Pulmonary nodules volumetry. Variability in Results After Manual Correction of Contours

Abstract

Objective. To study variability of volumetric of the pulmonary nodules volumetry after manual correction of their contours.Material and methods. Twenty-seven nodules uncircumscribed from the vascular structures and pleura were selected. A linear regression model was used to investigate the impact of the size of a nodule, the area of its contact with the adjacent structures on variability in results.Results. The linear regression model based on contact area and nodule size can correctly predict volumetry variability.Conclusion. Even after manual segmentation volumetry remain suitable method for size assessment of lung nodules. Segmentation should be made with the same person because of significant difference of interobserver and intraobserver variabilities.