VR-Based Support of Unipolar Depression - Institute of Experimental Psychophysiology

Virtual Reality-Based Support of Unipolar Depression

Support of Acute Therapy and Relapse Prevention in the Deep Psychological Treatment

Project Justification

Within the framework of cognitive behavioral therapy of anxiety disorders virtual realities (VRs) are already successfully used for a variety of anxiety disorders, such as specific phobias or panic disorders. In theory, VR also allows therapeutically supervised role-playing, whereby potentially negative effects on the therapist-client relationship are avoided by transferring the therapeutic interaction to a VR. However, the VR-based use of deep psychological and psychoanalytic therapies has been scarcely researched so far.

Application Scenario: Therapy

Our Approach

The EU-funded research project DeepVR investigates the potential of VR for the deep psychological treatment of unipolar depression. Supervised by therapists, patients are confronted with their central relationship conflict theme within a social role play situated in a VR environment. The goal is that patients explore new behaviors and evaluate their self-perception by repeatedly participating in VR role plays. During these social role plays, depression markers are extracted from speech data, to create a built-in speech corpus, as well as from eye-tracking data. A feasibility study will further evaluate the system’s usability, user experience, and user acceptance with a sample of patients with unipolar depression.

Scenario 1 of the VR environment: Therapy room and enviroment of the mindfulness exercise.

Scenario 2 of the VR environment: Interaction incl. course of conversation between therapist and patient.

Work Packages, Insights and Outcomes

Within the project, ixp is responsible for the requirement analysis to acquire the needs of potential user groups and stakeholders such as patients and therapists. Furthermore, ixp handles the conceptualization and design of suitable depression symptom indicators as well as therapy success markers derived from VR-immanent patient behavior. Another aspect here is the conceptualization of deep learning prediction models for relapse prevention. Machine learning and deep learning techniques are used to compute models based on mood, interest, and activity parameters of the patient’s behavior. These models can be used as indicators of recurrent episodes of depression and therefore as basis for an early warning system. To ensure the project’s overall success, ixp further administers the evaluation of the system’s effectiveness and user acceptance within the project.