A Schizophrenia Screening Method Based on Fine-grained Facial Features and Network Isolation Privacy Protection

Schizophrenia is a common and complex severe mental disorder, and early screening for this condition is critical for social harmony and stability, as well as personal health and growth. Based on facial image features of 85 schizophrenia patients, this paper proposes an early schizophrenia screening framework, along with the implementation of private cloud network isolation technology to safeguard user privacy. We first design a network security strategy based on a wireless sensor network using network physical isolation technology to ensure the security of user personal privacy. We further exploit three base classifiers: K-Nearest Neighbors, Support Vector Machine, and eXtreme Gradient Boosting, to formulate the schizophrenia screening decision process using facial image features. Finally, we explore the sensitivity of facial image features in schizophrenic individuals to classifier selection and time consumption.