The technology of using finger vein patterns is a biometric system that has a high level of security. By using the identification of blood vessels found on the human finger, when this data is needed it can be accessed immediately, and there is no possibility of being lost or forgotten. The condition of the blood vessels in the human body, precisely in the fat tissue, also makes it difficult for data to be stolen. A person's health data is quite crucial data. So we need a good security system. By using this blood vessel authentication, the data can only be accessed by the person concerned. With a finger vein pattern that is not easy to duplicate so it is suitable for creating a security system. In this research, we take advantage of the advantages of the finished vein pattern to create a patient data security system in the hospital. The system is made by utilizing one of the algorithms of an artificial neural network. This algorithm is oriented towards changing the value of weights and biases in the training process. From the training process, a model of an artificial neural network system is generated. The best labeling is for one finger vein identity, one label. The average time required for image recognition is 2.4 seconds. The best rejection result is 100% and the best acceptance is 81.67%.