Image processing is a fundamental technique in the quest to identify physical pathologies like tumors and others. Lung cancer is today one of the main causes of death in the world among both men and women, with an impressive rate of about five million deadly cases per year. The survival rate is strictly related to the stage in which the disease is diagnosed. Precise medical imaging and analysis enable early detection of the lung cancer, help determine its exact size and location and significantly improve the diagnosis and treatment.
One client requested our help to automatically segment images of the pulmonary system. This lung segmentation should include blood vessels, the lower respiratory tract and pulmonary vasculature.
By using images of the human body for clinical purposes in order to enable precise diagnosis and analysis of diseases, RSIP Vision’s medical imaging expertise has found the optimal answer to the challenge of producing a viable airways segmentation. We properly recognized that correct segmentation of airway vessels offers the most effective solution to our client’s search. Our scientists wrote software built on advanced segmentation techniques to exploit CT images: this software offers an efficient segmentation of the airways and pulmonary vasculature.
The recommended airways segmentation method uses the strategy of adaptive region growing. This is an iterative technique employed to identify connected regions of interest, which in this case are contiguous sets of voxels. These regions should obey carefully selected inclusion rules based on threshold values, and should be in accord to the notion of discrete connectivity. This strategy starts by choosing an initial site in the region called a seed point, which in this case is based within the trachea: this site is grown and, after examination of its neighbors, it includes in the growing region only those voxels which meet the predetermined inclusion rules. Each included voxel becomes then a new seed point for the following iteration. This process continues until no voxel is added to the growing area.
Our program differentiates airway vessel from non-airway vessels using sophisticated algorithms, computing airway probability and orientation similarity. Parameters which are compared and analyzed include brightness, contrast and tubeness measures threshold.
The solution we choose for the airways segmentation allows an accurate identification of airway vessels in a large portion of the tree. In addition it assures a very efficient elimination of false positives in the findings: the method extracts more than half of the total tree length for more than 2/3 of the cases of the cases, with less than 1% false positives for all given cases. This provides a very effective leakage avoidance, while at the same time offering worthwhile imaging to the physician for appropriate and timely analysis and treatment.
Better monitoring is direly needed today to reduce the dramatically high mortality rate from lung pathologies. The airways segmentation solution which RSIP Vision delivers is a major contribution to help the medical profession giving earlier and better answers to major lung diseases.