Which Type of Urban Greenway is More Beneficial to the Health of Hospitalized Schizophrenic Patients? A Comparative Study of Virtual Greenway Walking

Background: Mental health issues in China are increasingly severe, with a significant rise in the number of inpatients with schizophrenia. Green spaces are strongly associated with health benefits, and virtual reality (VR) technology has emerged as an effective tool to connect patients with natural environments. However, research examining the role of urban greenways in the rehabilitation of schizophrenia patients remains limited. Objectives: This study aims to identify the virtual urban greenway environmental factors that influence the recovery of hospitalized schizophrenia patients and to determine their relative contributions. Methods: Field investigations and photography were conducted along the Chengdu Tianfu Greenway. Using deep learning-based image semantic segmentation technology, greenway environmental factors were quantified. A regression model, combined with scores from the Perceived Restorativeness Scale, was employed to identify key restorative factors. Subsequently, VR models of the greenway incorporating varying proportions of these factors were developed to determine the combinations that produce optimal restorative effects. Conclusions: The abundance of rest facilities (ARF), green view index (GVI), color richness (CR), and abundance of artificial facilities (AAF) were identified as the key factors influencing the restorative effects of urban greenways. Notably, a virtual greenway featuring a high green view index (over 50%) combined with a high abundance of rest facilities (over 15%) yielded the most significant restorative benefits for patients. Trial registration:The research protocol was reviewed and approved by the Academic Ethics Committee of Sichuan Agricultural University. It is consistent with the principles of the Declaration of Helsinki and relevant ethical guidelines, permitting the involvement of human participants. Ethical Approval Date: November 2024 .