Multipoint Search Algorithm for Automatic Segmentation of Tooth from Digital Intra Oral Periapical Radiographs

Objectives: In the present study digitally recorded Intra Oral Periapical radiographs (IOPA) are investigated. Methods/Analysis: This paper presents an innovative algorithm for automatic segmentation of the digital IOPA radiograph in one or more qualified regions of interest. Segmentation of IOPA is done repetitively till all teeth in the image are extracted. These segmented images resulting from this algorithm will now have one tooth in one image. Findings: A radiograph with exactly one tooth in image is extremely useful for further study and analysis of individual tooth in many and diverse applications of computer assistance in processing of dental radiographs. The segmentation algorithm proposed in this paper, necessarily being a spatial segmentation method, separates regions of interest without any loss or alterations in the data. The algorithm works successfully on 76 digital IOPA radiographs out of 80 images in test data. Novelty/Improvements: The algorithm proposed in the paper is fully automatic and does not require any human intervention. The segmentation algorithm proposed separates one tooth with its perimeter region intact in one image. The number of output images generated are equal to number of teeth in input IOPA.