Fuzzy based Regional Thresholding for Cyst Segmentation in Dental Radiographs

Dental cysts are the major problem caused due to infection in the tooth. Cysts cause any symptoms rarely, if not treated they become infected secondarily. The image processing helps to improve the diagnostic outcomes and provide effective early diagnostic yield, To achieve the highest possible degrees of automatization for the computer-aided diagnosis or detection of diseases, the proposed methodology utilizes both fuzzy membership function of each pixel and local spatial information of the neighbourhood pixels. There are five steps in implementing the fuzzy-based multi-region thresholding. From the comparative analysis of the results obtained by different segmentation methods like histogram-based multilevel thresholding methods. The proposed hybrid multi-region segmentation performs well and the outputs show that it is automated, accurate and it shows the well-connected boundary of the cystic region and there are no isolated cyst pixels.