Virtual reality supported craving detection method in alcohol and substance addiction

Abstract

The invention relates to a computer-based craving detection method enabling the detection of craving occurring in alcohol and substance addiction by means of virtual reality systems.

Description

[0001] VIRTUAL REALITY SUPPORTED CRAVING DETECTION METHOD IN ALCOHOL AND SUBSTANCE ADDICTION

[0002] Technical Field

[0003] The invention relates to a computer-based craving detection method enabling the detection of craving occurring in alcohol and substance addiction by means of virtual reality systems.

[0004] Prior Art

[0005] One of the major challenges in the treatment of alcohol and substance addiction is to manage craving situations occurring in situations triggering the al cohol / sub stance use of addicts. In traditional methods, craving is detected with questionnaires based on patient declaration (for example, Visual Analog Scale - VAS) or clinical observations. However, these methods are limited due to being subjective and not providing real-time data.

[0006] The patent document numbered KR101830067B1 discloses a method for preventing immersion in commercially accessible games and other technological devices and a device providing content for this.

[0007] The patent document numbered CN117243608A discloses a biological signal test system based on substance intake cue rating.

[0008] When the studies existing in the prior art are examined, a need has been felt for the development of a computer-based craving detection method enabling the detection of craving occurring in alcohol and substance addiction by means of virtual reality systems.

[0009] Objectives of the Invention

[0010] The object of this invention is to develop a computer-based craving detection method enabling the detection of craving occurring in alcohol and substance addiction by means of virtual reality systems. Another object of this invention is to develop a computer-based craving detection method offering a research opportunity on realistic environments directed towards alcohol and substance addiction behaviors.

[0011] Detailed Description of the Invention

[0012] A computer-based craving detection method enabling the detection of craving occurring in alcohol and substance addiction by means of virtual reality systems, and comprises;

[0013] • obtaining biophysiological and neurophysiological data from users, o obtaining electrodermal activity (EDA), electrocardiography (ECG), electrooculography (EOG), electromyography (EMG) data as biophysiological data from the user, o obtaining quantitative electroencephalography (QEEG) and event- related potential (ERP) data as neurophysiological data from the user,

[0014] • subjecting the data to pre-processing with noise filtering and amplification methods,

[0015] • applying the feature extraction step by taking changes associated with craving into account (For example; severe craving if significant increase in GSR peaks, decrease in HRV and increase in eye movements are present; moderate craving if mild GSR response, partial decrease in HRV and slight change in ERP component are present; mild craving if minimal physiological changes and insensitive ERP response are present),

[0016] • applying support vector machines (SVM), random forest (RF), decision trees (DT), K-Nearest neighbor (KNN), and linear discriminant analysis (LDA) algorithms simultaneously to the feature vectors created using numerical values calculated from all data,

[0017] • classifying the craving according to maj ority voting by looking at the results produced by the five classification algorithms applied simultaneously, which will be one of the mild-moderate-severe classes. (Majority voting means complying with the decision of the majority when the decisions given by the classification algorithms used are different from each other. For example, when two algorithms give the mild decision and three algorithms give the moderate decision, the moderate result will be reached according to majority voting.)

[0018] In the method subject to the invention, firstly, virtual environment videos will be created by using stimuli that will not trigger craving, cues that will trigger craving at a high level, and stimuli that will provide relief after craving, in line with the data obtained as a result of interviews and survey applications conducted with alcohol and substance addicted patients. These are named as VRNeutrai, VR nggenng, and VRReiaxing, respectively.

[0019] The software used in the development of virtual reality (VR) environments plays an important role in a system containing a VR application. Based on the features and capabilities it provides, environment design has a decisive effect on the appearance and performance of the final product. The increasing use of virtual reality in various fields and the development of VR applications to perform different tasks in related fields have caused the emergence of numerous VR software development platforms. The selection of the software tool for VR development depends on the purpose and functions of the application being developed.

[0020] Today's game engines have the feature of integrating ready-made accessories to create efficient VR-based applications. Factors contributing to the popularity of game engines include broad community support, economic affordability, the user- friendliness of interfaces, and the lack of need for specific expertise to use them. Furthermore, the graphic features and quality provided by the game engine is another important factor. In this context, it was decided to use Unity in designing the virtual environment to be used in the method subject to the invention.

[0021] In the system where the method subject to the invention will be applied, a selection of VR glasses offering users a high-quality and rich virtual reality experience has been provided. Due to its high ecological validity, VR technology has emerged as a powerful tool for mental health research. Computer-generated VR environments offer realistic, dynamic, interactive, and complex real-life simulations and offer the opportunity for active participation. Virtual environment designs are quite successful in creating a high sense of presence in users by combining stereoscopic three-dimensional visual, auditory, olfactory, and tactile perceptions.[1]

[0022] Although VR simulations are widely used in research conducted on mental illnesses, the examination of VR environments in addiction processes is still at an early stage. VR studies conducted in recent years have focused on the roles of craving, psychophysiology, emotional states, cognition, and brain activity in addiction. VR is a new tool regarding how we will examine proximal multi-sensory cues, contextual environmental cues, and their interaction (complex cues) modulating addiction behaviors. VR allows for experimental designs under highly standardized, strictly controlled, predictable, and reproducible conditions. Furthermore, VR simulations are personalizable. Personalization, eye tracking, and neurobiological factors are the new future VR directions. The advancement of VR applications has produced ways to understand treatment mechanisms underlying addiction, which researchers have just started to explore. VR methods promise to achieve significant successes in the field of addiction. These methods are necessary to develop more effective, efficient, and preventive therapeutic strategies.

[0023] In the last few years, Virtual Reality Exposure Therapy (VRET) has been used effectively in psychological and psychiatric disorders such as phobia, anxiety disorders, and post-traumatic stress disorders.[2'6]Researchers have stated that the VR environment can trigger the desire to use in smokers and alcohol users and is effective in treatment.[7'9]On the other hand, in a few studies, virtual reality technology has been used in drug addicts. Studies have shown that VR cues can effectively arouse subjective desires and, as a result of this, physiological parameters can change.

[0010]

[0024] Drug addiction treatment should not only reduce physical addiction. More importantly, it should treat psychological addiction. The superiority of VR is its immersing the patient into an effective and immersive environment regarding triggering the desire by exposing the patient to a substance-triggering environment. However, studies on drug addiction and VR are still rare; most VRET studies focus on nicotine or alcohol addiction. In particular, studies concentrating on craving in addiction are quite limited.

[0025] Various data such as Electrocardiography (ECG) and Electrodermal activity (EDA) have been widely used in studies for the treatment or assessment of drug addiction. Heart Rate Variability (HRV) analysis using ECG signals has been accepted as an effective way of observing the response of the autonomic nervous system. In a review study conducted, sensors used for the purpose of assessing addiction and the studies conducted were examined.

[0010] In studies conducted by different teams, it has been stated that the drugs (substances) used have an effect on the P wave, T wave, and QRS complex in the ECG.

[1112] EDA is also used to monitor human psychological symptoms and is accepted as a reflection of physiological changes. Based on this, in a study conducted, they used machine learning methods to recognize four emotional states, namely calmness, sadness, fear, and happiness, using EDA data.

[0013] Granados et al. applied a deep convolutional neural network to classify a person's emotional state with a combined ECG and EDA dataset.

[0014] Mahmud et al., on the other hand, aimed to perform timely intervention by utilizing machine learning techniques to assess the severity of drug addiction of patients with the EDA data of opioid (a drug used for recreational purposes) users.

[0015] On the other hand, some studies utilized the eye tracking technique to evaluate visual attention. Lohan et al. applied sequential neural network and K-means clustering to distinguish the difference between children with autism spectrum disorder and typically developing children using eye tracking, pupil diameter, and gaze position data.

[0016] In another study, a consistent physiological pattern was detected using EDA, Accelerometer (ACC), and temperature sensors after opioid administration in the emergency department, and it was stated that there were significant differences between the sensor data of heavy and light opioid users.

[0017] In the study, it was stated that potential applications of biosensors in fields such as drug addiction treatment and pain management need to be investigated in more detail. Finally, a study was conducted among heroin-addicted patients who have not used substances for a long time, and it was stated that skin conductance, EMG, temperature, and blood pressure values showed significant changes after exposure in heroin-addicted patients experiencing long-term abstinence.

[0018]

[0026] Research and reviews conducted regarding addiction-based VR applications have been examined in detail. According to the results obtained from the studies, physiological parameters measured by means of various sensors present significant assessments in studies associated with addiction. However, the smallness of sample volumes, the heterogeneity of the demographic characteristics of the participants, their use of a limited number of sensors, and the obtained results not being generalizable are significant shortcomings of the existing studies. Consequently, it was decided to use ECG, EDA, EOG, and EMG data in the craving detection system to be designed in line with the objective.

[0027] The direct interpretation of raw data collected from sensors is difficult. If a machine learning-based classification is to be performed using sensor data, the sensor data needs to be digitized. In this section, the features (numerical attributes) to be obtained via each physiological measurement tool decided to be used in the system are explained. In the feature extraction step, the goal is to define features that can be calculated quickly, are useful, provide distinctive information, and enable accurate classification to be performed.

[0028] In the method subject to the invention, the data collection process will be performed twice as VRNeutrai and VRynggenng. In the VRNeutrai phase, a scenario consisting of neutral stimuli and not triggering craving, lasting approximately 15 min, will be shown to the patients, and during this time, EDA, ECG, EOG, and EMG data, along with QEEG and ERP measurements to be used for the purpose of evaluating brain activity, will be taken from the patients. The main objective in this phase is to ensure the evaluation of the changes that will occur when exposed to the virtual environment video containing stimulating cues (VRynggenng). In the VRynggenng data collection process, a virtual environment video containing cues that will trigger craving, lasting approximately 15 min, will be shown to the patients. During this time as well, EDA, ECG, EOG, and EMG data, along with QEEG and ERP measurements to be used for the purpose of evaluating brain activity, will be taken from the patients. After the VR nggenng application, a relaxing virtual environment video VRR pyxing) lasting approximately 15 min will be shown to the patients. All virtual environment videos will be shown to the patients in a 20 m2quiet room.

[0029] EDA (Electrodermal Activity): It is used for the determination of stress and emotional responses by measuring skin conductance changes.

[0030] ECG (Electrocardiography): It is used for Heart Rate Variability (HRV) analyses.

[0031] EOG (Electrooculography): It is used for the evaluation of attention and arousal level by monitoring eye movements.

[0032] EMG (Electromyography): It is used for analyzing motor responses by measuring muscle activity.

[0033] The obtained data are subjected to the data pre-processing steps specified below.

[0034] Noise filtering: Environmental noises are reduced using a 50 Hz notch filter.

[0035] DC level removal: Drifts in EOG and EMG signals are corrected.

[0036] Z-Score normalization: Statistical scaling of the data is ensured.

[0037] In addition to the data obtained from these sensors, QEEG and ERP values are calculated for the purpose of neurophysiological data collection.

[0038] QEEG (Quantitative Electroencephalography): It is used in performing the frequency analysis of brain waves.

[0039] ERP (Event-Related Potentials): It is used in the evaluation of brain responses directed towards substance cues by measuring cognitive responses.

[0040] Changes that may occur in brain activity in the craving situation triggered by cues in the virtual environment will be analyzed with QEEG and ERP measurements. With this aspect, the proposed system provides a significant contribution to the literature. As is known, unlike standard EEG devices, there is additional software in QEEG. This software can also be used for the purpose of obtaining ERP measurements. Therefore, quantitative parameters for the purpose of analyzing brain activities in a multifaceted and stimulated state are easily obtained with software support and will be able to be included in the feature vector to be applied to the input of machine learning techniques in a similar framework with sensor data.

[0041] In the method subject to the invention, since patients will be exposed to stimuli by means of the virtual environment, conditioned response to stimuli, attention to cues relating to the substance, and subjective desire assessment gain importance. For this purpose, studies based on Electroencephalography (EEG) and Event Related Potentials (ERP) have been taken into account. EEG is a non-invasive neurophysiology technique used to record brain activity. EEG measures the electrical activity on the brain surface by means of a series of electrodes. This activity consists of the sum of small electrical changes generated by neurons during neural transmission. EEG is used in the diagnosis of neurological disorders, in the diagnosis of epilepsy, in sleep research, and in cognitive neurophysiology studies by recording brain waves and potentials. EEG enables the real-time monitoring of brain activity and therefore provides a significant contribution to the understanding and evaluation of neurological functions. Continuous changes in the electrical potentials in the brain can be recorded by placing EEG electrodes on or near the head.

[0042] In the method subject to the invention, machine learning algorithms are used to classify craving levels using the data obtained from the sensors.

[0043] Raw data obtained from the sensors are analyzed by being converted into certain features.

[0044] EDA: It is used in the Peak Detection algorithm for SCR (Skin Conductance Response) detection.

[0045] ECG: It is used in the calculation of HRV parameters (SDNN, RMSSD, LF / HF ratio). EOG: It is used in eye blink duration and gaze orientation measurements.

[0046] EMG: It is used in the frequency and amplitude analysis of muscle activity.

[0047] QEEG: It is used in the evaluation of ratios between Delta, Theta, Alpha, Beta, and Gamma frequency bands.

[0048] ERP: It is used in the evaluation of the P300 component amplitude and latency.

[0049] In the method subject to the invention, craving levels are classified as mild, moderate, and severe. A decision is made in a manner compatible with clinical evaluations by feeding the features obtained from sensor data into the machine learning model. For example; if a significant increase in EDA peaks, a decrease in HRV, and an increase in eye movements are present, it can be decided that there is severe craving. The machine learning-based decision support mechanism subject to the patent will offer a much more detailed analysis using biophysiological and neurophysiological data.

[0050] References:

[0051] [1] Mazza, M., Kamml er- Sucker, K., Lemenager, T., Kiefer, F., & Lenz, B. Virtual reality: a powerful technology to provide novel insight into treatment mechanisms of addiction. Translational Psychiatry, 2021, 11(1), 617.

[0052] [2] B. K. Wiederhold et al., The treatment of fear of flying: A controlled study of imaginal and virtual reality graded exposure therapy,” IEEE Trans. Inf. Technol. Biomed., vol. 6, no. 3, pp. 218-223, Sep. 2002.

[0053] [3] Carlin, A.S., Hoffman, H. G. & Weghorst, S. “Virtual reality and tactile augmentation in the treatment of spider phobia: A case report,” Behav. Res. Ther., 1997, 35(2), pp. 153-158.

[0054] [4] Maples-Keller JL, Bunnell BE, Kim SJ, Rothbaum BO. The use of virtual reality technology in the treatment of anxiety and other psychiatric disorders. Harv. Rev. Psychiatry. 2017 May / Jun;25(3): 103-113. doi: 10.1097 / HRP.0000000000000138. [5] Rizzo, A., Hartholt, A., Grimani, M., Leeds, A., & Liewer, M. Virtual reality exposure therapy for combat-related posttraumatic stress disorder. Computer, 2014, 47(7), 31-37.

[0055] [6] Gongalves R, Pedrozo AL, Coutinho ES, Figueira I, Ventura P. Efficacy of virtual reality exposure therapy in the treatment of PTSD: a systematic review. PLoS One. 2012, 7(12):e48469. doi:

[0056] 10.1371 / journal. pone.0048469.

[0057] [7] Thompson-Lake D.G., Cooper K.N., Mahoney J. J. 3rd, Bordnick P.S., Salas

[0058] R., Kosten T.R., Dani J. A., De La Garza R. 2nd. Withdrawal Symptoms and Nicotine Dependence Severity Predict Virtual Reality Craving in Cigarette- Deprived Smokers. Nicotine Tob Res. 2015, 17(7):796-802. doi: 10.1093 / ntr / ntu245.

[0059] [8] Culbertson C. S., Shulenberger S., De La Garza R., Newton T.F., Brody A.L. Virtual reality cue exposure therapy for the treatment of tobacco dependence. J. Cyber Ther Rehabil. 2012, 5(1), 57-64. PMID: 25342999.

[0060] [9] Bordnick, P. S., Traylor, A., Copp, H. L., Graap, K. M., Carter, B., Ferrer,

[0061] M., & Walton, A. P. (2008). Assessing reactivity to virtual reality alcohol based cues. Addictive Behaviors, 2008, 33(6), 743-756. https: / / doi.Org / 10.1016 / j.addbeh.2007.12.010

[0062]

[0010] Oesterle, T. S., Karpyak, V. M., Coombes, B. J., Athreya, A. P., Breitinger,

[0063] S. A., Correa da Costa, S., & Gerberi, D. J. (2022). Systematic review: Wearable remote monitoring to detect nonalcohol / nonnicotine-related substance use disorder symptoms. The American Journal on Addictions, 31(6), 535-545.

[0064]

[0011] Carreiro, S., Wittbold, K., Indic, P., Fang, H., Zhang, J., & Boyer, E. W. (2016). Wearable biosensors to detect physiologic change during opioid use. Journal of Medical Toxicology, 12, 255-262.

[0065]

[0012] Natarajan, A., Parate, A., Gaiser, E., Angarita, G., Malison, R., Marlin, B., & Ganesan, D. (2013, September). Detecting cocaine use with wearable electrocardiogram sensors. In Proceedings of the 2013 ACM International Joint Conference on Pervasive and Ubiquitous Computing (pp. 123-132).

[0013] Zangroniz R, Martinez-Rodrigo A, Pastor JM, Lopez MT, Fernandez- Caballero A. Electrodermal activity sensor for classification of calm / distress condition. Sensors (Basel). 2017; 17: 10, 2324.

[0066]

[0014] Santamaria-Granados, L., Munoz-Organero, M., Ramirez-Gonzalez, G., Abdulhay, E., & Arunkumar, N. J. I. A. (2018). Using deep convolutional neural network for emotion detection on a physiological signals dataset (AMIGOS). IEEE Access, 7, 57-67.

[0067]

[0015] Mahmud, M. S., Fang, H., Wang, H., Carreiro, S., & Boyer, E. (2018, March). Automatic detection of opioid intake using wearable biosensor. In 2018 International Conference on Computing, Networking and Communications (ICNC) (pp. 784-788). IEEE.

[0068]

[0016] Lohan, K. S., Sheppard, E., Little, G., & Rajendran, G. (2018). Toward improved child-robot interaction by understanding eye movements. IEEE Transactions on Cognitive and Developmental Systems, 10(4), 983-992.

[0069]

[0017] Carreiro, S., Wittbold, K., Indic, P., Fang, H., Zhang, J., & Boyer, E. W. (2016). Wearable biosensors to detect physiologic change during opioid use. Journal of Medical Toxicology, 12, 255-262.

[0070]

[0018] Zhao, M., Fan, C., Du, J., Jiang, H., Chen, H., & Sun, H. (2012). Cue- induced craving and physiological reactions in recently and long-abstinent heroin-dependent patients. Addictive Behaviors, 37(4), 393-398.

Claims

CLAIMS1. A computer-based craving detection method enabling the detection of craving occurring in alcohol and substance addiction by means of virtual reality systems, characterized in that it comprises;- obtaining biophysiological and neurophysiological data from users, subjecting the data to pre-processing with noise filtering and amplification methods,- classifying the data according to maj ority voting by applying support vector machines (SVM), random forest (RF), decision trees (DT), K-Nearest neighbor (KNN), and linear discriminant analysis (LDA) algorithms to feature vectors created using values calculated from all data,- determining whether there is mild-moderate-severe craving in patients by interpreting the classification results.

2. The method according to claim 1, characterized in that electrodermal activity (EDA), electrocardiography (ECG), electrooculography (EOG), and electromyography (EMG) data are obtained from the user as biophysiological data.

3. The method according to claim 1, characterized in that quantitative electroencephalography (QEEG) and event-related potential (ERP) data are obtained from the user as neurophysiological data.