Fig. 3

Effect of prone position on ventilation/perfusion (V/Q) matching in two representative patients. From left to right, lung ventilation (blue–white gradient area), perfusion (red–yellow area), and V/Q matching are depicted. First row, supine position. Second row, prone position. First column, functional EIT image shows tidal ventilation distribution (blue–white gradient area). Second column, functional EIT image shows perfusion distribution (red–yellow area). Third column, functional EIT image shows the distribution of regional V/Q matching. Ventilated regions were defined as pixels with impedance changes higher than 20% of the maximum tidal impedance variation in the functional ventilation image. Perfused regions were defined as pixels higher than 20% of the maximum bolus-related impedance change in the functional perfusion image. Regions with high ventilation and low perfusion are marked in blue (denoted as dead space), low ventilation and high-perfusion regions in red (denoted as shunt), and good V/Q matching in gray (denoted as V/Q matching). The first patient with early ARDS A showed decreased percent of dorsal shunt and ventral dead space units, leading to decreased mismatch. The second patient with persistent ARDS B showed decreased percent of total shunt, leading to increased mismatch