Computer based diagnosis using artificial neural networks

We describe a package using parallel and distributed processing elements. This package called MICINE is based on the ARTi-Artificial Neural Network proposed by Carpenter and Grossberg of MIT, USA. It is quite tedious to go through and ascertain the exact disease by giving simply some symptoms. Here we have tried to design a package which will be more efficient than a well experienced Doctor. This system takes in a set of the symptoms from the patient and tries to get the best matching disease for the symptom-set. Then it gives the proper prescription for the medicines. We have implemented the package using an alternative computational paradigm which is inherently parallel and distributed. This system when converted to a fully automatic system can give treatments to any patients without the help of human beings. Thus it can reduce human errors