistension, RACH) is based on the estimation of the compliance changes in each element through the PEEP steps. Thus, the objective of the present study is to compare these published algorithms in order to establish whether they are equivalent in the quantification of changes in spatial distribution of OD and TR during mechanical ventilation. Therefore, EIT was collected from 20 mechanically ventilated healthy patients during a decremental positive end-expiratory pressure (PEEP) titration maneuver. Thereby, the proportions of TR and OD at each PEEP step were computed from EIT according with each algorithm. The results show that the estimations of TR and OD was significantly affected by PEEP changes. In addition, compared to the other methods, the estimates of TR were considerably higher with PVShapeinsp (more than 60% larger, p<0.001). While the higher proportion of OD was found in the PVShapeexp (more than 6% larger, p<0.001). According to the findings, this study demonstrates a significant discrepancy between the proportions of TR and OD quantified by these three EIT-based methods, and, in general, a disagreement between the inferences that can be deducted by each of these are not physiologically comparable. Nevertheless, the changes in the threshold used for the classification of PVShape regions indicated that the difference between the estimations of PVShapeexp and RACH are partially related to the threshold values initially used. | |
