Model Checking an Artificial Neural Networks System in Medical Diagnosis

With the progress in science and technology, many Applications of Artificial Neural Networks in medical fields have been developing rapidly. Considering that the advent of practical service seems to be inevitable, the performance evaluation of these systems is worthy. Therefore, this paper selects one typical ANNs-based system in medical diagnosis to verify its reliability. In order to check the reliability of this system, this paper introduces a tool called PRISM with the probabilistic model checking technique. Additionally, the self learning mechanism inside the ANNs will also be simulated to discern its usefulness. Our study comprehensively digs the impact of those modules in this system and discovers the positive function of the self learning on medical diagnosis.