Automatic retiming method based on genetic algorithm for the detection and the follow-up of dental lesions

In dental surgery, a great number of pathologies (cysts, granuloma…) presents clinical, radiological and evolutionary aspects considerably polymorphic. Medical imagery and particularly oral imagery, by the means of the digitized images and the image analysis algorithms, constitutes an essential element which leads to a precise presumptive diagnosis. The technique of retiming per subtraction can be thus used to recognize in two-dimensional images evolution of the pathological zones (lesions, tumours). From this point of view, a method of automatic retiming was developed by genetic algorithm to follow the evolution of the pathological zones after parodental treatment. In this article the traditional technique of retiming by specifying its disadvantages for our application and then an automatic model of retiming, which offers promising results, are presented. This tool will make it possible to distinguish, identify, and visualize automatically the form of the pathological structures and their evolution in time. An analysis of form and texture of these structures will allow the identification type of pathology.