A personalized biofeedback analgesia system in a virtual reality environment

Abstract

This invention discloses a personalized biofeedback analgesia system in a virtual reality environment, relating to the technical field of pain regulation. It includes a personalized feedback scenario generation module, which collects the test subject's pain experience data to form a desired scenario video resource library; a personalized pain biomarker mining module, which collects multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states to identify individual-specific pain biomarkers; an online pain monitoring module, which acquires the patient's current multimodal bioelectrical signals in real time, extracts pain biomarker response values, compares them with preset pain thresholds, and outputs a pain warning signal; and a pain intervention module, which, upon receiving the pain warning signal, matches and recommends personalized scenario videos generated based on the patient's characteristics from the desired scenario video resource library, and uses virtual reality to visually present the changes of corresponding regulatory elements in the recommended scenario videos in real time to alleviate pain. This invention has a more comprehensive data foundation, can meet the individual needs of patients, and reminds patients to take pain intervention measures when the pain signal is weak, achieving effective prevention and relief of pain.

Description

Technical Field

[0001] This invention relates to the technical field of pain regulation, and more specifically, to a personalized biofeedback analgesia system in a virtual reality environment. Background Technology

[0002] Pain is an unpleasant subjective feeling and emotional experience associated with actual or potential tissue damage, and it is one of the most common clinical symptoms. Statistics show that the number of patients with chronic pain is increasing rapidly and at a younger age, placing a heavy economic and psychological burden on patients and their families. Chronic pain is long-lasting, persistent, and difficult to cure, often accompanied by various complications, easily generating negative emotions, and severely reducing quality of life. Current pain management mainly relies on drug therapy, but this has problems such as drug dependence and significant side effects. Developing effective non-pharmacological analgesics will help alleviate patient suffering, improve their quality of life, and reduce the social medical burden.

[0003] Biofeedback is a therapeutic approach that assists an individual's self-regulation by monitoring their physiological processes and providing corresponding biological information. By learning and controlling external feedback signals provided by instruments, individuals can learn to self-regulate internal psychophysiological changes to prevent and treat specific diseases or optimize their physiological and psychological state. Based on the type of brain activity signals, biofeedback can be divided into peripheral neurophysiological feedback and central nervous system-based feedback. Currently, commonly used biofeedback techniques in clinical practice mainly target peripheral signals, such as electromyography (EMG), electrodermal activity (TEA), electrocardiography (ECG), and respiration. Central nervous system-based biofeedback, also known as neurofeedback, is a biofeedback technique within cognitive behavioral therapy. Neurofeedback technology is a method of neural oscillation modulation. Using neurofeedback training, pain nerve signal characteristics can be fed back to pain patients in visual or auditory form, enabling them to self-regulate their brain state based on the feedback, potentially achieving pain relief. A major problem with current biofeedback systems is significant individual variability; for example, approximately 60% of the population cannot achieve autonomous regulation of neural signals based on feedback signals, which may stem from the following two reasons.

[0004] First, most neurofeedback techniques present individual neurophysiological signals as simple two-dimensional signals (Hesam-Shariati et al., 2022), such as feeding back the amplitude of alpha nerve oscillations in the sensorimotor cortex or theta nerve oscillations in the prefrontal cortex to patients with chronic pain. Neurofeedback studies using multivariate pattern analysis (MVPA) have shown that significant effects can be achieved after only a few training sessions in improving attention and modulating perception (deBettencourt et al., 2015; Shibata et al., 2011). These studies suggest that using machine learning methods to extract personalized, pain-specific, and sensitive neurobiomarkers, and then feeding them back to pain patients, could potentially maximize the analgesic effects of neurofeedback training.

[0005] Secondly, most neurofeedback designs use visual feedback, displaying bars of height that increase or decrease based on brain activation levels. However, such simple and unengaging designs can make feedback sessions negative and tiring, necessitating the definition of more engaging feedback designs. The high immersion of technology makes it an ideal tool for addressing this issue. With the help of VR technology, we can present the neurophysiological signals of subjects in a virtual environment, providing a more engaging feedback experience and thus improving the effectiveness of neurofeedback training.

[0006] To address the issue of significant individual variability in neurofeedback training, this invention proposes the construction of a personalized biofeedback system. We will employ machine learning methods applied to multi-dimensional electrophysiological signals to uncover personalized pain biomarkers. Then, we will feed back the pain signals to the subjects in a VR environment, constructing a virtual reality biofeedback training system for pain relief. Summary of the Invention

[0007] To overcome the shortcomings of existing technologies, such as the inability to accurately detect human pain states, delayed pain intervention in mild pain states, and poor analgesic effects, this invention provides a personalized biofeedback analgesia system in a virtual reality environment.

[0008] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0009] This invention provides a personalized biofeedback analgesia system in a virtual reality environment, comprising:

[0010] The personalized feedback scenario generation module is used to collect the tester's pain experience data, generate desired scenario videos based on the pain experience data, and set corresponding adjustment elements to form a desired scenario video resource library; in subsequent practical applications, the original desired video resource library is continuously updated and supplemented based on the patient's pain experience data;

[0011] A personalized pain biomarker mining module is used to collect multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states, and to determine pain biomarkers from the multimodal bioelectrical signals based on machine learning methods.

[0012] The online pain monitoring module is used to acquire the patient's current multimodal bioelectrical signals in real time, extract pain biomarker response values ​​from the current multimodal bioelectrical signals, compare them with a preset pain threshold, and output a pain warning signal based on the comparison result; wherein, the pain threshold is set based on the pain biomarkers;

[0013] The pain intervention module is used to, upon receiving the pain warning signal, match the desired situation video resource library based on the pain questionnaire to obtain recommended situation videos; select a preset pain relief mode to adjust the adjustment elements corresponding to the recommended situation videos; and use virtual reality to visually present the changes of the adjustment elements in the recommended situation videos in real time to relieve pain.

[0014] This invention first establishes a personalized feedback scenario generation module to collect the test subject's pain experience data, generate desired scenario videos, and establish a desired scenario video resource library. Then, a personalized pain biomarker mining module simultaneously collects the patient's bioelectrical signals and corresponding pain questionnaires. Compared to traditional methods relying on patient verbal pain reports or collecting only single physiological signals, this provides a more comprehensive data foundation. Through machine learning, the bioelectrical signals are classified to extract the patient's personalized pain biomarkers, meeting the needs of individual patient differences. Next, an online pain monitoring module is set up to acquire the patient's current bioelectrical signals in real time, extract pain biomarker response values, compare them with preset pain thresholds, and output pain warning signals based on the comparison results, meeting the need for timely reminders and intervention, providing alerts before pain occurs. Finally, a pain intervention module is set up. Upon receiving the pain warning signal, it matches the desired scenario video resource library based on the pain questionnaire to obtain recommended scenario videos that are easier to relieve pain. A preset pain relief mode is selected to adjust the adjustment elements corresponding to the recommended scenario video, and virtual reality is used to visually present the changes of the adjustment elements in the recommended scenario video in real time to alleviate pain. Finally, the highly immersive nature of virtual reality is used to present the information to the patient in real time, enhancing the patient's involvement and thus achieving a better analgesic effect.

[0015] Preferably, the personalized pain biomarker mining module includes:

[0016] The multi-state information acquisition submodule is used to collect multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states. The pain questionnaire includes pain intensity scores and pain keywords.

[0017] The preprocessing submodule is used to perform preprocessing operations on multimodal bioelectric signals under different states to obtain preprocessed bioelectric signals;

[0018] The pain biomarker extraction submodule is used to input the preprocessed bioelectrical signals and the corresponding pain intensity scores into the constructed classifier to obtain the patient's pain biomarkers.

[0019] Preferably, the multi-state information acquisition submodule acquires multimodal bioelectrical signals and corresponding pain questionnaires from the patient in different states. The pain questionnaire includes pain intensity scores and pain keywords, including:

[0020] Bioelectrical signals of the patient were collected multiple times during the pain attack and non-pain attack periods, with the same preset duration.

[0021] During each bioelectric signal acquisition process, the patient selects the corresponding pain intensity score and pain keywords;

[0022] Each bioelectric signal and its corresponding pain intensity score are treated as a bioelectric data set with a pain intensity label, and all bioelectric data sets with pain intensity labels are combined to form a bioelectric dataset.

[0023] Preferably, the pain intensity score ranges from 0 to 10 points, where 0 points indicates no pain, 1 to 3 points indicates mild pain, 4 to 6 points indicates moderate pain, 7 to 9 points indicates severe pain, and 10 points indicates unbearable pain.

[0024] Preferably, the preprocessing submodule includes preprocessing operations such as filtering, bad segment removal, and independent principal component analysis.

[0025] Preferably, in the pain biomarker extraction submodule, the preprocessed bioelectrical signals and corresponding pain intensity scores are input into the constructed classifier to obtain the patient's pain biomarkers, including:

[0026] Bioelectric data labeled with pain intensity are used to form a bioelectric dataset, which is then input into a constructed classifier. This establishes a mapping relationship between different pain intensities and bioelectric signals, learns the characteristics of bioelectric signals under different pain intensities, and obtains pain biomarkers, including one or more of the following: rhythmic features, network features, cross-band coupling features, and microstate features.

[0027] This study collects bioelectrical signals from patients during and between pain attacks, as well as their subjective pain intensity scores, to more precisely measure the dynamic fluctuations of their pain state. Different patients exhibit different bioelectrical signals distinguishing between pain attacks and periods of non-attack. Machine learning is used to identify neural features that differentiate or characterize these two states, such as bioelectrical waves of a certain frequency or brain activity patterns, including frequency, amplitude, temporal characteristics, and the strength of functional connectivity between different brain regions. These serve as personalized pain markers for each patient. When the pain state changes, the patient provides an updated pain intensity score via keyboard, ranging from 0 to 10. Each bioelectrical signal and its corresponding pain intensity score constitutes a bioelectrical data set labeled with pain intensity. The study further categorizes the bioelectrical signals caused by each patient's pain physiological state to uncover personalized pain biomarkers: the collected bioelectrical signals are preprocessed, and then the bioelectrical data labeled with pain intensity is used as input to train a classifier. During training, the classifier learns the characteristics of bioelectrical signals under different pain states to effectively distinguish them. In this way, the classifier can establish a mapping relationship between different pain states and bioelectrical signals, thereby finding personalized pain biomarkers for each patient. Extracting these markers can help us understand each patient's pain response more accurately, providing strong support for personalized pain assessment and management.

[0028] Preferably, the online pain monitoring module includes:

[0029] The threshold setting submodule is used to set the maximum response value of pain biomarkers during the pain-free period as the pain threshold and send it to the detection and evaluation submodule.

[0030] The detection and evaluation submodule is used to acquire the patient's multimodal bioelectrical signals in real time, extract pain biomarker response values ​​from the multimodal bioelectrical signals, compare the pain biomarker response values ​​with a preset pain threshold, and generate a pain warning signal when the pain biomarker response value is greater than the pain threshold, and send it to the warning submodule.

[0031] The early warning submodule is used to output the pain early warning signal in a variety of preset early warning modes.

[0032] Preferably, the preset warning methods in the warning submodule include visual warning, tactile warning and auditory warning.

[0033] In the threshold setting submodule, the patient's bioelectrical signal during the pain-free period is used as a baseline to calculate the maximum response value of its pain biomarkers, which serves as the pain threshold. This pain threshold can also be dynamically adjusted according to the patient's condition to ensure the accuracy of pain status assessment. In the detection and assessment submodule, the patient's current bioelectrical signal is acquired in real time, and the response value of the pain biomarkers is extracted and compared with the pain threshold. Once the pain threshold is exceeded, a pain warning signal is generated and output through the warning submodule in various preset warning modes to promptly remind the patient that pain intervention measures are needed. This helps patients to be reminded before pain occurs and take appropriate measures to reduce pain or prevent pain from worsening.

[0034] Preferably, the pain intervention module includes:

[0035] The scenario matching submodule is used to, upon receiving the pain warning signal, traverse each desired scenario video in the desired scenario video resource library, calculate the similarity between the pain keywords in the pain questionnaire and the topic tags of the desired scenario video, sort the similarity from largest to smallest, and use the top few desired scenario videos as recommended scenario videos.

[0036] The adjustment mode selection submodule is used by the patient to select an active adjustment pain relief mode or a passive viewing pain relief mode; in the active adjustment pain relief mode, the corresponding adjustment element is adjusted according to the magnitude of the response value of the pain biomarker; in the passive viewing pain relief mode, the corresponding adjustment element changes autonomously according to preset conditions.

[0037] The virtual reality intervention submodule is used to visually present changes in corresponding adjustment elements in the recommended context video in real time using virtual reality to alleviate pain.

[0038] The user can choose to passively watch recommended scenario videos to relieve pain, or actively adjust the corresponding regulatory elements based on the response values ​​of pain biomarkers to achieve pain relief through task training. Furthermore, virtual reality is used to visually present the changes of the corresponding regulatory elements in the recommended scenario videos in real time, enhancing the patient's involvement and achieving better analgesic effects.

[0039] The preset adjustment elements include one or more of the target object's speed, position, sharpness, and volume. For example, if the recommended scenario video is a snow scene video, the adjustment element is the speed at which snowflakes fall. When the response value of the pain biomarker is large, the snowflakes fall faster, and when the response value of the pain biomarker decreases, the snowflakes fall slower. If the recommended scenario video is a family photo, the adjustment element is the photo's sharpness. When the response value of the pain biomarker is large, the family photo is blurry, and when the response value of the pain biomarker decreases, the family photo becomes sharper.

[0040] Compared with the prior art, the beneficial effects of the technical solution of the present invention are:

[0041] This invention first establishes a personalized feedback scenario generation module to collect the test subject's pain experience data, generate desired scenario videos, and establish a desired scenario video resource library. Then, a personalized pain biomarker mining module simultaneously collects the patient's bioelectrical signals and corresponding pain questionnaires. Compared to traditional methods relying on patient verbal pain reports or collecting only single physiological signals, this provides a more comprehensive data foundation. Through machine learning, the bioelectrical signals are classified to extract the patient's personalized pain biomarkers, meeting the needs of individual patient differences. Next, an online pain monitoring module is set up to acquire the patient's current bioelectrical signals in real time, extract pain biomarker response values, compare them with preset pain thresholds, and output pain warning signals based on the comparison results, meeting the need for timely reminders and intervention, providing alerts before pain occurs. Finally, a pain intervention module is set up. Upon receiving the pain warning signal, it matches the desired scenario video resource library based on the pain questionnaire to obtain recommended scenario videos that are easier to relieve pain. A preset pain relief mode is selected to personalize the adjustment elements corresponding to the recommended scenario video to better match the current patient characteristics. Virtual reality is used to visually present the changes of the adjustment elements in the recommended scenario video in real time to alleviate pain. Finally, by leveraging the high immersion of virtual reality, personalized situational videos are presented to patients in real time based on changes in current brain activity, enhancing patient involvement and thus achieving better analgesic effects. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the structure of a personalized biofeedback analgesia system in a virtual reality environment as described in Example 1;

[0043] Figure 2 This is an execution flowchart of a personalized biofeedback analgesia system in a virtual reality environment as described in Example 1;

[0044] Figure 3This is a schematic diagram of the structure of a personalized biofeedback analgesia system in a virtual reality environment as described in Example 2.

[0045] Figure 4 This is the execution flowchart of the personalized feedback scenario generation module described in Example 2;

[0046] Figure 5 This is the execution flowchart of the multi-state information acquisition submodule described in Example 2;

[0047] Figure 6 This is a schematic diagram of the pain biomarkers described in Example 2;

[0048] Figure 7 This is the execution flowchart of the online pain monitoring module described in Example 2;

[0049] Figure 8 This is the execution flowchart of the pain intervention module described in Example 2. Detailed Implementation

[0050] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent.

[0051] To better illustrate this embodiment, some parts in the accompanying drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions;

[0052] It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings.

[0053] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0054] Example 1

[0055] This embodiment provides a personalized biofeedback analgesia system in a virtual reality environment, such as Figure 1 As shown, it includes:

[0056] The personalized feedback scenario generation module is used to collect the tester's pain experience data, generate expected scenario videos based on the pain experience data, set corresponding adjustment elements, form an expected scenario video resource library, and continuously update and supplement the original expected video resource library based on the patient's pain experience data.

[0057] A personalized pain biomarker mining module is used to collect multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states, and to determine pain biomarkers from the multimodal bioelectrical signals based on machine learning methods.

[0058] The online pain monitoring module is used to acquire the patient's current multimodal bioelectrical signals in real time, extract pain biomarker response values ​​from the current multimodal bioelectrical signals, compare them with a preset pain threshold, and output a pain warning signal based on the comparison result; wherein, the pain threshold is set based on the pain biomarkers;

[0059] The pain intervention module is used to, upon receiving the pain warning signal, match the desired situation video resource library based on the pain questionnaire to obtain recommended situation videos; select a preset pain relief mode to adjust the adjustment elements corresponding to the recommended situation videos; and use virtual reality to visually present the changes of the adjustment elements in the recommended situation videos in real time to relieve pain.

[0060] In the specific implementation process, such as Figure 2 As shown, a personalized feedback scenario generation module is first set up to collect the test subject's pain experience data, generate desired scenario videos, and establish a desired scenario video resource library. Then, a personalized pain biomarker mining module simultaneously collects the patient's bioelectrical signals and corresponding pain questionnaires. Compared to traditional methods that rely on patients' verbal descriptions of pain or only collect single physiological signals, this provides a more comprehensive data foundation. Through machine learning, the bioelectrical signals are classified to extract the patient's personalized pain biomarkers, meeting the needs of individual patient differences. Next, an online pain monitoring module is set up to acquire the patient's current bioelectrical signals in real time, extract pain biomarker response values, compare them with preset pain thresholds, and output pain warning signals based on the comparison results, meeting the need for timely reminders and interventions, providing alerts before pain occurs. Finally, a pain intervention module is set up. Upon receiving the pain warning signal, it matches the desired scenario video resource library based on the pain questionnaire to obtain recommended scenario videos that are easier to relieve pain. A preset pain relief mode is selected to adjust the adjustment elements corresponding to the recommended scenario video, and virtual reality is used to visually present the changes of the adjustment elements in the recommended scenario video in real time to alleviate pain. Finally, the highly immersive nature of virtual reality is used to present the information to the patient in real time, enhancing the patient's involvement and thus achieving a better analgesic effect.

[0061] Example 2

[0062] This embodiment provides a personalized biofeedback analgesia system in a virtual reality environment, such as Figure 3 As shown, it includes:

[0063] A personalized feedback scenario generation module is used to collect pain experience data from test subjects, generate desired scenario videos based on the pain experience data, and set corresponding adjustment elements to form a desired scenario video resource library; including, for example... Figure 4 As shown:

[0064] A questionnaire survey was collected from several healthy test subjects and chronic pain test subjects; the questionnaire survey included at least a description of the expected scenario for pain relief;

[0065] Using the Wensheng video tool, the desired scene description is converted into a corresponding desired situation video, and at least one adjustment element is set accordingly;

[0066] Cluster all the desired scenario videos to obtain the topic categories and topic tags of the desired scenario videos;

[0067] Save all the desired scenario videos according to their theme categories to obtain a desired scenario video resource library;

[0068] The existing video resource library is continuously updated and supplemented based on patients' pain experience data.

[0069] A personalized pain biomarker mining module is used to collect multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states, and to determine pain biomarkers from the multimodal bioelectrical signals based on machine learning methods; including:

[0070] The multi-state information acquisition submodule is used to collect multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states. The pain questionnaire includes pain intensity scores and pain keywords; specifically, such as... Figure 5 As shown:

[0071] Bioelectrical signals of the patient were collected multiple times during the pain attack and non-pain attack periods, with the same preset duration.

[0072] During each bioelectric signal acquisition process, the patient selects the corresponding pain intensity score and pain keywords;

[0073] Each bioelectric signal and its corresponding pain intensity score are treated as a bioelectric data set with a pain intensity label, and all bioelectric data sets with pain intensity labels are combined to form a bioelectric dataset.

[0074] The pain intensity score ranges from 0 to 10 points, where 0 points indicates no pain, 1 to 3 points indicates mild pain, 4 to 6 points indicates moderate pain, 7 to 9 points indicates severe pain, and 10 points indicates unbearable pain.

[0075] The preprocessing submodule is used to perform preprocessing operations on multimodal bioelectric signals under different states to obtain preprocessed bioelectric signals;

[0076] The preprocessing operations include filtering, bad segment removal, and independent principal component analysis.

[0077] The pain biomarker extraction submodule is used to input preprocessed bioelectrical signals and corresponding pain intensity scores into a constructed classifier to obtain the patient's pain biomarkers; specifically:

[0078] Bioelectrical data labeled with pain intensity are used to form a bioelectrical dataset, which is then input into a constructed classifier. This establishes a mapping relationship between different pain intensities and bioelectrical signals, learns the characteristics of bioelectrical signals under different pain intensities, and obtains pain biomarkers, including one or more of the following: rhythmic features, network features, cross-band coupling features, and microstate features. Figure 6 As shown.

[0079] The study collects bioelectrical signals from patients during and between pain attacks, as well as their subjective pain intensity scores, to more precisely measure the dynamic fluctuations in their pain state. Once the pain state changes, the patient provides an updated pain intensity score via keyboard, ranging from 0 to 10. Each bioelectrical signal and its corresponding pain intensity score is treated as a bioelectrical data set labeled with pain intensity. For each patient, the bioelectrical signals induced by different pain physiological states are categorized to uncover personalized pain biomarkers: the collected bioelectrical signals are preprocessed, and then the bioelectrical data labeled with pain intensity is used as input to train a classifier. During training, the classifier learns the characteristics of bioelectrical signals under different pain states to effectively distinguish them. Labeled pain intensity data can be used as a supervisory signal during model training to help the model learn the mapping relationship between electrophysiological data and pain intensity. Machine learning methods, such as Support Vector Machines (SVM), Random Forest, Convolutional Neural Networks (CNN), and Naive Bayes classifiers, will be used to process multi-channel electrophysiological data and extract features related to pain intensity. To improve model reliability, we will address the challenge posed by the high dimensionality of bioelectrical data in risk factor selection. We propose an automatic kernel function selection technique based on Bayesian inference, utilizing the probabilistic correlations between dimensions to automatically identify important risk factors. This will extract personalized pain biomarkers for each patient, meeting the needs of individual patient differences. In this way, the classifier can establish a mapping relationship between different pain states and bioelectrical signals, thereby finding personalized pain biomarkers for each patient. The extraction of these biomarkers helps us more accurately understand each patient's pain response, providing strong support for personalized pain assessment and management.

[0080] The online pain monitoring module is used to acquire the patient's current multimodal bioelectrical signals in real time, extract pain biomarker response values ​​from the current multimodal bioelectrical signals, compare them with a preset pain threshold, and output a pain warning signal based on the comparison result; wherein, setting the pain threshold based on the pain biomarkers includes:

[0081] The threshold setting submodule is used to set the maximum response value of pain biomarkers during the pain-free period as the pain threshold and send it to the detection and evaluation submodule.

[0082] The detection and evaluation submodule is used to acquire the patient's multimodal bioelectrical signals in real time, extract pain biomarker response values ​​from the multimodal bioelectrical signals, compare the pain biomarker response values ​​with a preset pain threshold, and generate a pain warning signal when the pain biomarker response value is greater than the pain threshold, and send it to the warning submodule.

[0083] The early warning submodule is used to output the pain early warning signal in a variety of preset early warning modes.

[0084] like Figure 7 As shown, in the threshold setting submodule, the patient's bioelectrical signal during the non-painful period is used as a baseline to calculate the maximum response value of its pain biomarkers, which is then used as the pain threshold. This pain threshold can also be dynamically adjusted according to the patient's condition to ensure the accuracy of pain status assessment. In the detection and assessment submodule, the patient's current bioelectrical signal is acquired in real time, and the response value of the pain biomarkers is extracted from it and compared with the pain threshold. Once the pain threshold is exceeded, a pain warning signal is generated and output through the warning submodule in various preset warning modes to promptly remind the patient that pain intervention measures are needed. This helps patients to be reminded before pain occurs and take corresponding measures to reduce the pain or prevent the pain from worsening.

[0085] The pain intervention module, upon receiving the pain warning signal, matches a pain questionnaire against a desired scenario video resource library to obtain recommended scenario videos; selects a preset pain relief mode to adjust the adjustment elements corresponding to the recommended scenario videos; and utilizes virtual reality to visually present the changes of the adjustment elements in the recommended scenario videos in real time to alleviate pain. Figure 8 As shown, it includes:

[0086] The scenario matching submodule is used to, upon receiving the pain warning signal, traverse each desired scenario video in the desired scenario video resource library, calculate the similarity between the pain keywords in the pain questionnaire and the topic tags of the desired scenario video, sort the similarity from largest to smallest, and use the top few desired scenario videos as recommended scenario videos.

[0087] The adjustment mode selection submodule is used by the patient to select an active adjustment pain relief mode or a passive viewing pain relief mode; in the active adjustment pain relief mode, the corresponding adjustment element is adjusted according to the magnitude of the response value of the pain biomarker; in the passive viewing pain relief mode, the corresponding adjustment element changes autonomously according to preset conditions.

[0088] The virtual reality intervention submodule is used to visually present changes in corresponding adjustment elements in the recommended context video in real time using virtual reality to alleviate pain.

[0089] The user can choose to passively watch recommended scenario videos to relieve pain, or actively adjust the corresponding regulatory elements based on the response values ​​of pain biomarkers to achieve pain relief through task training. Furthermore, virtual reality is used to visually present the changes of the corresponding regulatory elements in the recommended scenario videos in real time, enhancing the patient's involvement and achieving better analgesic effects.

[0090] The preset adjustment elements include one or more of the target object's speed, position, sharpness, and volume. For example, if the recommended scenario video is a snow scene video, the adjustment element is the speed at which snowflakes fall. When the response value of the pain biomarker is large, the snowflakes fall faster, and when the response value of the pain biomarker decreases, the snowflakes fall slower. If the recommended scenario video is a family photo, the adjustment element is the photo's sharpness. When the response value of the pain biomarker is large, the family photo is blurry, and when the response value of the pain biomarker decreases, the family photo becomes sharper.

[0091] Example 3

[0092] This embodiment provides a personalized biofeedback analgesia system in a virtual reality environment, mainly explaining the principles of the personalized feedback scenario generation module and the pain intervention module: We simultaneously perform personalized VR customization based on individual pain characteristics. The specific implementation method is as follows: We plan to recruit 100 healthy subjects and 100 chronic pain patients, and collect pain-related personality traits and happiness scenario preference questionnaires for pain relief through questionnaire surveys. The main questionnaires are: pain sensitivity questionnaire, pain fear questionnaire, pain catastrophizing questionnaire, pain alertness and awareness questionnaire, pain anxiety symptom scale, and happiness scenario preference questionnaire. The Happiness Scene Preference Questionnaire contains two questions: For healthy participants, "Imagine you are currently experiencing a very painful period. What are 5-6 scenes you would most like to see at this time? Please describe them in as much detail as possible." For patients with chronic pain, "Please recall as many as possible 5-6 scenes that would best alleviate your pain experience when you are experiencing chronic pain." Subsequently, using text-based video tools such as Sora, the scene descriptions were converted into corresponding scene videos, ultimately generating 200-300 scene videos. The text descriptions underwent preprocessing, including text cleaning and conversion to numerical form, such as using TF-IDF (Term Frequency-Inverse Document Frequency) or word embedding techniques. Through cluster analysis and topic models (such as Latent Dirichlet Allocation, LDA), the samples were grouped according to pain traits, and the main themes of happiness scenes were identified to explore the differences in scene descriptions among different pain trait groups. A similarity model was trained, for example using machine learning algorithms (such as K-Nearest Neighbors, K-NN, or Support Vector Machine, SVM), to learn the relationship between pain questionnaire scores and video features. The recommendation system is built based on content-based recommendation (based on the matching degree between video features and the questionnaire scores of the participants) and collaborative filtering recommendation (based on historical user feedback data), that is, predicting and recommending the three most matching videos based on the questionnaire scores.

[0093] The adjustment mode selection submodule allows patients to choose between an active adjustment mode or a passive viewing mode for pain relief. In the active adjustment mode, the corresponding adjustment elements are adjusted based on the magnitude of the pain biomarker response values. In the passive viewing mode, the corresponding adjustment elements change autonomously according to preset conditions. The autonomous selection of the adjustment mode allows users to passively watch recommended scenario videos for pain relief, or actively adjust the corresponding adjustment elements based on the magnitude of the pain biomarker response values ​​through task training to achieve pain relief. Furthermore, virtual reality is used to visually present the changes of the corresponding adjustment elements in the recommended scenario videos in real time, enhancing patient involvement and achieving a better analgesic effect. The preset adjustment elements include one or more of the target object's speed, position, sharpness, and volume. For example, if the recommended scenario video is a snow scene video, the adjustment element is the speed at which snowflakes fall. When the response value of the pain biomarker is large, the snowflakes fall faster, and when the response value of the pain biomarker decreases, the snowflakes fall slower. If the recommended scenario video is a family photo, the adjustment element is the photo's sharpness. When the response value of the pain biomarker is large, the family photo is blurry, and when the response value of the pain biomarker decreases, the family photo becomes sharper.

[0094] Finally, virtual reality is used to visually present the changes of corresponding adjustment elements in the recommended context video in real time. Virtual Reality (VR) is a computer system that integrates computer graphics, optoelectronic imaging technology, sensing technology, computer simulation, artificial intelligence, and other technologies to create a realistic experience with multiple senses such as sight, hearing, touch, smell, and taste. People immerse themselves in the virtual environment using various interactive devices, interacting with entities within the virtual environment and experiencing sensations equivalent to those in the real physical environment. Therefore, the high immersion of VR technology makes it an ideal medium for presenting patients' neurophysiological signals in neurofeedback technology.

[0095] VR-based personalized neurofeedback enables users to more intuitively and concretely perceive changes in their real-time state, guiding them to learn to shape their neural activity in a desired manner to a certain extent. During use, patients wear virtual reality devices and enter a digital environment; once the system detects that the pain biomarker response value exceeds the pain threshold, an intervention program is initiated; the high immersion of VR enhances individual involvement, thereby achieving better analgesic effects.

[0096] The same or similar labels correspond to the same or similar parts;

[0097] The terms used to describe positional relationships in the accompanying drawings are for illustrative purposes only and should not be construed as limiting this patent.

[0098] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A personalized biofeedback analgesia system in a virtual reality environment, characterized in that, include: The personalized feedback scenario generation module is used to collect the tester's pain experience data, generate desired scenario videos based on the pain experience data, and set corresponding adjustment elements to form a desired scenario video resource library. A personalized pain biomarker mining module is used to collect multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states, and to determine pain biomarkers from the multimodal bioelectrical signals based on machine learning methods. The personalized pain biomarker mining module includes: The multi-state information acquisition submodule is used to collect multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states. The pain questionnaire includes pain intensity scores and pain keywords. The preprocessing submodule is used to perform preprocessing operations on multimodal bioelectric signals under different states to obtain preprocessed bioelectric signals; The pain biomarker extraction submodule is used to input the preprocessed bioelectrical signals and the corresponding pain intensity scores into the constructed classifier to obtain the patient's pain biomarkers; The online pain monitoring module is used to acquire the patient's current multimodal bioelectrical signals in real time, extract pain biomarker response values ​​from the current multimodal bioelectrical signals, compare them with a preset pain threshold, and output a pain warning signal based on the comparison result; wherein, the pain threshold is set based on the pain biomarkers; The pain intervention module is used to, upon receiving the pain warning signal, match the desired situation video resource library based on the pain questionnaire to obtain recommended situation videos; select a preset pain relief mode to adjust the adjustment elements corresponding to the recommended situation videos, and use virtual reality to visually present the changes of the adjustment elements in the recommended situation videos in real time to relieve pain. The pain intervention module includes: The scenario matching submodule is used to, upon receiving the pain warning signal, traverse each desired scenario video in the desired scenario video resource library, calculate the similarity between the pain keywords in the pain questionnaire and the topic tags of the desired scenario video, sort the similarity from largest to smallest, and use the top few desired scenario videos as recommended scenario videos. The adjustment mode selection submodule is used by the patient to select an active adjustment pain relief mode or a passive viewing pain relief mode; in the active adjustment pain relief mode, the corresponding adjustment element is adjusted according to the magnitude of the response value of the pain biomarker; in the passive viewing pain relief mode, the corresponding adjustment element changes autonomously according to preset conditions. The virtual reality intervention submodule is used to visually present changes in corresponding adjustment elements in the recommended context video in real time using virtual reality to alleviate pain.

2. The personalized biofeedback analgesia system in a virtual reality environment according to claim 1, characterized in that, The multi-state information acquisition submodule collects multimodal bioelectrical signals and corresponding pain questionnaires from patients in different states. The pain questionnaire includes pain intensity scores and pain keywords, including: Bioelectrical signals of the patient were collected multiple times during the pain attack and non-pain attack periods, with the same preset duration. During each bioelectric signal acquisition process, the patient selects the corresponding pain intensity score and pain keywords; Each bioelectric signal and its corresponding pain intensity score are treated as a bioelectric data set with a pain intensity label, and all bioelectric data sets with pain intensity labels are combined to form a bioelectric dataset.

3. The personalized biofeedback analgesia system in a virtual reality environment according to claim 2, characterized in that, The pain intensity score ranges from 0 to 10 points, where 0 points indicates no pain, 1 to 3 points indicates mild pain, 4 to 6 points indicates moderate pain, 7 to 9 points indicates severe pain, and 10 points indicates unbearable pain.

4. The personalized biofeedback analgesia system in a virtual reality environment according to claim 1, characterized in that, The preprocessing submodule includes preprocessing operations such as filtering, bad segment removal, and independent principal component analysis.

5. The personalized biofeedback analgesia system in a virtual reality environment according to claim 3, characterized in that, In the pain biomarker extraction submodule, the preprocessed bioelectrical signals and corresponding pain intensity scores are input into the constructed classifier to obtain the patient's pain biomarkers, including: Bioelectric data labeled with pain intensity are used to form a bioelectric dataset, which is then input into a constructed classifier. This establishes a mapping relationship between different pain intensities and bioelectric signals, learns the characteristics of bioelectric signals under different pain intensities, and obtains pain biomarkers, including one or more of the following: rhythmic features, network features, cross-band coupling features, and microstate features.

6. The personalized biofeedback analgesia system in a virtual reality environment according to claim 2, characterized in that, The online pain monitoring module includes: The threshold setting submodule is used to set the maximum response value of pain biomarkers during the pain-free period as the pain threshold and send it to the detection and evaluation submodule. The detection and evaluation submodule is used to acquire the patient's multimodal bioelectrical signals in real time, extract pain biomarker response values ​​from the multimodal bioelectrical signals, compare the pain biomarker response values ​​with a preset pain threshold, and generate a pain warning signal when the pain biomarker response value is greater than the pain threshold, and send it to the warning submodule. The early warning submodule is used to output the pain early warning signal in a variety of preset early warning modes.

7. The personalized biofeedback analgesia system in a virtual reality environment according to claim 6, characterized in that, The pre-defined warning methods in the warning submodule include visual warning, tactile warning, and auditory warning.

8. The personalized biofeedback analgesia system in a virtual reality environment according to claim 1, characterized in that, The personalized feedback scenario generation module collects the test subject's pain experience data, generates desired scenario videos based on the pain experience data, and sets corresponding adjustment elements to form an iterative desired scenario video resource library, including: A questionnaire survey was collected from several healthy test subjects and chronic pain test subjects; the questionnaire survey included at least a description of the expected scenario for pain relief; Using the Wensheng video tool, the desired scene description is converted into a corresponding desired situation video, and at least one adjustment element is set accordingly; Cluster all the desired scenario videos to obtain the topic categories and topic tags of the desired scenario videos; Save all the desired scenario videos according to their theme categories to obtain a desired scenario video resource library.

Images