Clinical diagnosis is an unavoidable part of the treatment process for a healthy society. Computer Diagnosis has brought the scenario to great extents from where the whole clinical practice used to be. Now for it to evolve further we need advancements at the very premium standards like the improvements in accuracy and precision of a diagnosis results. Inflow treatment procedures require computationally sound results in the best possible time complexity. In this paper, we propose a rough set based network method that can help a neural network to reduce the decision computational time, as the rough set attribute reductions will help the network to gain better efficiency by removing the unwanted attribute relations from the dataset. To demonstrate this method we use a modified version of the Wisconsin Breast Cancer Dataset obtained from the UCI Repositories. © 2015, International Journal of Pharmacy and Technology. All rights reserved.