ANALYSIS OF LUNG DENSITY BY COMPUTED-TOMOGRAPHY BEFORE AND DURING GENERAL-ANESTHESIA

Pulmonary structure was analysed by means of computed tomography (CT) in 20 lung-healthy patients, relating tissue density to the attenuation value (AV) of a picture element. Regional density of pulmonary tissue (r(lung)) was determined using mean lung density in five regions of interest (ROI1-5) (sector method). Vertical and horizontal distributions of x-ray attenuation were analysed by density profiles, relating AV values to evenly distributed and normalised length scales. In group I (n = 12), CT-densitometry was obtained in awake, supine patients and after induction of general anaesthesia. In group II (n = 8), the effect of mechanical ventilation with positive end-expiratory pressure (PEEP, 1.0 kPa [10 cmH2O]) was studied. In the awake state, a vertical tissue density difference between the top and the bottom of the lung was found in all patients, accounting for a mean of 0.235 g . cm-3 (right lung) and 0.199 g . cm-3 (left lung). Only minor changes were seen in the horizontal lung density profiles. After induction of anaesthesia, x-ray attenuation of ROI1-4 showed no significant differences when compared with the awake state. The basal lung areas (ROI5) revealed a significantly increased tissue density (P less-than-or-equal-to 0.0 1), reaching mean values of 0.94 g . cm-3 (right lung) and 0.814 g . cm-3 (left lung). Similarly, vertical density profiles showed a markedly enhanced r(lung) of the bottom of the lung in all patients, interpreted as atelectasis. The amount of atelectasis accounted for 4.8 +/- 2.6% (right lung) and 4.7 +/- 2.1% (left lung) of the intrapulmonary area. There was no evidence of ''non-gravitational'' inhomogeneity of density distribution seen in the horizontal density profiles. After application of PEEP, basal lung densities decreased significantly, although small basal densities remained in most patients (2.2 7 +/- 2.57% of right intrapulmonary area [P less-than-or-equal-to 0.01], 2.2 +/- 2.37% left intrapulmonary area [P less-than-or-equal-to 0.01]). Calculated alveolar recruitment was 7.7 cm2 and 8.4 cm2, whereas expansion of both lungs was smaller (4.3 cm2 and 4.4 cm2 [right and left lung]). Mean density of aerated tissue had decreased by 25%, and both horizontal and vertical attenuation profiles revealed an even distribution of r(lung). Analysis of r(lung) provides useful information about regional pulmonary morphology during anaesthesia and may be related to lung function.