Virtual reality (VR) is widely used for medical purposes, particularly, for assessment and treatment of phobic anxiety disorders. The article provides a review and critical analysis of methods and tools for assessing phobic anxiety disorders via registration and digital processing of physiological signals, and adaptation of various VR scenes to the user's actual state. The article presents a technique for studying phobic anxiety disorders in VR environment that enables VR scenes to be adapted for a particular user, considering his individual features, and behavior parameters. Adaptation of VR scenes is carried out by selecting an optimal set of hidden signal patterns registered with electrocardiogram and electroencephalogram signals.