Recently, cloud computing gained an important role in healthcare services due to its ability to improve the healthcare performance. However, the optimal selection of virtual machines to process a medical request represents a big challenge. For that, it proposes a new model for healthcare based on cloud environment using Parallel Particle Swarm Optimization to optimize the virtual machines selection. In addition, a new model for chronic kidney disease diagnosis and prediction is proposed to measure the performance of our virtual machines model. The prediction model of chronic kidney disease is implemented using two consecutive techniques, which are linear regression and neural network.