METHOD AND SYSTEM OF SEGMENTATION OF LUNG FOCI IMAGES Russian patent published in 2018 - IPC A61B6/03 G06T7/00 G06K9/50 

FIELD: data processing.

SUBSTANCE: group of inventions relates to the field of computed tomography (CT) data processing and can be used for segmentation of images of pulmonary foci. Chest CT data containing images are received. First stage of segmentation of CT data is performed, at which the chest region is determined. Identified bone structures and soft tissues of the mediastinum with the exception of the lung segment are removed from the images. CT-density parameters of the structures in the given center window and the width of the Hounsfield range of the image areas obtained after the first segmentation stage are determined. Average value of their CT-density parameter is determined. Areas of images with a CT-density parameter above the average one are determined. Second segmentation stage is performed at which a set of heterogeneous structures is determined on the images obtained after the first segmentation stage. Filtration of the resulting structures, consisting of two stages, is performed. First, the points on the image of the obtained structures are determined and the average curvature of the resulting structures is determined based on the points mentioned. Further, structures are determined whose average curvature index is close to or lies in the range of 0 to 0.05 of the curvature of the circumference, the area of which is equivalent to the area of the structure being analyzed. System for segmenting images of the lung foci comprises at least one processor and at least one memory. This memory contains machine-readable instructions, which, when executed by at least one processor, carry out this method of segmentation of images of the lung foci.

EFFECT: group of inventions provides an increase in the accuracy of detection of lung foci due to automatic segmentation of CT images.

3 cl, 5 dwg