Keywords: fuzzy neural networks, ANFIS, oculography, virtual reality, VR exposure therapy, emotion classification, fuzzy distress index, risk assessment, ordered fuzzy numbers
UDC 004.85:004.946
DOI: 10.26102/2310-6018/2026.62.11.008
The current stage of VR development is characterized by exponential growth in the implementation of VR technologies in healthcare. The ability of VR systems to understand and respond to user emotions opens up revolutionary possibilities in healthcare and psychotherapy. Russian and international research in this field demonstrates that VR technologies can provide personalized and adaptive therapy for anxiety disorders, depression, and Post-Traumatic Stress Disorder. These technologies can monitor a patient's condition in real time and dynamically adjust the difficulty of exercises within cognitive-behavioral therapy, exposure therapy, and neurorehabilitation. Of particular relevance is current research aimed at implementing adaptive monitoring technologies for the emotional states of virtual reality (VR) users based on artificial intelligence approaches. This approach will enable the transition from static to truly adaptive and immersive systems based on affective computing and advances in artificial intelligence. This paper analyzes a method for adaptive monitoring of the emotional states of virtual reality (VR) users based on fuzzy neural networks, ANFIS (Adaptive Neuro-Fuzzy Inference System). The proposed approach was tested using eye-tracking data available in the open-source VREED dataset. The study utilizes hybrid learning, combining the least-squares method for configuring inference parameters and gradient descent with an adaptive Adam step for configuring Gaussian membership function parameters. The neural network output generates a fuzzy distress index (FDI), which enables real-time decisions about adapting or terminating VR sessions. The proposed ANFIS architecture achieves Accuracy = 0.77 and F1 = 0.77 on a four-class classification task using a two-dimensional affect model (valence × activation), significantly outperforming the results of using FNN on ordered fuzzy numbers (OFN), KNN, and decision trees. The statistical significance of the differences is confirmed by the McNemar test.
1. Parsons Th.D., Rizzo A.A. Affective outcomes of virtual reality exposure therapy for anxiety and specific phobias: A meta-analysis. Journal of Behavior Therapy and Experimental Psychiatry. 2008;39(3):250–261. https://doi.org/10.1016/j.jbtep.2007.07.007
2. Opriş D., Pintea S., Garcia-Palacios A., Botella C., Szamosközi Ş., David D. Virtual reality exposure therapy in anxiety disorders: a quantitative meta-analysis. Depression and Anxiety. 2012;29(2):85–93. https://doi.org/10.1002/da.20910
3. Mevlevioğlu D., Tabirca S., Murphy D. Real-Time Classification of Anxiety in Virtual Reality Therapy Using Biosensors and a Convolutional Neural Network. Biosensors. 2024;14(3):131. https://doi.org/10.3390/bios14030131
4. Bekler M., Yilmaz M., Ilgın H.E. Assessing Feature Importance in Eye-Tracking Data within Virtual Reality Using Explainable Artificial Intelligence Techniques. Applied Sciences. 2024;14(14):6042. https://doi.org/10.3390/app14146042
5. Alharbi H. Explainable feature selection and deep learning based emotion recognition in virtual reality using eye tracker and physiological data. Frontiers in Medicine. 2024;11:1438720. https://doi.org/10.3389/fmed.2024.1438720
6. Tabbaa L., Searle R., Bafti S.M., et al. VREED: Virtual Reality Emotion Recognition Dataset Using Eye Tracking and Physiological Measures. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies. 2021;5(4):178. https://doi.org/10.1145/3495002
7. Apiecionek L. Fuzzy Neural Networks – A Review with Case Study. Applied Sciences. 2025;15(13):6980. https://doi.org/10.3390/app15136980
8. Jang J.-S.R. ANFIS: Adaptive-Network-based Fuzzy Inference System. IEEE Transactions on Systems, Man, and Cybernetics. 1993;23(3):665–685. https://doi.org/10.1109/21.256541
9. Rahman M.A., Brown D.J., Mahmud M., et al. Enhancing biofeedback-driven self-guided virtual reality exposure therapy through arousal detection from multimodal data using machine learning. Brain Informatics. 2023;10(1):14. https://doi.org/10.1186/s40708-023-00193-9
10. Apiecionek Ł., Moś R., Ewald D. Fuzzy Neural Network with Ordered Fuzzy Numbers for Life Quality Technologies. Applied Sciences. 2023;13(6):3487. https://doi.org/10.3390/app13063487
11. Małolepsza O., Mikołajewski D., Prokopowicz P. Adaptation of fuzzy systems based on ordered fuzzy numbers: a review of applications and development prospects. Electronics. 2025;14(12):2341. https://doi.org/10.3390/electronics14122341
12. Tan J.Ch.M., Cao Q., Quek Ch. FE-RNN: A fuzzy embedded recurrent neural network for improving interpretability of underlying neural network. Information Sciences. 2024;663:20276. https://doi.org/10.1016/j.ins.2024.120276
Keywords: fuzzy neural networks, ANFIS, oculography, virtual reality, VR exposure therapy, emotion classification, fuzzy distress index, risk assessment, ordered fuzzy numbers
For citation: Pozdnyakov D.A., Azarnova T.V. Fuzzy neural networks for emotional state classification and risk assessment in VR exposure therapy. Modeling, Optimization and Information Technology. 2026;14(11). URL: https://moitvivt.ru/ru/journal/article?id=2612 DOI: 10.26102/2310-6018/2026.62.11.008 (In Russ).
© Pozdnyakov D.A., Azarnova T.V. Статья опубликована на условиях лицензии Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NS 4.0)Received 14.09.2026
Revised 03.11.2026
Accepted 12.11.2026