| Summary: | Purpose. To develop the automated expert support system for optic nerve head morphological description in normal conditions and in pathology.Methods. The proposed expert support system is based on the integration algorithm of luminance samples along the diagonal, it allows to detect the optic nerve head border. On the basis of this algorithm the method for solving the following tasks of fundus image processing have been proposed: detecting of the optic nerve head border, method of the morphological description of the optic nerve head boundary, method of the determining the value of the disk excavation. An experimental study of the parameters effect on the effectiveness of the optic nerve head detecting method was made.Results. The effectiveness assessment of the proposed border detection algorithm on the optic nerve head model has showed that the amount of overlap averaged 0.985, which indicates high quality. It was found that the algorithm for estimating the diameter of the single-sided optic nerve head image is sufficiently resistant to changes in such parameters as the influence of the noise level in the scene and the offset of the strobe center coordinates of the samples accumulation from the image center coordinates. Evaluation of the efficiency of the optic nerve head borders morphological description has showed that the value of the first-order derivative of the result of accumulation of luminance readings diagonally for images of optic nerve head with blurred boundaries is 2 times smaller than for images of optic nerve head with clear boundaries. The effectiveness of the method of selecting the border for assessment the disk excavation size was examined. It was obtained that the error in estimating the magnitude of excavation amounted to an average of 8.43 %.Conclusions. Тhe presented expert support system allows to automate the process of optic disk morphological description, in particular, such parameters as the state of the border and the size of the disc excavation. This method can be used to create medical expert systems and software for fundus images processing.
|
|---|
