Clinical research wearable device monitoring system and method based on artificial intelligence

Abstract

The invention discloses a clinical research wearable device monitoring system and method based on artificial intelligence, and relates to the technical field of signal monitoring. Firstly, a pressure value of a contact surface between a wearable device and skin is collected in real time through a pressure monitoring module; drawing a two-dimensional pressure fluctuation curve and automatically identifying an abnormal pressure interval and a non-pressure interval; the signal compensation module acquires a corresponding PPG digital signal in real time, and compensates the signal in the abnormal pressure interval by combining the pressure fluctuation curve; the non-voltage region prediction module performs intelligent filling on signals in a non-voltage region by using historical PPG data; the displacement detection module is responsible for judging the moving direction and distance of the equipment; finally, the digital signal adjusting module determines a new detection area according to the displacement information, calculates the blood volume ratio of the original area to the new area by means of a preset blood volume model, and corrects PPG digital signals collected in the new area according to the blood volume ratio, so that the continuity and accuracy of a measurement result are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of signal monitoring, and specifically to a monitoring system and method for clinical research wearable devices based on artificial intelligence. Background Art

[0002] The existing clinical-level wearable monitoring technology based on artificial intelligence mainly focuses on: removing noise through traditional signal processing and motion artifact recognition algorithms such as filtering, high / low pass, and band stop; on the other hand, using machine learning or deep learning models to evaluate the quality of PPG waveforms, score abnormal segments, and compensate for missing signals using methods such as prediction or autoencoders.

[0003] In addition, with the rapid development of wearable devices and mobile health monitoring technology, photoplethysmography (PPG) has been widely used in the monitoring of physiological parameters such as real-time heart rate and blood oxygen saturation due to its non-invasive, low-power, and portability. However, in actual clinical and daily exercise scenarios, existing PPG-based wearable monitoring systems are often severely interfered by factors such as unstable wearing pressure, slight device displacement, and short-term detachment from skin contact, resulting in signal distortion, tomography, and pseudo-pulse phenomena, thereby affecting the continuous and accurate acquisition and analysis of long-term dynamic physiological data. At the same time, it is difficult to achieve an integrated closed-loop from pressure monitoring to pressure-free prediction, displacement detection, and cross-region signal calibration. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a monitoring system and method for clinical research wearable devices based on artificial intelligence, which solves the problems in the background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A monitoring system for clinical research wearable devices based on artificial intelligence, comprising:

[0006] A pressure monitoring module, configured to monitor the pressure value between the wearable device and the original detection area, establish a two-dimensional pressure value fluctuation curve graph, and determine the abnormal pressure interval and the pressure-free interval according to the pressure value fluctuation curve;

[0007] A signal compensation module, configured to obtain the generated PPG digital signal in real time, establish a two-dimensional PPG signal curve graph, and combine it with the two-dimensional pressure value fluctuation curve graph to obtain the PPG digital signal in the abnormal pressure interval for signal compensation;

[0008] A pressure-free area prediction module, configured to fill the PPG digital signal in the pressure-free interval according to the historical PPG digital signal in the pressure-free interval;

[0009] A displacement detection module, configured to determine the displacement direction and displacement distance of the wearable device;

[0010] A digital signal adjustment module is used to determine a new detection area according to the displacement direction and displacement distance of the wearable device, determine the blood volume of the new detection area according to a preset model, calculate the ratio coefficient of the blood volume of the original detection area to the blood volume of the new detection area, and adjust the PPG digital signal generated in the new detection area according to the ratio coefficient.

[0011] As a further solution of the present invention: in the pressure monitoring module, the specific content of monitoring the pressure value between the monitoring wearable device and the original detection area, establishing a two-dimensional pressure value fluctuation curve graph, and determining the abnormal pressure interval and the non-pressure interval according to the pressure value fluctuation curve is as follows:

[0012] When establishing the two-dimensional pressure value fluctuation curve graph, the abscissa represents the time line, and the ordinate represents the pressure value;

[0013] Compare the pressure value with a preset value G1 to determine the abnormal pressure interval:

[0014] If the pressure value exceeds the preset value G1, determine the time point T1 until the pressure value is less than or equal to the preset value G1, and determine the time point T2. The time interval between the time point T1 and the time point T2 is determined as the abnormal pressure interval; no other processing is performed in other cases;

[0015] Obtain the time interval when the pressure value is 0 and determine it as the non-pressure interval.

[0016] As a further solution of the present invention: in the signal compensation module, the specific method of obtaining the PPG digital signal in the abnormal pressure interval for signal compensation is as follows:

[0017] Determine the pressure difference when each PPG digital signal is generated in the abnormal pressure interval; where the pressure difference is expressed as the difference between the actual pressure value at this moment and the preset value G1;

[0018] According to the pressure difference, perform signal compensation on each PPG digital signal through a preset weighted formula.

[0019] As a further solution of the present invention: in the non-pressure area prediction module, the specific method of predicting and filling the PPG digital signal in the non-pressure interval according to the historical PPG digital signal is as follows:

[0020] Determine the specific time period of the non-pressure interval as the periodic time period;

[0021] Intercept the n periodic time periods before the start of the non-pressure interval, where n is a preset value;

[0022] Obtain the first PPG digital signal in n periodic time periods, and calculate its average value, denoted as Bp; then calculate the absolute value of the average value Bp and the first PPG digital signal in the i-th periodic time period; and denote it as Di, where 1 ≤ i ≤ n;

[0023] Obtain the absolute values Di of n first PPG digital signals, and calculate their average value to obtain the floating value Fp;

[0024] At the same time, calculate the difference between the average value Bp and the first PPG digital signal in the i-th periodic time period, and denote it as Ci. Establish a trend curve based on the difference Ci, where the abscissa represents the i-th difference and the ordinate represents the specific difference;

[0025] According to the trend curve, obtain the n-th difference coordinate point and the (n - 1)-th difference coordinate point, connect the two coordinate points and calculate their slope k, and determine the first PPG digital signal in the non-pressure interval according to the slope.

[0026] As a further solution of the present invention: the specific method for determining the first PPG digital signal in the non-pressure interval according to the slope is:

[0027] If the slope k is positive, determine the first PPG digital signal X1 in the non-pressure interval. At this time: X = Bp + k × Fp;

[0028] If the slope k is negative, determine the first PPG digital signal X1 in the non-pressure interval. At this time: X = Bp - k × Fp;

[0029] If the slope is 0, determine the first PPG digital signal in the non-pressure interval as X1: X = Bp.

[0030] As a further solution of the present invention: it further includes:

[0031] After determining the first PPG digital signal in the non-pressure interval, obtain the second PPG digital signal in n periodic time periods, repeat the same process of determining the first PPG digital signal in the non-pressure interval, and determine the second PPG digital signal in the non-pressure interval, and so on, until the last PPG digital signal in the non-pressure interval is determined.

[0032] As a further solution of the present invention: in the displacement detection module, the specific method for determining the displacement direction and displacement distance of the wearable device is:

[0033] Embed two slender flexible resistive strain sensing bands in the wearable device;

[0034] Obtain the tensile forces of the two sensing bands in real time, and denote them as Fr and Fl respectively;

[0035] Determine the displacement direction according to the changes in the pulling forces Fr and Fl and the different wearing positions:

[0036] Then, according to the change values ΔFr and ΔFl of the pulling forces Fr and Fl, combined with Hooke's law, determine the displacement distance x of the wearable device through the following formula:

[0037]

[0038] where K is the elastic coefficient of the sensing belt.

[0039] As a further solution of the present invention: in the digital signal adjustment module, the specific method of calculating the ratio coefficient of the blood volume in the original detection area and the blood volume in the new detection area and adjusting the PPG digital signal generated in the new detection area according to the ratio coefficient is as follows:

[0040] Obtain the ratio coefficient and the PPG digital signal generated in the new detection area, multiply the two to obtain the adjusted PPG digital signal, and record it in the two-dimensional PPG signal curve graph.

[0041] A monitoring method for a wearable device in clinical research based on artificial intelligence includes:

[0042] Step 1: Monitor the pressure value between the wearable device and the original detection area, establish a two-dimensional pressure value fluctuation curve graph, and determine the abnormal pressure interval and the pressure-free interval according to the pressure value fluctuation curve;

[0043] Step 2: Real-time obtain the generated PPG digital signal, establish a two-dimensional PPG signal curve graph, and combine it with the two-dimensional pressure value fluctuation curve graph to obtain the PPG digital signal in the abnormal pressure interval for signal compensation;

[0044] Step 3: Fill the PPG digital signal in the pressure-free interval according to the historical PPG digital signal in the pressure-free interval;

[0045] Step 4: Determine the displacement direction and displacement distance of the wearable device;

[0046] Step 5: Determine the new detection area according to the displacement direction and displacement distance of the wearable device, determine the blood volume size of the new detection area according to the preset model, calculate the ratio coefficient of the blood volume in the original detection area and the blood volume in the new detection area, and adjust the PPG digital signal generated in the new detection area according to the ratio coefficient.

[0047] The present invention provides a monitoring system and method for a wearable device in clinical research based on artificial intelligence. Compared with the prior art, it has the following beneficial effects:

[0048] The present invention realizes the high reliability and continuity guarantee of wearable PPG signals. The pressure monitoring and signal compensation module can identify and correct in real time the blood volume change artifacts caused by excessive wearing pressure, while the pressure-free area prediction module, when the device becomes loose or has poor contact, intelligently fills in the missing waveforms based on historical signals to ensure that the monitoring curve is complete without breaks.

[0049] Meanwhile, the displacement detection and digital signal adjustment module accurately captures the minute movements of the device during wearing, and relocates the detection area and dynamically corrects the blood volume model accordingly, ensuring the comparability and accuracy of signal amplitudes and physiological parameters under different wearing states. Overall, the invention provides an end-to-end, anti-interference and adaptive technical solution for long-term dynamic physiological monitoring at the clinical research level. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The present invention will be further described below in conjunction with the accompanying drawings.

[0051] Figure 1 is the structural framework diagram of a clinical research wearable device monitoring system based on artificial intelligence according to the present invention;

[0052] Figure 2 is the step flow chart of a clinical research wearable device monitoring method based on artificial intelligence according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0054] Embodiment 1

[0055] Please refer to Figure 1 , the present invention provides a clinical research wearable device monitoring system based on artificial intelligence, including;

[0056] A pressure monitoring module, configured to monitor the pressure value between the wearable device and the detection area, establish a two-dimensional pressure value fluctuation curve graph, and determine an abnormal pressure interval and a pressure-free interval according to the pressure value fluctuation curve;

[0057] It should be noted that the original detection area is defined as the area that comes into long-term contact with the skin by the wearable device after wearing is completed when wearing the wearable device; according to the principle of Photoplethysmography (PPG): usually, a green light LED (sometimes infrared light or near-infrared light) and a photodiode are integrated on the back of the device; the light emitted by the LED irradiates the capillaries under the skin in the detection area, and as the blood flows in the capillaries with the heartbeat, the periodic change in blood volume will cause a slight change in the intensity of the reflected or transmitted light; after the photodiode captures these changes in light intensity, it is converted into an electrical signal, and through filtering and peak detection algorithms, the number of heartbeats per minute (i.e., heart rate) can be obtained.

[0058] Monitoring the pressure value between the monitoring wearable device and the original detection area, establishing a two-dimensional pressure value fluctuation curve graph, and determining the specific content of the abnormal pressure interval and the pressure-free interval according to the pressure value fluctuation curve are as follows:

[0059] When establishing the two-dimensional pressure value fluctuation curve graph, the abscissa represents the time line, and the ordinate represents the pressure value.

[0060] Comparing the pressure value with the preset value G1 to determine the abnormal pressure interval:

[0061] If the pressure value exceeds the preset value G1, determine the time point T1 until the pressure value is less than or equal to the preset value G1, and then determine the time point T2. The time interval between the time point T1 and the time point T2 is determined as the abnormal pressure interval; no other processing is done in other cases.

[0062] Obtain the time interval when the pressure value is 0 and determine it as the pressure-free interval. If the pressure value is always 0, it means that the wearable device is being worn, and this situation is excluded.

[0063] Real-time monitoring of the pressure value in the area where the wearable device contacts the skin, establishing a pressure fluctuation curve with time as the abscissa and pressure value as the ordinate, can accurately identify the periods of excessive pressure (abnormal pressure interval) or zero pressure (pressure-free interval); this not only can timely detect measurement distortion caused by improper wearing tightness or position, but also provides a key interval division basis for subsequent signal processing, ensuring the reliability of data from the source.

[0064] The signal compensation module obtains the generated PPG digital signal in real time, establishes a two-dimensional PPG signal curve graph, and combines it with the two-dimensional pressure value fluctuation curve graph to obtain the PPG digital signal in the abnormal pressure interval for signal compensation.

[0065] The specific method of obtaining the PPG digital signal in the abnormal pressure interval for signal compensation is as follows:

[0066] Determine the pressure difference when generating each PPG digital signal in the abnormal pressure range; wherein, the pressure difference is expressed as the difference between the actual pressure value at this moment and the preset value G1;

[0067] According to the pressure difference, perform signal compensation on each PPG digital signal through a preset weighting formula;

[0068] Specifically, if the pressure difference is Hr when a certain PPG digital signal is generated, and this PPG digital signal is Sv, then determine the PPG digital signal after compensation for this PPG digital signal at this moment through the preset weighting formula Dv = Sv + αHr; wherein, α is the weighting coefficient, which is specifically determined by professional staff;

[0069] It should be noted that when the pressure is too high, due to the effect of the pressure, the blood volume in the detection area will decrease. The photodiode in the wearable device will read the reflected light intensity every millisecond (for example, 500 times per second) and turn it into a long string of digital points: 0.12, 0.15, 0.13, …; when the wearable device applies too much pressure to the skin at the detection position, the underlying capillaries and veins will be squeezed, and the blood will be "pushed" to adjacent areas or deeper tissues, and the light reflected back from the detection area may be significantly reduced; this "blood volume trough" caused by mechanical squeezing is manifested as a trough lower than the normal pulse peak in the PPG signal, and even a pseudo-pulse with a very small "peak" amplitude may appear;

[0070] After detecting the abnormal pressure range, perform real-time compensation on the corresponding PPG digital signal through a preset weighting formula according to the difference between the actual pressure value at this moment and the preset threshold G1; it can effectively correct the signal deviation caused by the reduction of capillary blood volume due to excessive pressure - that is, the "blood volume trough" or pseudo-pulse, thereby improving the measurement accuracy and stability of physiological parameters such as heart rate;

[0071] A pressure-free area prediction module, used to fill the PPG digital signal in the pressure-free range according to the historical PPG digital signal in the pressure-free range;

[0072] The specific method of predicting and filling the PPG digital signal in the pressure-free range according to the historical PPG digital signal in the pressure-free range is as follows:

[0073] Determine the specific time period of the pressure-free range, and use this time period as the periodic time period;

[0074] Intercept the first n periodic time periods before the start of the pressure-free range, where n is a preset value, which is specifically determined by professional staff;

[0075] Obtain the first PPG digital signal in n cycle periods, calculate its average value, denoted as Bp; then calculate the absolute value of the average value Bp and the first PPG digital signal in the i-th cycle period; and denote it as Di, where 1 ≤ i ≤ n;

[0076] Obtain the absolute values Di of n first PPG digital signals, calculate their average value, and obtain the floating value Fp;

[0077] At the same time, calculate the difference between the average value Bp and the first PPG digital signal in the i-th cycle period, and denote it as Ci. Establish a trend curve based on the difference Ci, where the abscissa represents the i-th difference and the ordinate represents the specific difference;

[0078] According to the trend curve, obtain the n-th difference coordinate point and the (n - 1)-th difference coordinate point, connect the two coordinate points and calculate their slope k, and determine the first PPG digital signal in the non-pressure interval according to the slope:

[0079] If the slope k is positive, determine the first PPG digital signal X1 in the non-pressure interval. At this time: X = Bp + k × Fp;

[0080] If the slope k is negative, determine the first PPG digital signal X1 in the non-pressure interval. At this time: X = Bp - k × Fp;

[0081] If the slope is 0, determine the first PPG digital signal in the non-pressure interval as X1: X = Bp;

[0082] After determining the first PPG digital signal in the non-pressure interval, obtain the second PPG digital signal in n cycle periods, repeat the above process, determine the second PPG digital signal in the non-pressure interval, and so on until the last PPG digital signal in the non-pressure interval is determined;

[0083] For the non-pressure interval where the pressure value remains zero, by intercepting the historical PPG signals of the first N cycles in this interval, apply the prediction algorithm based on the average value, floating value, and trend slope to each sampling point in turn to fill in and infer the missing PPG digital signals. This not only ensures the continuity of the signal curve but also avoids data breaks caused by brief loosening or loose wearing, improving the overall monitoring integrity;

[0084] A displacement detection module for determining the displacement direction and displacement distance of the wearable device;

[0085] The specific method for determining the displacement direction and displacement distance of the wearable device is as follows:

[0086] Embed two slender flexible resistive strain sensing bands in the wearable device;

[0087] Obtain the tensile forces of two sensing bands in real time, and denote them as Fr and Fl respectively;

[0088] Determine the displacement direction according to the changes in the tensile forces Fr and Fl and the different wearing positions:

[0089] When the wearable device is a watch: when both the tensile force Fr and the tensile force Fl decrease, the displacement direction of the wearable device is to the right (i.e., the palm direction); when both the tensile force Fr and the tensile force Fl increase, the displacement direction of the wearable device is to the left (i.e., the joint direction);

[0090] Then, according to the change values of the tensile forces Fr and Fl, combined with Hooke's law, determine the displacement distance of the wearable device;

[0091] Specifically, obtain the change values of the tensile forces Fr and Fl as ΔFr and ΔFl, and determine the displacement distance x of the wearable device through the following formula:

[0092]

[0093] where K is the elastic coefficient of the sensing band;

[0094] Arrange two flexible resistive strain sensing bands inside the device to obtain the tensile force changes on the left and right sides in real time, and combine the wearing position information to accurately determine the displacement direction and distance of the device; convert the tensile force change into a displacement amount through Hooke's law, and this module provides accurate spatial position information for the subsequent positioning and signal adjustment of the new detection area;

[0095] The digital signal adjustment module is used to determine the new detection area according to the displacement direction and displacement distance of the wearable device, determine the blood volume size of the new detection area according to the preset model, calculate the ratio coefficient of the blood volume of the original detection area to the blood volume of the new detection area, and adjust the PPG digital signal generated in the new detection area according to the ratio coefficient;

[0096] It should be noted that the blood volume preset model takes the displacement direction and displacement distance of the wearable device as the core input (auxiliary features such as skin pressure and temperature can be additionally combined), and through physical optical calculation (obtaining the change in optical path and absorption coefficient in the new geometry according to diffusion approximation / Monte Carlo simulation) or data-driven regression (directly mapping the displacement features to the blood volume using a trained regression tree / neural network), after fitting the individual parameters in the initialization calibration stage, the blood volume size of the new detection area can be directly output by reading the displacement vector in real time. The blood volume preset model is specifically implemented through existing technologies and will not be elaborated here;

[0097] The specific method of calculating the ratio coefficient of the blood volume of the original detection area to the blood volume of the new detection area and adjusting the PPG digital signal generated in the new detection area according to the ratio coefficient is as follows;

[0098] Obtain the PPG digital signal generated by the ratio coefficient and the new detection area, multiply the two to obtain the adjusted PPG digital signal, and record it in the two-dimensional PPG signal curve graph;

[0099] Based on the displacement detection result, call the blood volume preset model to calculate the blood volume ratio coefficient between the original detection area and the new detection area; multiply this coefficient by the PPG digital signal collected in the new area to correct the signal amplitude difference caused by the change of the detection area, and achieve the consistency and comparability of cross-area data; this module ensures that the monitoring data can still maintain high precision after the device is moved or re-worn.

[0100] Embodiment 2

[0101] Refer to Figure 2 , in the specific implementation process of this embodiment, on the basis of Embodiment 1, and the difference from Embodiment 1 is that this embodiment also provides a monitoring method for a clinical research wearable device based on artificial intelligence:

[0102] Step 1: Monitor the pressure value between the wearable device and the original detection area, establish a two-dimensional pressure value fluctuation curve graph, and determine the abnormal pressure interval and the pressure-free interval according to the pressure value fluctuation curve;

[0103] Step 2: Real-time obtain the generated PPG digital signal, establish a two-dimensional PPG signal curve graph, and combine it with the two-dimensional pressure value fluctuation curve graph to obtain the PPG digital signal in the abnormal pressure interval for signal compensation;

[0104] Step 3: Fill the PPG digital signal in the pressure-free interval according to the historical PPG digital signal in the pressure-free interval;

[0105] Step 4: Determine the displacement direction and displacement distance of the wearable device;

[0106] Step 5: Determine the new detection area according to the displacement direction and displacement distance of the wearable device, determine the blood volume size of the new detection area according to the preset model, calculate the ratio coefficient of the blood volume of the original detection area to the blood volume of the new detection area, and adjust the PPG digital signal generated in the new detection area according to the ratio coefficient.

[0107] Embodiment 3

[0108] In the specific implementation process of this embodiment, it includes all the implementation processes of the above two groups of embodiments.

[0109] Some of the data in the above formula are numerically calculated after removing their dimensions, and the content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0110] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A wearable device monitoring system for clinical research based on artificial intelligence, characterized in that: include: The pressure monitoring module is used to monitor the pressure value of the wearable device and the original detection area, establish a two-dimensional pressure value fluctuation curve, and determine the abnormal pressure interval and the pressure-free interval according to the pressure value fluctuation curve; The signal compensation module obtains the generated PPG digital signal in real time, establishes a two-dimensional PPG signal curve graph, and combines the two-dimensional pressure value fluctuation curve graph to obtain the PPG digital signal in the abnormal pressure range for signal compensation; The unpressurized area prediction module is used to fill the PPG digital signal of the unpressurized area according to the historical PPG digital signal in the unpressurized area; A displacement detection module is used to determine the displacement direction and distance of the wearable device; The digital signal adjustment module is used to determine the new detection area according to the displacement direction and displacement distance of the wearable device, determine the blood volume of the new detection area according to the preset model, calculate the ratio coefficient of the blood volume of the original detection area to the blood volume of the new detection area, and adjust the PPG digital signal generated in the new detection area according to the ratio coefficient.

2. According to claim 1, a wearable device monitoring system for clinical research based on artificial intelligence is characterized in that: In the pressure monitoring module, the pressure values ​​of the monitoring wearable device and the original detection area are established to establish a two-dimensional pressure value fluctuation curve. According to the pressure value fluctuation curve, the specific contents of the abnormal pressure interval and the pressure-free interval are determined as follows: When establishing a two-dimensional pressure value fluctuation curve graph, the horizontal axis represents the time line and the vertical axis represents the pressure value; Compare the pressure value with the preset value G1 to determine the abnormal pressure range: If the pressure value exceeds the preset value G1, the time point T1 is determined, and when the pressure value is less than or equal to the preset value G1, the time point T2 is determined, and the time interval between the time point T1 and the time point T2 is determined as the abnormal pressure interval; no processing is performed in other cases; The time interval when the pressure value is 0 is obtained and determined as the pressure-free interval.

3. The artificial intelligence-based clinical research wearable device monitoring system according to claim 2, characterized in that: The signal compensation module obtains the PPG digital signal in the abnormal pressure interval for signal compensation in a specific manner as follows: Determine the pressure difference when each PPG digital signal is generated in the abnormal pressure interval; wherein the pressure difference is represented by the difference between the actual pressure value at that moment and the preset value G1; According to the pressure difference, each PPG digital signal is compensated by a preset weighting formula.

4. The artificial intelligence-based clinical research wearable device monitoring system according to claim 1, characterized in that: In the pressure-free area prediction module, the specific method of predicting the PPG digital signal filling the pressure-free interval in the pressure-free interval according to the historical PPG digital signal is: Determine the specific time period of the pressure-free interval and take this time period as the cycle period; Intercept the n period before the start of the pressure-free interval, where n is a preset value; Obtain the first PPG digital signal in the n cycle period and calculate its average value, recorded as Bp; then calculate the average value Bp and the absolute value of the first PPG digital signal in the i-th cycle period; and record it as Di, where 1≤i≤n; Obtain the absolute value Di of the n first PPG digital signals and calculate their average value to obtain the floating value Fp; At the same time, the difference between the average value Bp and the first PPG digital signal in the i-th cycle period is calculated and recorded as Ci, and a trend curve is established according to the difference Ci, where the abscissa represents the i-th difference and the ordinate represents the specific difference; According to the trend curve, the nth difference coordinate point and the n-1th difference coordinate point are obtained, and the slope k is calculated by connecting the two coordinate points. The first PPG digital signal in the pressure-free interval is determined based on the slope.

5. The artificial intelligence-based clinical research wearable device monitoring system according to claim 4, characterized in that: The specific method of determining the first PPG digital signal in the pressure-free interval according to the slope is: If the slope k is a positive number, determine the first PPG digital signal X1 in the pressure-free interval, at this time: X = Bp + k × Fp; If the slope k is a negative number, determine the first PPG digital signal X1 in the pressure-free interval, at this time: X = Bp-k×Fp; If the slope is 0, the first PPG digital signal in the pressure-free interval is determined to be X1: X=Bp.

6. The artificial intelligence-based clinical research wearable device monitoring system according to claim 5, characterized in that: Also includes: After determining the first PPG digital signal in the no-pressure interval, obtain the second PPG digital signal in the n-cycle period, repeat the same process of determining the first PPG digital signal in the no-pressure interval, determine the second PPG digital signal in the no-pressure interval, and so on, until the last PPG digital signal in the no-pressure interval is determined.

7. The artificial intelligence-based clinical research wearable device monitoring system according to claim 1, characterized in that: In the displacement detection module, the specific method of determining the displacement direction and displacement distance of the wearable device is: Two thin and flexible resistive strain sensing strips are embedded in the wearable device; The tension of the two sensing belts is obtained in real time and recorded as Fr and Fl respectively; Determine the displacement direction according to the changes in the tension Fr, Fl and the different wearing positions: Then, according to the change values ​​ΔFr and ΔFl of the pulling forces Fr and Fl, combined with Hooke's law, the displacement distance x of the wearable device is determined by the following formula: Where K is the elastic coefficient of the sensor belt.

8. The artificial intelligence-based clinical research wearable device monitoring system according to claim 1, characterized in that: In the digital signal adjustment module, the specific method of calculating the ratio coefficient of the blood volume in the original detection area and the blood volume in the new detection area, and adjusting the PPG digital signal generated in the new detection area according to the ratio coefficient is as follows; The ratio coefficient and the PPG digital signal generated by the new detection area are obtained, and the adjusted PPG digital signal is obtained by multiplying the two, and recorded in a two-dimensional PPG signal curve graph.

9. A clinical research wearable device monitoring method based on artificial intelligence, applied to a clinical research wearable device monitoring system based on artificial intelligence according to any one of claims 1 to 8, characterized in that: include: Step 1: Monitor the pressure values ​​of the wearable device and the original detection area, establish a two-dimensional pressure value fluctuation curve, and determine the abnormal pressure interval and the pressure-free interval according to the pressure value fluctuation curve; Step 2: Obtain the generated PPG digital signal in real time, and establish a two-dimensional PPG signal curve graph, combined with the two-dimensional pressure value fluctuation curve graph, obtain the PPG digital signal in the abnormal pressure range for signal compensation; Step 3: Fill the PPG digital signal of the pressure-free interval in the pressure-free interval according to the historical PPG digital signal; Step 4: Determine the displacement direction and distance of the wearable device; Step 5: Determine the new detection area according to the displacement direction and displacement distance of the wearable device, determine the blood volume of the new detection area according to the preset model, calculate the ratio coefficient of the blood volume in the original detection area to the blood volume in the new detection area, and adjust the PPG digital signal generated in the new detection area according to the ratio coefficient.

Images