Personalized Urination Activity Recognition Based on a Recurrent Neural Network Using Smart Band

This urination behavior recognition technology showed high accuracy and could be applied in clinical settings to characterize patients' voiding patterns. As wearable devices are developed and generalized, algorithms detecting consistent sequential body movement patterns reflecting specific physiologic behavior might be a new methodology for studying human physiologic behavior.