Practical satiety sense identification method based on physiological parameters

Abstract

The invention provides an actual satiety sense recognition method based on physiological parameters and a wearable intelligent health terminal, the wearable intelligent health terminal comprises a wrist terminal and a plurality of patches, the wrist terminal is worn on the wrist, and the small patches are worn on the abdomen. Comprising the following steps that 1, an intelligent health terminal monitors abdominal impedance, pulse waves, a blood oxygen value, a blood pressure value and a body movement value in real time; 2, at the time close to dining, biological impedance, pulse wave and blood pressure data of the wrist and the abdomen are obtained respectively, the data are summarized to a wrist terminal, and a wrist processor carries out data analysis; 3, according to the blood pressure data, the abdominal blood perfusion index is judged, the parsympathetic nerve excitation degree and the change value of the abdominal biological impedance are calculated through heart rate variability, and whether excessive satiation occurs or not is judged. The method has the advantages that the satiety state is automatically recognized by monitoring the physiological parameters in an undisturbed mode, the user is reminded, and good diet health management of the user is facilitated.

Description

Technical Field

[0001] The present invention relates to the field of health technologies, and particularly to a method for recognizing actual satiety based on physiological parameters. Background Art

[0002] Satiety is the feeling of not needing to eat during the process from the start to the end of eating. Due to the lag of the feeling, after being full, blood surges into the stomach, causing the brain to generate a satiety signal, thus making the human body not need to eat. At this time, enough food has been eaten, and eating enough food not only burdens the digestive organs.

[0003] One of the traditional Chinese health concepts and ways is to eat only until 70% full at each meal. When one is 70% full, the body has completely absorbed the nutrients in the food, and 70% fullness needs to be judged by one's own feeling, and this judgment usually has a certain deviation.

[0004] Judging satiety by feeling is not very accurate and is usually greatly affected by other factors, including subjective and objective factors. For example, when in a good mood, one is prone to overeating, and when in a bad mood, the food intake is insufficient. In addition, the weather also affects the food intake. The food intake is low in hot weather and high in cold weather. These factors make it impossible to judge an individual's satiety through subjective feeling. Summary of the Invention

[0005] By detecting parameter values such as impedance changes, blood flow direction, blood oxygen changes, blood pressure changes, and heart rate variability in the stomach of the present invention, and then combining with a deep learning algorithm to judge whether the 70% fullness level is reached, scientific and healthy eating is realized through the following technical solutions.

[0006] A method for recognizing actual satiety based on physiological parameters, wearing a wearable intelligent health terminal, wearing a wrist terminal on the wrist and a small patch on the abdomen, including the following steps: Step 1, the intelligent health terminal continuously monitors abdominal impedance, pulse wave, blood oxygen value, blood pressure value, and body movement value; Step 2, when it is judged as a non-sleep state according to the body movement value and it is close to the meal time, respectively obtain the bio-impedance, pulse wave, and blood pressure data of the wrist and abdomen. The data is aggregated to the wrist terminal through Bluetooth, and the data is analyzed by the wrist terminal processor; Step 3, calculate the blood perfusion through the pulse wave data, respectively analyze the blood perfusion indexes of the wrist and abdomen. After a full meal, the blood perfusion index of the abdomen will increase; after a full meal, it will also cause the parasympathetic nerve to be excited, and calculate the change in the degree of parasympathetic nerve excitement through heart rate variability; at the same time, due to the change in the volume of the stomach, the abdominal impedance will also change accordingly; analyze the change trends of parameters such as blood oxygen and blood pressure to judge whether overeating occurs; Step 4: After a preliminary judgment of satiety, a pop-up window is shown to the user to confirm whether they are full, and the user's confirmation result is used as the label for the above parameter set. The parameter set and the label of whether the user is full are input into a deep learning model, which includes but is not limited to common deep learning models such as convolutional neural networks and principal component analysis.

[0007] The beneficial effects of the present invention are: by non-intrusively monitoring physiological parameters to automatically identify the satiety state and giving reminders to users, it is beneficial for users to manage their diet and health well. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 is a flowchart of the actual satiety recognition method according to an embodiment of the present invention.

[0009] Figure 2 is a schematic diagram of a wrist-type terminal and an abdominal patch according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0010] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, the present invention can be implemented in many different ways defined and covered by the claims.

[0011] An actual satiety recognition method based on physiological parameters, wearing a wearable intelligent health terminal, which includes a wrist-type terminal and several patches. The wrist-type terminal is worn on the wrist, and multiple small patches are worn on the abdomen, including the following steps: Step 1: The intelligent health terminal continuously monitors abdominal impedance, pulse wave, blood oxygen value, blood pressure value, and body movement value; attention should be paid to the measurement methods below.

[0012] Step 2: Near meal time, the bioimpedance, pulse wave, and blood pressure data of the wrist and abdomen are respectively obtained. The data is aggregated to the wrist terminal, and the data analysis is performed by the wrist processor. Step 3: According to the blood pressure data, the abdominal blood perfusion index is judged, the degree of parasympathetic nerve excitation is calculated through heart rate variability, and the change value of abdominal bioimpedance is used to judge whether overeating occurs.

[0013] Among them, heart rate variability refers to the minute changes in the time intervals between two heartbeats, which is the result of the modulation of the autonomic nervous system on the sinus node automaticity. The steps of calculating the degree of parasympathetic nerve excitation through heart rate variability are: Collect electrocardiogram data: Collect electrocardiogram data through an electrocardiogram device, and record continuous heartbeats and time intervals.

[0014] Calculate heart rate variability: Analyze the electrocardiogram data, calculate the time intervals between adjacent heartbeats to form a time series, and then calculate the heart rate variability through time domain or frequency domain analysis methods.

[0015] Evaluating the degree of parasympathetic nerve excitation: Heart rate variability reflects the regulation of the autonomic nervous system on the heart. When the parasympathetic nerve is excited, the heart rate variability usually increases. Therefore, by comparing the heart rate variability index with the reference value or normal range, the degree of parasympathetic nerve excitation can be evaluated.

[0016] In the embodiment of the present invention, the wrist-type terminal starts the photoelectric pulse wave sensor, emits a light signal to the skin, obtains the light signal reflected by the skin, and according to the peak and trough positions of the reflected pulse wave signal, obtains a series of peak time intervals, namely RR intervals, calculates the HRV, calculates the standard deviation of the continuous RR intervals, and obtains the SDNN. The higher the SDNN, the more excited the sympathetic nerve is, and the more excited the human body is.

[0017] In the embodiment of the present invention, the method for measuring abdominal impedance includes: Attach 2 patches to the left wrist and attach the other 2 patches to the left ankle; Connect all the patches to the wrist-type terminal. The wrist-type terminal is powered on to emit low-voltage electricity, measures the resistivity ρ and dielectric constant ε between the patches, and obtains the body impedance. The method for measuring wrist impedance includes: Attach 2 patches to the left wrist and attach the other 2 patches to the right wrist; Connect all the patches to the wrist-type terminal. The wrist-type terminal is powered on to emit low-voltage electricity, measures the resistivity ρ and dielectric constant ε between the patches, and obtains the wrist impedance.

[0018] In the embodiment of the present invention, the method for measuring pulse wave includes: The wrist-type terminal starts the photoelectric pulse wave sensor, emits a light signal to the skin, obtains the light signal reflected by the skin, and inputs it into the algorithm model to obtain the pulse wave.

[0019] In the embodiment of the present invention, the method for measuring blood oxygen value includes: The wrist-type terminal starts the photoelectric pulse wave sensor, emits a light signal to the skin, obtains the light intensity reflected by the skin, and inputs it into the algorithm model to obtain the blood oxygen value.

[0020] In the embodiment of the present invention, the method for measuring blood pressure value includes: The wrist-type terminal analyzes according to the amplitude and time of the pulse wave, calculates the systolic blood pressure and diastolic blood pressure of the user, and obtains the blood pressure value.

[0021] In the embodiment of the present invention, the method for measuring blood pressure value includes: The wrist-type terminal obtains the value of the gyroscope, judges the movement and posture of the user according to the value, and judges whether the user is in a sleeping state according to the movement and posture.

[0022] If it is judged as a non-sleeping state according to the body movement value, when it is judged as a sleeping state, the satiety degree judgment is not executed.

[0023] When in a non-sleep state, the degree of parasympathetic nerve excitation is calculated through heart rate variability and the change value of abdominal bioimpedance based on the abdominal blood perfusion index to determine whether the user is full.

[0024] The abdominal blood perfusion index is a calculated value obtained from the light volume signal for blood oxygen saturation monitoring and the diastolic and systolic of peripheral pulsatile blood vessels, reflecting the ability of peripheral blood perfusion; it is an index ranging from 0 to 100 that is transformed by mathematical methods through computer processing from the volume waveform collected by a pulse oximetry probe and is the ratio of the pulsatile signal to the non-pulsatile signal measured in a specific measurement area.

[0025] The calculation formula for the abdominal blood perfusion index (API) is: API = (maximum blood flow velocity of the abdominal aorta - minimum blood flow velocity of the abdominal aorta) / maximum blood flow velocity of the abdominal aorta × 100%. Among them, the maximum blood flow velocity of the abdominal aorta refers to the maximum blood flow velocity in the abdominal aorta during systole, and the minimum blood flow velocity of the abdominal aorta refers to the minimum blood flow velocity in the abdominal aorta during diastole.

[0026] The larger the value of API, the faster the systolic blood flow velocity of the abdominal artery and the more the blood flow, indicating that the blood perfusion of the intra-abdominal organs is more sufficient. If the value of API is smaller, it indicates that the systolic blood flow velocity of the abdominal artery slows down and the blood flow decreases, which may be caused by insufficient blood perfusion of the intra-abdominal organs.

[0027] Turn on the wrist terminal, and through the optical sensor at the bottom of the terminal, emit light sources with two different wavelengths of 660 and 940 nm to obtain the PPI value.

[0028] After a preliminary judgment of fullness, a pop-up window is shown to the user to confirm whether they are full.

[0029] Use the user's confirmation result as a label for the above parameter set {wrist blood perfusion, abdominal blood perfusion, abdominal impedance, heart rate variability, blood oxygen, blood pressure...}, and input the parameter set and the label of whether full into a deep learning model, which includes but is not limited to common deep learning models such as convolutional neural networks and principal component analysis.

[0030] In the embodiments of the present invention, A deep learning-based model is established to simultaneously process and predict multiple physiological parameters (such as wrist blood perfusion, abdominal blood perfusion, abdominal impedance, heart rate variability, blood oxygen, and blood pressure), including the following steps: Step 1: Data collection and preprocessing: Collect a large amount of multi-channel physiological signal data containing these physiological parameters.

[0031] Clean the data to remove outliers and noise.

[0032] Preprocess the data, such as normalization, interpolation, etc., to adapt to the input of the model.

[0033] Step 2: Feature extraction: Use traditional signal processing techniques to extract features from physiological signals. For example, extract heart rate variability features from electrocardiograms and blood oxygen saturation from blood oxygen signals. These features can include features in the frequency, time, spectrum, and time-frequency domains.

[0034] Step 3: Model architecture: Select or design a deep learning model, such as a convolutional neural network (CNN), a recurrent neural network (RNN), or a long short-term memory network (LSTM).

[0035] The model should be able to process multiple input channels (i.e., multiple physiological parameters) simultaneously and predict or classify the output.

[0036] Step 4: Training and optimization: Train the model using labeled data. Use cross-validation, early stopping, and other optimization techniques to prevent overfitting. Use an appropriate loss function (such as mean squared error, cross-entropy, etc.) to optimize the model.

[0037] Step 5: Evaluation and validation: Evaluate the performance of the model on an independent validation set. Use metrics such as accuracy, recall, F1-score, etc. to evaluate the prediction ability of the model.

[0038] Step 6: Model deployment and monitoring: Deploy the trained model to practical applications. Monitor the performance of the model and retrain or update the model if necessary.

[0039] Step 7: Interaction interface: Ask the user whether they are full, to what extent, etc., and verify with the data in the model after the results are input. Continuously improve and expand the model with the availability of more data and the progress of technology.

[0040] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for recognizing actual satiety based on physiological parameters, wearing a wearable intelligent health terminal, which includes a wrist-type terminal and several patches. Wear the wrist-type terminal on the wrist and wear multiple small patches on the abdomen, characterized in that It includes the following steps: Step 1: The intelligent health terminal monitors the abdominal impedance, pulse wave, blood oxygen value, blood pressure value and body movement value in real time; Step 2: At a time close to mealtime, the bio-impedance, pulse wave and blood pressure data of the wrist and abdomen are obtained respectively. The data is aggregated to the wrist terminal, and the data analysis is performed by the processor of the wrist terminal; Step 3: Judge the abdominal blood perfusion index according to the blood pressure data, calculate the degree of parasympathetic nerve excitement through heart rate variability, and the change value of abdominal bio-impedance to judge whether overeating occurs.

2. The actual satiety recognition method according to claim 1, wherein Judge as a non-sleep state according to the body movement value. When it is judged as a sleep state, the satiety judgment is not executed.

3. The actual satiety recognition method according to claim 1, characterized in that, After initially judging overeating, a pop-up window is shown to the user to confirm whether they are full, and the user's confirmation result is used as the label of the above parameter set.

4. The actual satiety recognition method according to claim 3, wherein The parameter set and the label of whether being full are input into the deep learning model, which includes a convolutional neural network and principal component analysis.

Images