ABSTRACT
Early diagnosis is especially important, because cataracts develop slowly and many patients are not aware of their gradual loss of sight until very late in the disease when vision is irreversibly affected. Therefore, improving methods for cataract screening is an important field of research. This research involves the combination of engineering and medical methods in order to create physical solutions which eliminate suffering caused by cataract-related blindness. These methods must address cost effectiveness and reliability issues [4]. One way to address these issues is to use inexpensive computer technology to design and develop automated cataract classification systems for diagnosis support. Computer-based systems come with their own set of challenges, which result from the immense flexibility of modern processing machines [5]. One of these shortcomings is state space explosion and nondeterminism of large-scale systems. To overcome, or at least manage, these challenges, this type of medical support system must be designed with a formal and model-driven design methodology [6,7].
