A smoking behavior intervention and assessment method and electronic device

Abstract

The present invention provides a smoking behavior intervention and assessment method and electronic device, comprising: a client obtaining a user's personalized daily smoking scenes and behavior information; the client constructing personalized smoking imagery instructions and generating AI-synthesized audio; the client obtaining the user's EEG signals while performing smoking imagery under the guidance of the AI-synthesized audio, and uploading them to a cloud server for storage; the client using the AI-synthesized audio to perform smoking behavior intervention training based on smoking imagery; after the smoking behavior intervention training, the client obtaining the user's EEG signals while performing smoking imagery under the guidance of the AI-synthesized audio, and uploading them to a cloud server for storage; the cloud server pre-processing, feature extraction, and comparative analysis of the EEG signals before and after the smoking behavior intervention training, and outputting quantified intervention training effects to the client. The present invention facilitates users to independently conduct smoking behavior intervention training and can effectively reduce their smoking cravings.

Description

Technical Field

[0001] The present invention relates to the fields of psychology intervention technology and electroencephalogram signal processing technology, and in particular to a smoking behavior intervention and evaluation method and electronic equipment. Background Art

[0002] In the field of smoking behavior intervention, behavioral intervention methods have advantages such as no side effects and no drug tolerance, and therefore have very broad application prospects. Among behavioral intervention methods, imagery-based behavioral intervention training has received widespread attention due to its low cost, simple operation and significant effect. Usually, imagery-based behavioral intervention training needs to be carried out under the operation and guidance of professionals. Its labor cost is high, the use environment is also limited, and professionals are relatively scarce. It cannot meet the smoking behavior intervention training needs of a large number of users. There is a lack of an intelligent and automated method and system that facilitates users to conduct smoking behavior intervention training independently. In addition, the current imagery-based behavioral intervention training lacks feedback from objective physiological indicators, which often makes it impossible for users to evaluate the true effect of the intervention training. Summary of the Invention

[0003] This invention provides a smoking behavior intervention and assessment method and electronic device. These methods primarily address the lack of intelligent and automated imagery-based behavioral intervention training methods and systems in the smoking behavior intervention field, as well as the lack of objective physiological indicator feedback in imagery-based behavioral intervention training. This method enables users to independently conduct smoking behavior intervention training, and also provides a physiological indicator analysis method based on EEG signal characteristics, which can provide users with quantitative feedback on the intervention training effects. This invention offers advantages such as autonomy, universal applicability, simple user operation, and a high degree of intelligence and automation.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] A smoking behavior intervention and assessment method comprises the following steps:

[0006] S1, the client obtains the user's personalized daily smoking scene and behavior information;

[0007] S2, the client constructs personalized smoking imagery guidance and generates AI-synthesized audio;

[0008] S3, before smoking behavior intervention training, the client obtains the user's EEG signals of smoking imagery under the guidance of AI synthesized audio and uploads them to the cloud server for storage;

[0009] S4, the client uses AI-synthesized audio to conduct smoking behavior intervention training based on smoking imagery;

[0010] S5, after the smoking behavior intervention training, the client obtains the EEG signals of the user performing smoking imagery under the guidance of AI synthesized audio and uploads them to the cloud server for storage;

[0011] S6, the cloud server preprocesses, extracts features and conducts comparative analysis on the EEG signals before and after the smoking behavior intervention training, and outputs the quantified intervention training effect to the client.

[0012] Furthermore, S1 includes: on the client, the user selects an experience in daily life that makes him most want to smoke, and enters key information of this experience, including time and place, what he was doing before smoking, things that triggered the idea of ​​smoking, physiological reactions when he wanted to smoke, and actions and feelings when smoking.

[0013] Furthermore, S2 includes: the client integrates the information input by the user into a complete imagery instruction according to a standard structural framework; the standard structural framework is: the first part describes the time and place, the second part describes what is being done before smoking, the third part describes the things that trigger the idea of ​​smoking, the fourth part describes the physiological reaction when wanting to smoke, and the fifth part describes the actions and feelings when smoking; then, the imagery instruction is uploaded to the AI ​​intelligent voice generation system to generate AI synthesized audio, the tone is female, and the speaking speed is 140 words / minute.

[0014] Furthermore, S3 includes: the client obtains the EEG signal of the user performing smoking imagery under the guidance of AI synthesized audio, and uploads it to the cloud server for storage.

[0015] Furthermore, the client connects to a wireless Bluetooth headset via Bluetooth, plays AI synthesized audio to the user, and guides the user to engage in smoking imagery for 5 minutes to induce the user's craving for smoking; at the same time, the client connects to a smart wearable EEG device via Bluetooth, uses dry electrodes and samples the user's EEG signals at a sampling rate of 1000 Hz, amplifies, digitizes and filters the signals through an amplifier, and obtains the EEG signals of the user engaging in smoking imagery under the guidance of AI synthesized audio before smoking behavior intervention training; then, the client uploads the EEG signals to the cloud server for storage.

[0016] Furthermore, the S4 includes: providing the user with two imagery-based smoking behavior intervention trainings on the client side, and the user can freely choose any one of the smoking behavior intervention trainings.

[0017] Furthermore, the two imagery-based smoking behavior intervention trainings are as follows: the first intervention training is an imagery-based exposure method, which lasts about 60 minutes. The client connects to a wireless Bluetooth headset via Bluetooth, plays AI-synthesized audio to the user, and guides the user to engage in smoking imagery for 5 minutes, followed by a 1-minute break. The above process is repeated 10 times before the intervention training ends.

[0018] The second type of intervention training is based on the extraction-extinction method of imagery, which lasts about 75 minutes. The client connects to the wireless Bluetooth headset via Bluetooth, plays AI synthesized audio to the user, guides the user to engage in 5 minutes of smoking imagery, and then lets the user rest for 10 minutes. Next, guide the user to engage in 5 minutes of smoking imagery, and then let the user rest for 1 minute, and repeat the process; after 10 times, the intervention training ends.

[0019] Furthermore, S5 includes: the client obtains the EEG signal of the user who is performing smoking imagery under the guidance of AI synthesized audio, and uploads it to the cloud server for storage. Specifically, the client connects to a wireless Bluetooth headset via Bluetooth, plays AI synthesized audio to the user, and guides the user to perform smoking imagery for 5 minutes to induce the user's smoking craving; at the same time, the client connects to a smart wearable EEG device via Bluetooth, uses dry electrodes and samples the user's EEG signal at a sampling rate of 1000 Hz, amplifies, digitizes, and filters the signal through an amplifier, and obtains the EEG signal of the user who is performing smoking imagery under the guidance of AI synthesized audio after smoking behavior intervention training; the client then uploads the EEG signal to the cloud server for storage.

[0020] Furthermore, the S6 includes: on the cloud server, for the EEG signals before and after the smoking behavior intervention training, the preprocessing steps are: high-pass filtering and low-pass filtering of the EEG signals; reducing the sampling rate of the EEG signals to 250 Hz; performing average re-referencing; using a traditional recursive least squares algorithm to correct blink artifacts; segmenting the EEG signals, each segment being 1 second long; using an automatic detection algorithm to remove segments containing noise; and performing feature extraction on the EEG signals after the preprocessing. The specific steps are: first calculating the global explained variance GFP, the specific formula is as follows:

[0021]

[0022] In the above formula, N represents the number of electrodes, μ i represents the potential of electrode i at a given time point, Represents the average potential of all electrodes, and the output of the formula is the global explained variance time series;

[0023] Next, the EEG maps corresponding to all peaks in the global explained variance time series were input into the k-means clustering algorithm, where k was set to 4, resulting in four standard EEG map categories A, B, C, and D. The spatial correlation coefficient between each peak and the four standard maps was calculated, and the peak was assigned to the category with the highest spatial correlation coefficient. The troughs between peaks were considered to be the transition time points between different categories. Finally, the average duration of category C in the global explained variance time series was calculated. After calculating the average duration of category C for the EEG signals under the two conditions before and after the smoking behavior intervention training, the quantitative intervention training effect indicator E was calculated using the formula:

[0024]

[0025] Since the average duration of category C is an EEG signal feature closely related to smoking cravings, the numerical value of the smoking behavior intervention training effect indicator E reflects the degree of change in smoking cravings; after the calculation is completed, the cloud server sends indicator E to the client, and the client provides feedback on the quantified training intervention effect to the user.

[0026] The present invention also provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned smoking behavior intervention and evaluation method when executing the program.

[0027] Beneficial effects:

[0028] This invention makes imagery-based behavioral intervention training intelligent and automated, significantly resolving the issue of a shortage of specialized behavioral intervention training professionals unable to meet the needs of a large number of users. It facilitates users to independently conduct smoking behavior intervention training, effectively reducing smoking cravings. Furthermore, it provides a method for analyzing physiological indicators based on EEG signal characteristics, providing users with quantitative feedback on the effectiveness of intervention training. This invention offers advantages such as autonomy, universal applicability, simple user operation, intelligence, and a high degree of automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 The present invention is a flowchart of a smoking behavior intervention and evaluation method. DETAILED DESCRIPTION

[0030] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0031] like Figure 1 As shown, a smoking behavior intervention and assessment method of the present invention specifically includes the following steps:

[0032] S1, the client obtains the user's personalized daily smoking scene and behavior information;

[0033] S2, the client constructs personalized smoking imagery guidance and generates AI-synthesized audio;

[0034] S3, before smoking behavior intervention training, the client obtains the user's EEG signals of smoking imagery under the guidance of AI synthesized audio and uploads them to the cloud server for storage;

[0035] S4, the client uses AI-synthesized audio to conduct smoking behavior intervention training based on smoking imagery;

[0036] S5, after the smoking behavior intervention training, the client obtains the EEG signals of the user performing smoking imagery under the guidance of AI synthesized audio and uploads them to the cloud server for storage;

[0037] S6, the cloud server preprocesses, extracts features and conducts comparative analysis on the EEG signals before and after the smoking behavior intervention training, and outputs the quantified intervention training effect to the client.

[0038] Specifically, S1 involves the user selecting a daily experience in their life when they most wanted to smoke and entering key information about the experience, including the time and location, what they were doing before smoking, what triggered the urge to smoke, their physiological reactions to the urge, and the movements and feelings they felt while smoking. To standardize the input, the client provides five standard description examples for users' reference.

[0039] S2 includes: the client integrates the user input information into a complete imagery instruction according to a standard structural framework. The standard structural framework is: the first part describes the time and place, the second part describes what is being done before smoking, the third part describes the things that trigger the idea of ​​smoking, the fourth part describes the physiological reaction when wanting to smoke, and the fifth part describes the actions and feelings when smoking. Subsequently, the imagery instruction is uploaded to the AI ​​intelligent voice generation system to generate AI synthesized audio (duration 5 minutes), the tone is female, and the speaking speed is 140 words / minute.

[0040] The S3 includes: the client obtains the EEG signal of the user who is performing smoking imagery under the guidance of AI synthesized audio, and uploads it to the cloud server for storage. Specifically, the client connects to a wireless Bluetooth headset (such as a Sennheiser wireless Bluetooth headset) via Bluetooth, plays AI synthesized audio to the user, and guides the user to perform a 5-minute smoking image to induce the user's smoking craving. At the same time, the client connects to a smart wearable EEG device (such as Muse and Emotive brain ring devices) via Bluetooth, uses dry electrodes and samples the user's EEG signal at a sampling rate of 1000HZ, amplifies, digitizes and filters the signal through an amplifier, and obtains the EEG signal of the user who is performing smoking imagery under the guidance of AI synthesized audio before smoking behavior intervention training. Subsequently, the client uploads the EEG signal to the cloud server for storage.

[0041] Wherein, the smoking image is:

[0042] Imagery is a conscious, vivid simulation of a stimulus or event in the brain, resulting in a vivid sensory experience. Imagery can be divided into guided imagery and free imagery. Guided imagery relies on relevant instructions, so the imagery content can be manipulated. In the field of smoking behavior research, smoking-related imagery is a conditioned stimulus that can effectively induce smoking cravings in smokers. When smokers experience smoking cravings, they experience a series of physiological and psychological reactions, including an increase in subjective craving scores, changes in heart rate, and changes in EEG signal characteristics.

[0043] The S4 includes: on the client side, providing users with two imagery-based smoking behavior intervention trainings, and users can freely choose any one of the smoking behavior intervention trainings. Specifically, the first intervention training is an imagery-based exposure method, which lasts about 60 minutes. Its main form is: the client connects to a wireless Bluetooth headset (such as a Sennheiser wireless Bluetooth headset) via Bluetooth, plays AI synthesized audio to the user, guides the user to perform a 5-minute smoking image, and then lets the user rest for 1 minute, and repeats the above process (ie, 5 minutes of smoking imagery and 1 minute of rest). After 10 times, the intervention training ends. The second type of intervention training is based on the extraction-extinction method of imagery, which lasts about 75 minutes. Its main form is: the client connects to a wireless Bluetooth headset (such as a Sennheiser wireless Bluetooth headset) via Bluetooth, plays AI synthesized audio to the user, guides the user to engage in 5 minutes of smoking imagery, and then lets the user rest for 10 minutes. Next, guide the user to engage in 5 minutes of smoking imagery, and then let the user rest for 1 minute. After repeating the process (i.e. 5 minutes of smoking imagery and 1 minute of rest) 10 times, the intervention training ends.

[0044] The principles of the two behavioral intervention trainings are as follows:

[0045] (1) Imagery-based exposure

[0046] This method is based on the principle of counterconditioning. Its basic idea is: under experimental conditions, conditioned stimuli related to smoking (such as smoking image stimuli) are repeatedly presented to induce smoking cravings, but cigarettes are not provided and actual smoking behavior does not occur. Therefore, the connection between smoking-related conditioned stimuli and smoking cravings can be gradually weakened, thereby achieving the purpose of smoking behavior intervention.

[0047] (2) Imagery-based extraction-fading method

[0048] This method is based on the principle of memory reconsolidation, that is, smoking-related memories will be activated by smoking-related stimuli (such as smoking image stimuli). After activation, the memory will enter an unstable state and need to go through a reconsolidation process to return to a steady state. External intervention during the unstable state (such as repeated presentation of smoking-related stimuli) can interfere with the reconsolidation process, thereby weakening smoking-related memories and achieving the purpose of smoking behavior intervention.

[0049] The S5 includes: the client obtains the EEG signal of the user performing smoking imagery under the guidance of AI synthesized audio, and uploads it to the cloud server for storage. Specifically, the client connects to a wireless Bluetooth headset (such as a Sennheiser wireless Bluetooth headset) via Bluetooth, plays AI synthesized audio to the user, and guides the user to perform 5 minutes of smoking imagery to induce the user's smoking craving. At the same time, the client connects to a smart wearable EEG device (such as Muse and Emotive brain ring devices) via Bluetooth, uses dry electrodes and samples the user's EEG signal at a sampling rate of 1000HZ, amplifies, digitizes and filters the signal through an amplifier, and obtains the EEG signal of the user performing smoking imagery under the guidance of AI synthesized audio after smoking behavior intervention training. Subsequently, the client uploads the EEG signal to the cloud server for storage.

[0050] The S6 includes: the cloud server pre-processes, extracts features and conducts comparative analysis on the EEG signals before and after the smoking behavior intervention training, and outputs the quantitative intervention training effect to the client. Specifically, on the cloud server, for the EEG signals before and after the smoking behavior intervention training, the pre-processing steps are: high-pass filtering (2HZ) and low-pass filtering (20HZ) of the EEG signals; reducing the sampling rate of the EEG signals to 250HZ; performing average re-reference; using the traditional recursive least squares algorithm to correct blink artifacts; segmenting the EEG signals, each segment is 1 second long; using an automatic detection algorithm to eliminate segments containing noise, and the algorithm detection standard is that segments with EEG amplitude changes exceeding ±100mV are considered to contain noise. Next, feature extraction is performed on the EEG signals after the pre-processing is completed, and the specific steps are: first calculate the global explained variance (GFP), and the specific formula is as follows:

[0051]

[0052] In the above formula, N represents the number of electrodes, μ i represents the potential of electrode i at a given time point, Represents the average potential of all electrodes, and the output of the formula is a time series of global explained variance.

[0053] Next, the EEG maps corresponding to all peaks in the global explained variance time series were input into the k-means clustering algorithm, where k was set to 4, resulting in four standard EEG map categories: A, B, C, and D. The spatial correlation coefficient between each peak and the four standard maps was calculated, and the peak was assigned to the category with the highest spatial correlation coefficient. The troughs between peaks were considered to be the transition time points between different categories. Finally, the average duration (in milliseconds) of category C in the global explained variance time series was calculated. After calculating the average duration of category C for the EEG signals before and after the smoking behavior intervention training, the quantitative intervention training effect indicator E was calculated using the formula:

[0054]

[0055] Because the average duration of category C is an EEG signal characteristic closely related to smoking cravings, the numerical value of the smoking behavior intervention training effect indicator E reflects the degree of change in smoking cravings. After calculation, the cloud server sends indicator E to the client, which provides users with quantitative feedback on the training intervention effect.

[0056] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for intervention and assessment of smoking behavior, characterized in that: The steps include: S1, the client obtains the user's personalized daily smoking scene and behavior information; S2, the client constructs personalized smoking imagery guidance and generates AI-synthesized audio; S3, before smoking behavior intervention training, the client obtains the user's EEG signals of smoking imagery under the guidance of AI synthesized audio and uploads them to the cloud server for storage; S4, the client uses AI-synthesized audio to conduct smoking behavior intervention training based on smoking imagery; S5, after the smoking behavior intervention training, the client obtains the EEG signals of the user performing smoking imagery under the guidance of AI synthesized audio and uploads them to the cloud server for storage; S6. The cloud server preprocesses, extracts features, and conducts comparative analysis on the EEG signals before and after the smoking behavior intervention training, and outputs the quantified intervention training effect to the client. This includes: on the cloud server, the EEG signals before and after the smoking behavior intervention training are preprocessed as follows: high-pass filtering and low-pass filtering of the EEG signals; reducing the sampling rate of the EEG signals to 250 Hz; performing average re-referencing; using the traditional recursive least squares algorithm to correct blink artifacts; segmenting the EEG signals, each with a length of 1 second; using an automatic detection algorithm to remove segments containing noise; and extracting features from the EEG signals after preprocessing. The specific steps are as follows: first, calculating the global explained variance (GFP), using the following formula: （1） In the above formula, N represents the number of electrodes. represents the potential of electrode i at a given time point, Represents the average potential of all electrodes, and the output of the formula is the global explained variance time series; Next, the EEG maps corresponding to all peaks in the global explained variance time series were input into the k-means clustering algorithm, where k was set to 4, resulting in four standard EEG map categories A, B, C, and D. The spatial correlation coefficient between each peak and the four standard maps was calculated, and the peak was assigned to the category with the highest spatial correlation coefficient. The troughs between peaks were considered to be the transition time points between different categories. Finally, the average duration of category C in the global explained variance time series was calculated. After calculating the average duration of category C for the EEG signals under the two conditions before and after the smoking behavior intervention training, the quantitative intervention training effect indicator E was calculated using the formula: （2） Since the average duration of category C is an EEG signal feature closely related to smoking cravings, the numerical value of the smoking behavior intervention training effect indicator E reflects the degree of change in smoking cravings; after the calculation is completed, the cloud server sends indicator E to the client, and the client provides feedback on the quantified training intervention effect to the user.

2. The method for intervention and assessment of smoking behavior according to claim 1, characterized in that: The S1 includes: on the client, the user selects an experience in daily life that makes him most want to smoke, and enters key information of this experience, including time and place, what he was doing before smoking, things that triggered the idea of ​​smoking, physiological reactions when he wanted to smoke, and actions and feelings when smoking.

3. The method for intervention and assessment of smoking behavior according to claim 2, characterized in that: The S2 includes: the client integrates the information input by the user into a complete imagery instruction according to a standard structural framework; the standard structural framework is: the first part describes the time and place, the second part describes what is being done before smoking, the third part describes the things that trigger the idea of ​​smoking, the fourth part describes the physiological reaction when wanting to smoke, and the fifth part describes the actions and feelings when smoking; then, the imagery instruction is uploaded to the AI ​​intelligent voice generation system to generate AI synthesized audio.

4. The method for intervention and assessment of smoking behavior according to claim 3, characterized in that: The client connects to a wireless Bluetooth headset via Bluetooth, plays AI-synthesized audio to the user, and guides the user to engage in smoking imagery for 5 minutes to induce smoking cravings. Simultaneously, the client connects to a smart wearable EEG device via Bluetooth, samples the user's EEG signals using dry electrodes at a sampling rate of 1000 Hz, and amplifies, digitizes, and filters the signals through an amplifier to obtain EEG signals of the user engaging in smoking imagery under the guidance of AI-synthesized audio before smoking behavior intervention training. The client then uploads the EEG signals to the cloud server for storage.

5. The method for intervention and assessment of smoking behavior according to claim 4, characterized in that: The S4 includes: providing two imagery-based smoking behavior intervention trainings to the user on the client side, and the user can freely choose any one of the smoking behavior intervention trainings.

6. The method for intervention and assessment of smoking behavior according to claim 5, characterized in that: The two imagery-based smoking behavior intervention trainings are as follows: The first intervention training is an imagery-based exposure method, which lasts for 60 minutes. The client connects to a wireless Bluetooth headset via Bluetooth, plays AI-synthesized audio to the user, and guides the user to engage in smoking imagery for 5 minutes, followed by a 1-minute break. The above process is repeated 10 times before the intervention training ends. The second type of intervention training is based on the extraction-extinction method of imagery, which lasts about 75 minutes. The client connects to the wireless Bluetooth headset via Bluetooth, plays AI synthesized audio to the user, guides the user to engage in 5 minutes of smoking imagery, and then lets the user rest for 10 minutes. Next, guide the user to engage in 5 minutes of smoking imagery, and then let the user rest for 1 minute, and repeat the process; after 10 times, the intervention training ends.

7. The method for intervention and assessment of smoking behavior according to claim 6, characterized in that: The S5 includes: the client obtains the EEG signal of the user who is performing smoking imagery under the guidance of AI synthesized audio, and uploads it to the cloud server for storage; specifically, the client connects to a wireless Bluetooth headset via Bluetooth, plays the AI ​​synthesized audio to the user, and guides the user to perform smoking imagery for 5 minutes to induce the user's smoking craving; at the same time, the client connects to an intelligent wearable EEG device via Bluetooth, uses dry electrodes and samples the user's EEG signal at a sampling rate of 1000 Hz, amplifies, digitizes and filters the signal through an amplifier, and obtains the EEG signal of the user who is performing smoking imagery under the guidance of AI synthesized audio after smoking behavior intervention training; then, the client uploads the EEG signal to the cloud server for storage.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the smoking behavior intervention and assessment method according to any one of claims 1 to 7 are implemented.

Images