An automated and hybrid method for cyst segmentation in dental X-ray images

The dental X-ray image has poor contrast and uneven exposure which results in the lack of reliable separation between various parts of teeth, which makes the segmentation of cyst very tedious. A unique hybrid automated technique has been proposed to detect and extract the cystic region using circularly symmetric isophote properties and fast marching method.The isophote curvature is the curve connecting the same intensity pixels. Each Isophote curvature line has an isocenter associated with it. Among them, the isocenter which is having a maximum response in the isophote center map to be concluded as the most likely estimate for locating the cystic region. This Maximum IsoCenter (MIC) is the seed point to the model-based segmentation of fast marching method. The fast marching algorithm (FMM) is like Dijkstra’s algorithm, and it follows the shortest path from seed area, where the information flows outward only. It works systematically to make it fast, and it is a one-pass method because each point is touched only once mainly. This fast marching method extracts the cystic region boundary very effectively and efficiently. This two-stage hybrid method is an automated, robust, and fast method for solving the complex problem of cyst segmentation. The average execution time calculated is 2.8 s and the accuracy achieved is 95%. The performance outcomes show that the proposed segmentation technique has the high correlation with the manual method. Therefore, the combination of model-based and feature based segmentation of the dental X-ray images has great potential in diagnosis of dental diseases and plays a significant role in the development of automated systems. The automated segmentation computerizes or automates the diagnostic method so that huge number of patients can be monitored with the same assured accuracy. High-speed computers are helpful in attaining fast and precise results. To extend the patient care to remote areas, the faster communication is possible by computer networks