Modeling of glaucoma induced changes in the retina and neural net assisted diagnosis

A system for modeling and early detection of glaucoma induced changes in the human retina is described. The system includes a modeling tool for design of semi-realistic retinal pictures, that may either be used for educational purposes or (as here) as a laboratory for controlled signal processing experiments. The detection system includes preprocessing algorithms for elimination of intensity variations and other artefacts. The final segmentation step is based on a cellular neural network.<
>