RESUMO
OBJECTIVES@#To estimate lung volume in normal population by a combination of multi-slice spiral CT and Pulmo software.@*METHODS@#Scans of the lung were performed on 45 normal individuals using Emotion 16-slice spiral CT and Pulmo software. A measurement instrument of work station was used to measure the vertical diameter of left and right lungs, the maximum anteroposterior and transverse diameters at the layers of apex pulmonis 1 cm below, arcus aortae, subcarinal and diaphragmatic dome, and the lung volume. Regression analysis was performed on normal lung volume and each measured parameter by SPSS 20.0 and an optimal model was selected.@*RESULTS@#The goodness of fit between the best curvilinear equations of the normal bilateral lung volumes was 0.981. The goodness of fit between the curvilinear equations of the normal bilateral lung volume and the product of the vertical diameter and the maximum transverse diameter of diaphragmatic dome were 0.977 and 0.972, respectively. Fifteen cases were selected to make a retrospective test on the 3 models. No significant difference was found between the estimated and measured with Pulmo software lung volumes.@*CONCLUSIONS@#The normal lung volume before injury can be estimated by the curvilinear equations established in this study. The percentage of lung compression can be measured based on the volume of compressed lung, which can provide a reference for the identification of lung compression degree.