Adjustment of positive end-expiratory pressure (PEEP) is a core component of lung-protective ventilation in acute respiratory distress syndrome (ARDS). The conventional invasive method for guiding PEEP by measuring transpulmonary pressure uses an esophageal catheter. This proof-of-concept study investigated whether noninvasive regional information from electrical impedance tomography (EIT) could estimate the transpulmonary pressure at which alveolar collapse begins (TPP = 0 mbar) and thus serve as a surrogate measure to guide PEEP.
The experimental series included 14 pigs. ARDS was induced by repeated lung lavage followed by a 2-hour period of ventilator-induced lung injury (VILI) ventilation to create an injurious lung condition. The study performed measurements both in healthy lungs and after the ARDS induction to compare the behaviour of regional EIT-derived parameters across lung conditions.
During a descending pressure-controlled ramp from 50 to 0 mbar, the investigators recorded airway and esophageal pressures and acquired EIT images. Using the EIT data, they constructed regional pressure–flow curves at the pixel level for a candidate dependent lung region. From these regional curves they calculated deflation parameters and identified characteristic points intended to mark changes in regional aeration and the onset of alveolar collapse (derecruitment).
Transpulmonary pressure (TPP) in this study was defined and calculated as the difference between end-expiratory airway pressure and the corresponding esophageal pressure obtained via an esophageal catheter (the invasive reference method cited from prior clinical work). The PEEP corresponding to a TPP of 0 mbar was used as the comparator for the EIT-derived characteristic point associated with the beginning of collapse.
Analysis of the descending pressure ramp revealed that the beginning of alveolar collapse in the dependent region of the EIT images could be determined at approximately 25 mbar based on regional characteristic points in both healthy and injured lung conditions. Quantitative correlation results reported in the abstract were:
Healthy lungs: a strong correlation between the second characteristic point (from regional deflation curves) and the PEEP corresponding to TPP = 0 mbar (r = 0.929, B = 1.005, R2 = 0.862, p = 0.007).
ARDS lungs: an intermediate correlation for the same relationship (r = 0.817, B = 0.894, R2 = 0.668, p = 0.025).
These statistics indicate the second regional characteristic point derived from EIT pressure–flow data aligned closely with invasively measured TPP-derived PEEP in healthy lungs and showed a meaningful but weaker association in the ARDS model.
The findings demonstrate that EIT can identify regional characteristic points in the dependent lung region that correlate with regional TPP changes. This suggests a potential pathway to titrate PEEP guided by noninvasive, regional EIT metrics rather than relying solely on invasive esophageal pressure measurements. Such an approach could support optimization of lung-protective ventilation by targeting the pressure threshold for collapse at a regional level.
The abstract does not provide detailed limitations, long-term outcomes, or stepwise procedures for clinical implementation. Sample-size rationale, possible confounders, technical specifics of the pixel-level curve derivation, and how regional characteristic points were defined algorithmically are not reported in the abstract. Full-text access would be required for these methodological details and for a comprehensive assessment of translational readiness.
In a porcine proof-of-concept ARDS model, regional deflation parameters derived from EIT pressure–flow curves correlated with invasively measured TPP-derived PEEP, particularly in healthy lungs and to a lesser degree in injured lungs. The study supports further investigation into EIT-guided PEEP titration as a noninvasive strategy to identify regional thresholds for alveolar collapse and to refine lung-protective ventilation. Additional work will be needed to report full methods, validate the approach across larger samples and diverse injury patterns, and define clinical protocols for human translation.
Note: The abstract is the source of all statements above. Details not reported in the abstract (for example, specific algorithmic steps, full methodological parameters, or limitations beyond those noted) are not available in the source document and therefore were not added here. The article is published under a Creative Commons Attribution license according to the PubMed record.