Wearable physical sign monitoring device for newborns

Abstract

The invention relates to the technical field of intelligent medical treatment, and discloses a neonatal wearable physical sign monitoring device which comprises a plurality of collection points used for collecting neonatal physical sign signals, and the collection points comprise PPG sensor units which are arranged at different key positions of neonatal socks and have independent monitoring functions; the PPG sensor unit transmits collected photoelectric pulse signals to the processor module for processing, the processor module carries out fusion calculation on the signals of all collection points, and the final output of the oxyhemoglobin saturation and the heart rate of the newborn is obtained by integrating the data of all the collection points; the technical problems that an existing monitoring device for monitoring newborn physiological data is poor in wearing stability and poor in monitoring result accuracy are solved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical technology, and in particular to a wearable vital sign monitoring device for newborns. Background Art

[0002] In today's era of rapid technological advancement, the field of intelligent healthcare is undergoing unprecedented change. Artificial intelligence, as a core driving force, is profoundly transforming the way healthcare is monitored and diagnosed. As people's emphasis on health management continues to grow, wearable physiological data monitoring devices have emerged and achieved significant development, demonstrating significant maturity and widespread application value in the field of adult health monitoring.

[0003] Currently, monitoring technology based on photoplethysmography (PPG) is widely used in adult wearable devices. For example, smartwatches, with built-in PPG sensors, can continuously monitor a user's heart rate, blood oxygen saturation, and other vital physiological indicators in real time. These watches typically utilize advanced sensor technology and algorithms to accurately capture PPG signals from the wrist. Combined with artificial intelligence algorithms, these signals are analyzed and processed to provide users with detailed health reports and early warning information. Furthermore, finger-clip PPG monitoring devices are also common in medical settings. They can quickly and accurately measure parameters such as a patient's blood oxygen saturation, providing doctors with important diagnostic information.

[0004] Apnea is a common symptom in newborns. If left untreated, prolonged hypoxia can cause brain damage, significantly impacting the newborn's intellectual development and, in severe cases, even death. However, existing wearable physiological data monitoring technologies struggle to meet the unique physiological needs and monitoring requirements of newborns.

[0005] First, newborns have smaller bones and significantly different body structures than adults. Traditional adult wearable devices, such as watches and bracelets, are designed for size and wearing style based on adult physiology and are not directly suitable for newborns.

[0006] Secondly, newborns are active and move their bodies widely. Traditional monitoring devices are typically fixed and lack sufficient flexibility and stability. These devices can easily fall off or shift during newborn activity, resulting in interrupted or inaccurate monitoring signals and an inability to provide continuous and reliable physiological data to caregivers.

[0007] Furthermore, newborns are in a period of rapid growth and development, and their physiological characteristics and body dimensions undergo significant changes within a short period of time. Significant individual differences exist among newborns, including in weight, height, and body shape. Traditional monitoring methods are difficult to adapt to these dynamic physiological characteristics and individual differences of newborns, resulting in biased monitoring results that fail to accurately reflect the newborn's true physiological condition. Summary of the Invention

[0008] The present invention aims to provide a wearable vital sign monitoring device for newborns to solve the technical problems of poor wearing stability and poor accuracy of monitoring results of existing monitoring equipment for monitoring newborn physiological data.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: it includes multiple collection points for collecting newborn vital sign signals, and the collection points include PPG sensor units arranged at different key positions of the newborn socks and having independent monitoring functions; the PPG sensor units transmit the collected photoelectric pulse signals to the processor module for processing, and the processor module performs fusion calculations on the signals of each collection point, and obtains the final output of the newborn's blood oxygen saturation and heart rate by integrating the data of each collection point.

[0010] The principles and advantages of this solution are as follows: In practical application, this wearable neonatal vital sign monitoring device, based on photoplethysmography (PPG), collects vital sign signals from newborns by placing multiple independent PPG sensor units at strategic locations on a newborn's socks. These PPG sensors capture the photoelectric pulse signals beneath the skin of the newborn's feet, which contain physiological information such as the newborn's heart rate and blood oxygen saturation. The collected photoelectric pulse signals are transmitted to a processor module, which uses a specific algorithm to fuse and calculate the signals from each collection point. During the calculation process, the data from each collection point is comprehensively considered, and by analyzing the characteristics and interrelationships of the signals from different collection points, noise interference is removed and effective physiological information is extracted, ultimately providing accurate output of the newborn's blood oxygen saturation and heart rate. Compared with single-point monitoring, multi-point monitoring allows for mutual correction and verification, making the calculation results more stable and reliable, accurately reflecting the newborn's true physiological condition, and providing a more accurate basis for diagnosis.

[0011] Using a sock as a carrier, the sock's inherent elasticity allows it to fit snugly around the newborn's feet. Compared to traditional monitoring devices that are worn on the newborn's stomach or head, socks fit more closely to the newborn's skin and are less likely to fall off or shift due to the newborn's frequent movements in the incubator. Even with the newborn's large movements, the sock effectively secures the PPG sensor unit, ensuring the continuity and stability of the monitoring signal and providing reliable physiological data to the caregiver.

[0012] Multi-point monitoring can better adapt to the differences in body size and physiological characteristics of newborns, preventing the problem of a single collection point shifting and failing to collect data properly due to different alignment with the newborn's feet. Even if a single collection point is affected by changes in the position of the collection point due to different newborn body differences, the other collection points will still function normally, thus ensuring the accuracy of the monitoring results. This effectively solves the problem that traditional monitoring methods have difficulty adapting to the dynamic changes in the physiological characteristics and individual differences of newborns.

[0013] Traditional medical monitoring devices typically need to be worn on critical areas of the newborn's body, such as the abdomen and head, potentially interfering with other nursing procedures or home medical devices. This innovative device integrates the monitoring module into a sock, eliminating the need for additional equipment to be worn on a newborn's critical body part. This reduces interference with home care procedures while significantly improving ease of use and comfort for the newborn, making it particularly suitable for home monitoring scenarios.

[0014] Preferably, as an improvement, an airbag is provided at the interlayer position of the sock where the PPG sensor unit is installed, and each PPG sensor unit is located on the side of the corresponding airbag that is in close contact with the skin of the newborn; by controlling the inflation and deflation of the airbag, the airbag is expanded or contracted, so that the PPG sensor unit is in close contact with the skin of the newborn when the airbag is inflated, and the pressure on the skin of the newborn is reduced when the airbag is contracted.

[0015] The beneficial effect of this improvement is that by controlling the inflation and deflation of the airbag, the PPG sensor unit can be closely attached to the newborn's skin when the airbag is inflated, ensuring good contact between the sensor and the skin, effectively improving the quality of photoelectric pulse signal acquisition and, in turn, improving detection accuracy. When the airbag is deflated, it reduces pressure on the newborn's skin, preventing damage caused by prolonged pressure, improving the wearer's comfort, and also ensuring the health and safety of the newborn during monitoring.

[0016] Preferably, as an improvement, a collection point evaluation unit is provided. When the PPG sensor unit is worn for the first time, the collection point evaluation unit controls each PPG sensor unit and its corresponding airbag to work in sequence, and obtains the PPG signal collected by each PPG sensor unit respectively; the collection point evaluation unit extracts a variety of feature information from the PPG signal, including waveform features, statistical features, nonlinear features, dynamic template features and obvious interference features, and extracts multi-scale time series information, and inputs these feature information into the trained quality evaluation model to evaluate the quality of the PPG signal. The PPG signal with unqualified quality is eliminated and does not participate in the subsequent physiological information extraction and analysis.

[0017] The beneficial effect of this improvement is that when the PPG sensor unit is first worn, the collection point evaluation unit comprehensively evaluates the signals collected by each sensor. By extracting multiple feature information and multi-scale time series information and using a trained quality assessment model to make judgments, it can accurately identify and remove substandard PPG signals from subsequent physiological information extraction and analysis. This effectively prevents substandard signals from interfering with monitoring results, improves the accuracy and reliability of subsequent physiological information extraction and analysis, and provides more accurate diagnostic evidence for caregivers.

[0018] Preferably, as an improvement, the collection point evaluation unit sorts the PPG sensor units corresponding to the PPG signals whose quality is evaluated as qualified, and controls each PPG sensor unit and its corresponding airbag to work in sequence according to the sorting; each PPG sensor unit collects signals in sequence and transmits them to the processor module for processing.

[0019] The beneficial effect of this improvement is that it sorts the sensor units corresponding to PPG signals that have been assessed as qualified, and controls each sensor unit and its corresponding airbag to operate sequentially according to the order, thus preventing the prolonged operation of a single sensor unit in a single location from causing localized pressure on the newborn's skin. Each sensor unit sequentially collects signals and transmits them to the processor module for processing, making the monitoring process more scientific and reasonable. This ensures comprehensive and continuous monitoring, maximizes protection for the newborn's skin, improves wearing comfort, and helps extend the life of the monitoring device.

[0020] Preferably, as an improvement, a collection point correction unit is provided, which obtains the blood oxygen saturation and heart rate data of all qualified PPG sensor units calculated for the first time in the data processing unit, and assigns a weight to each qualified PPG sensor unit according to the quality level of each qualified PPG sensor unit; the blood oxygen saturation and heart rate data obtained by weighted averaging of each qualified PPG sensor unit are used as the preliminary output; based on the difference between the weighted average fused data and the actual data of each PPG sensor unit, a correction function is designed for each qualified PPG sensor unit, and in the continuous monitoring and collection of the PPG sensor unit, the correction function is optimized using the accumulated collection data; the PPG signal collected by each qualified PPG sensor unit is calculated by the data processing unit, and then corrected by the correction function to obtain the final output.

[0021] The beneficial effect of this improvement is that the acquisition point correction unit assigns weights based on the quality level of each qualified PPG sensor unit, uses the weighted average data as the initial output, and designs a correction function for each sensor unit based on the difference between the weighted average fused data and the actual data of each sensor unit. During continuous monitoring and acquisition, the correction function is continuously optimized using accumulated acquisition data, so that the PPG signal collected by each sensor unit is calculated by the data processing unit and then corrected by the correction function, resulting in a more accurate final output. This correction method fully accounts for the quality differences between different sensor units. Through dynamic adjustment and optimization, it effectively improves the accuracy of monitoring results and provides more reliable diagnostic information for caregivers.

[0022] Preferably, as an improvement, a neonatal assessment unit is provided, which judges the output blood oxygen saturation and heart rate data and presets the normal range of blood oxygen saturation and heart rate; when the blood oxygen saturation or heart rate exceeds the preset range, the neonatal assessment unit drives the intervention module to perform patting stimulation for rescue, and activates the photoelectric alarm to remind the guardian.

[0023] The beneficial effect of this improvement is that the neonatal assessment unit judges the output blood oxygen saturation and heart rate data, sets a preset normal range, and when the monitoring data exceeds the preset range, it can promptly activate the intervention module to perform a tapping stimulation for rescue and activate a photoelectric alarm to alert parents. This automated monitoring and intervention mechanism can immediately detect abnormalities in newborns and take appropriate measures, buying precious time for rescue, effectively improving the safety level of newborns and reducing the risk of brain damage and death caused by prolonged hypoxia and other causes.

[0024] Preferably, as an improvement, the intervention module includes a tapping unit and an alarm unit; The beating unit includes a housing, a silicone wall, an electromagnet, a striking hammer, a torsion spring, and a power drive circuit. The housing is disposed within the interlayer of a newborn's sock and is located in the area around the sole of the newborn's foot, and is used to house the electromagnet, striking hammer, torsion spring, and power drive circuit. The silicone wall is disposed on the contact surface between the housing and the sole of the newborn's foot. The torsion spring movably fixes the striking hammer, allowing it to rotate along the torsion direction of the torsion spring. The electromagnet is fixedly mounted within the housing and suspended relative to the striking surface of the hammer. The power drive circuit is disposed within the housing and is used to adjust the operating current of the electromagnet to control the strength with which the electromagnet attracts the hammer. The alarm unit triggers an audible and visual alarm when it detects an abnormality, notifying the guardian to take timely measures.

[0025] The beneficial effects of this improvement are as follows: the tapping unit adjusts the operating current of the electromagnet through the power drive circuit, accurately controlling the electromagnet's attraction to the hammer, thereby adjusting the magnitude of the striking force; and by adjusting the on and off time of the electromagnet, the striking frequency can be controlled. This adjustable design allows the tapping stimulation to be personalized according to the actual situation of the newborn, achieving effective rescue results while avoiding excessive harm to the newborn. When the alarm unit detects an abnormality, it triggers an audible and visual alarm, notifying the guardian to take timely measures, ensuring that parents are aware of the newborn's abnormal condition immediately, improving the timeliness and effectiveness of rescue.

[0026] Preferably, as an improvement, the waveform characteristics include amplitude standard deviation, baseline sum, rise time standard deviation and fall time standard deviation; statistical characteristics include kurtosis and skewness; nonlinear characteristics include Shannon entropy; dynamic template characteristics include calculating the distance between all peak signal segments in the sample and the template; obvious interference characteristics include pulse signal characteristics and random noise characteristics, the pulse signal characteristics are detected by setting the signal range, and the random noise characteristics are detected by calculating the coefficient of variation of the pulse signal.

[0027] The beneficial effects of this improvement are as follows: the amplitude standard deviation (A_std) reflects the fluctuation of the PPG signal amplitude, which helps to judge the stability of the signal; the baseline sum (D) can reflect the baseline level of the signal, and an unstable baseline may affect the accuracy of the signal; the rise time standard deviation (RT_std) and fall time standard deviation (DT_std) describe the degree of dispersion of the rise and fall times of the signal peaks and troughs. These features can reflect the waveform characteristics of the PPG signal from different angles and provide an important basis for signal quality assessment.

[0028] Kurtosis (K) measures the sharpness of the signal distribution, while skewness (S) reflects the degree of asymmetry. These two features provide insights into the signal's distribution characteristics and help identify anomalies. Shannon entropy (SE) assesses the amount of information in the PPG signal, reflecting its complexity and providing valuable insights into signal quality. By calculating the distance (T_D) between all peak signal segments in a sample and a template, it can describe variations in PPG signal quality, helping to promptly detect abnormal signal fluctuations. The impulse signal feature (IMPULSE) detects impulse interference within a defined signal range, while the random noise feature (CV) calculates the coefficient of variation of the pulse signal to assess the level of random noise. These features accurately identify significant interference within the PPG signal, providing more comprehensive information for signal quality assessment, thereby improving the accuracy and reliability of signal quality assessment and ensuring the accuracy of subsequent physiological information extraction and analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a structural block diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following is further described in detail through specific implementation methods: Example Basically as attached Figure 1 As shown, a wearable vital sign monitoring device for newborns, comprising: The sensor module includes a PPG sensor unit and an airbag unit.

[0031] The PPG sensor unit, or photoplethysmography (PPG), uses a standalone monitoring sensor that integrates a light source and a reflected light receiver. During operation, the light source illuminates the newborn's skin, measures the reflected PPG signal, and analyzes changes in the absorbed light to determine the newborn's blood oxygen and heart rate.

[0032] There are multiple independent monitoring sensors installed in the socks of the newborn. Their distribution locations include the toes, soles, heels and other key positions in the newborn socks to ensure all-round monitoring. According to the size of the newborn's feet and monitoring needs, 3-5 PPG sensor units can be configured to ensure coverage of key monitoring areas. Each independent monitoring sensor includes an LED light source and a photodetector for emitting and receiving light. The LED light source includes a red light source and an infrared light source. The wavelength of the red light source is 660nm, and the wavelength of the infrared light source is 940nm. The red light source and the infrared light source work continuously and alternately. Each independent monitoring sensor is connected to the wireless transmission module through a flexible circuit board to ensure the stability and reliability of data transmission. The flexible circuit board has good flexibility and bending resistance, and can adapt to the activities of the newborn's feet. The flexible circuit board adopts a standardized interface design with the sensor unit and the wireless transmission module to facilitate installation and maintenance The airbag unit is located within the sock's interlayer, where the individual monitoring sensors are mounted. Made of medical-grade silicone, the airbags offer excellent elasticity and biocompatibility, ensuring safety for newborns. The individual monitoring sensors are located on the skin-facing side of the airbags. When inflated, they maintain close contact with the newborn's skin, ensuring accurate measurements. Each airbag corresponding to an individual monitoring sensor is connected to an air supply pump via an air tube. The air supply pump utilizes a compact, low-noise micro pump.

[0033] The airbag is equipped with an electronically controlled valve, whose opening and closing is controlled by the processor module. When the valve is closed, gas supplied by the air pump is retained in the airbag, causing it to expand and place the independent monitoring sensor in close contact with the newborn's skin. When the valve is opened, the gas in the airbag is discharged to the outside world, reducing pressure on the newborn's skin.

[0034] The wireless transmission module is used to wirelessly transmit the photoelectric pulse signal collected by the independent monitoring sensor to the processor module using low-power Bluetooth (BLE) or other wireless communication technologies.

[0035] The processor module receives data from independent monitoring sensors and pre-processes the PPG signal through filtering to remove interference. It then calculates blood oxygen saturation based on the absorption ratio of red and infrared light by oxyhemoglobin and deoxyhemoglobin. It also determines heart rate by calculating the peak value of the PPG signal per unit time. The processor module includes a collection point evaluation unit, a data sorting unit, a collection point correction unit, and a neonatal assessment unit.

[0036] The collection point evaluation unit controls each independent monitoring sensor and its corresponding airbag to inflate and work in sequence when the PPG sensor unit is worn for the first time, obtains the PPG signals collected by each independent monitoring sensor, evaluates the quality of each PPG signal, eliminates unqualified PPG signals, sorts the independent monitoring sensors corresponding to qualified PPG signals, and controls each independent monitoring sensor and its corresponding airbag to work in sequence according to the sorting, so as to avoid long-term pressure on a single position of the newborn.

[0037] The quality score of the PPG signal by the acquisition point evaluation unit includes: Various feature information is extracted from PPG signals, including waveform features, statistical features, nonlinear features, dynamic template features, and significant interference features. Waveform features include amplitude standard deviation (A_std), baseline sum (D), rise time standard deviation (RT_std), and fall time standard deviation (DT_std). These features are obtained by calculating the peaks and troughs of the PPG signal and their corresponding time and amplitude variations. Statistical features include kurtosis (K) and skewness (S), which measure the concentration and asymmetry of the signal distribution. Nonlinear features include Shannon entropy (SE), which assesses the information content of the PPG signal. Dynamic template features describe the variation in PPG signal quality by calculating the distance (T_D) between all peak signal segments in a sample and the template. Significant interference features include impulse signal features (IMPULSE) and random noise features (CV). Impulse signal features are detected by setting a signal range, while random noise features are detected by calculating the coefficient of variation of the pulse signal.

[0038] To extract multi-scale temporal information from PPG signals, a multi-scale CNN network is used to extract local information at different scales. A BiLSTM network is then used to extract multi-scale temporal information. Five CNN modules with different convolution kernel sizes (32, 64, 128, 256, and 512) are used. Each module includes a 1D CNN network, 1D max pooling, batch normalization, and a Reluctant Unit (ReLU) activation function. Local information of the PPG signal is extracted using convolution kernels of different scales. A BiLSTM network is used to extract multi-scale temporal information from the PPG signal in both forward and backward passes. The BiLSTM network consists of input, forget, and output gates, which enable information transmission and discarding.

[0039] The multi-class features and multi-scale temporal information for each location are fused and input into the trained quality assessment model to obtain a PPG signal quality assessment result for each location. Based on the quality assessment results, the PPG signal quality at each location is assessed. If the PPG signal quality assessment result for a location is unqualified, the PPG signal at that location is discarded and not used in subsequent physiological information extraction and analysis.

[0040] This quality assessment model is trained using a large amount of PPG signal data. This training data assigns each data set a quality grade based on its quality, manually annotated by experts. Multi-class features and multi-scale temporal information from these data sets are extracted to identify relational data features. This model is trained using a random forest learning model, which learns the relationship between features in the training data and quality grade labels to predict quality labels for new data. Based on the distribution of features, the quality assessment model finds one or more decision boundaries to separate data of different quality levels. For new PPG signals, the model extracts features and then assigns them to a quality category based on these features and the decision boundaries. The quality assessment model outputs quality grades: Grade 1, Grade 2, Grade 3, and Unacceptable. If the quality assessment model outputs an Unacceptable quality grade, the blood oxygen and heart rate data calculated from the PPG signal are considered to have large errors and should be discarded.

[0041] The data processing unit obtains the PPG signal from the qualified independent monitoring sensor and calculates the blood oxygen saturation and heart rate data.

[0042] Blood oxygen saturation is calculated by determining the red-to-infrared absorption ratio (R) in the PPG signal. The red light absorption ratio (A) is the ratio of the AC component of the received reflected red light to the DC component of the reflected red light. The infrared light absorption ratio (B) is the ratio of the AC component of the received reflected infrared light to the DC component of the reflected infrared light. The R value is the ratio of the red light absorption ratio (A) to the infrared light absorption ratio (B). Blood oxygen saturation is calculated using the R value and pre-calibrated absorption curves for oxyhemoglobin and deoxyhemoglobin, either by lookup or empirical formula. Heart rate calculation involves detecting peaks in each PPG signal and determining the heart rate by counting the number of peaks per unit time.

[0043] The acquisition point correction unit obtains the blood oxygen saturation and heart rate data from all qualified independent monitoring sensors initially calculated by the data processing unit and assigns a weight to each qualified independent monitoring sensor based on its quality level. The weighted average of the blood oxygen saturation and heart rate data from each qualified independent monitoring sensor is used as the final output.

[0044] Based on the difference between the weighted average fusion data and the actual data from each independent monitoring sensor, a correction function is designed for each qualified independent monitoring sensor. This correction function can employ methods such as linear regression and polynomial fitting. Accumulated collected data is used to optimize the correction function, enabling adaptive parameter adjustment. The PPG signal collected by each qualified independent monitoring sensor is calculated by the data processing unit and then corrected using the correction function to obtain the final output.

[0045] The neonatal assessment unit judges the output blood oxygen saturation and heart rate data. When the blood oxygen saturation or heart rate exceeds the preset range, it drives the intervention module to perform tapping stimulation for rescue and activates the photoelectric alarm to remind the guardian.

[0046] intervention module, including a slapping unit and an alarm unit; The striking unit comprises a housing, a silicone wall, an electromagnet, a hammer, and a torsion spring. The torsion spring flexibly secures the hammer, while the electromagnet is suspended relative to the hammer's striking surface. The housing houses the electromagnet, hammer, and torsion spring. The housing is positioned within the interlayer of a newborn's sock, located in the area surrounding the newborn's sole. The silicone wall is located at the contact surface between the housing and the sole. The electromagnet is fixedly mounted within the housing, and the hammer can rotate in the direction of the torsion spring's torque. The housing also houses a power drive circuit that regulates the electromagnet's operating current to control the electromagnet's pull. The electromagnet's pull controls the hammer's striking force, thereby determining the striking force of the striking unit. As the electromagnet's suction increases, the hammer is attracted to the electromagnet, overcoming the torsion force and moving closer to the electromagnet. The greater the electromagnet's pull, the greater the angle at which the hammer rotates toward the electromagnet. When the electromagnet is powered off, the electromagnet's suction force disappears, and the hammer is quickly rotated and rebounded toward the silicone wall by the torsion force of the torsion spring, thereby tapping the soles of the newborn's feet.

[0047] The alarm unit will trigger an audible and visual alarm to notify the guardian when a serious abnormality is detected.

[0048] The working process of a wearable vital signs monitoring device for newborns is as follows: Put the socks with PPG sensor units and airbags on the newborn, adjust the position of the socks to ensure that each sensor unit is located in the appropriate key position of the newborn's foot and the airbag can function normally.

[0049] When the PPG sensor unit is first donned, the collection point evaluation unit activates, controlling each PPG sensor unit and its corresponding airbag to operate sequentially. Each sensor unit collects PPG signals when the airbag is inflated against the newborn's skin and stops collecting signals when the airbag is deflated. This completes the signal collection process for all sensor units.

[0050] The acquisition point evaluation unit extracts various features from each collected PPG signal. This extracted feature information is fed into a trained quality assessment model to evaluate the quality of each PPG signal. If a signal's quality assessment fails, the signal is discarded and excluded from subsequent physiological information extraction and analysis.

[0051] The processor module processes the signals collected by each qualified PPG sensor unit to preliminarily calculate blood oxygen saturation and heart rate data. The acquisition point correction unit takes this preliminarily calculated data and assigns a weight to each sensor unit based on its quality level. The weighted average of the blood oxygen saturation and heart rate data from each qualified PPG sensor unit is used as the preliminary output.

[0052] Based on the difference between the weighted average fused data and the actual data of each PPG sensor unit, a correction function is designed for each qualified PPG sensor unit.

[0053] The collection point evaluation unit sorts the PPG sensor units corresponding to the PPG signals whose quality is evaluated as qualified. The sorting basis can be factors such as the location of the sensor unit and the quality of the historical collection signals.

[0054] The PPG sensor units and their corresponding airbags are controlled to operate sequentially. Each sensor unit collects signals from the newborn's skin when the airbag expands, and reduces pressure on the newborn's skin when the airbag deflates. Each sensor unit collects signals sequentially and transmits them to the processor module for processing.

[0055] The PPG signals collected by each qualified PPG sensor unit are calculated by the data processing unit and then corrected using a correction function to obtain and output the final blood oxygen saturation and heart rate data. During the continuous monitoring and collection of the PPG sensor unit, the correction function is continuously optimized using the accumulated collected data.

[0056] The neonatal assessment unit evaluates the output of oxygen saturation and heart rate data, setting the normal range for these values. If oxygen saturation or heart rate exceeds these limits, it is considered abnormal. Once this is determined to be an abnormality, the neonatal assessment unit immediately activates the intervention module for emergency response and issues an alarm.

[0057] The above is only an embodiment of the present invention, and the common knowledge such as the specific technical solutions and / or characteristics in the solution are not described in detail here. It should be pointed out that for those skilled in the art, without departing from the technical solution of the present invention, several variations and improvements can be made, which should also be regarded as the scope of protection of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the description can be used to interpret the content of the claims.

Claims

1. A wearable vital sign monitoring device for newborns, characterized by: It includes multiple collection points for collecting newborn vital signs signals, and the collection points include PPG sensor units set at different key positions of the newborn socks and with independent monitoring functions; the PPG sensor units transmit the collected photoelectric pulse signals to the processor module for processing, and the processor module fuses and calculates the signals of each collection point, and obtains the final output of the newborn's blood oxygen saturation and heart rate by integrating the data of each collection point.

2. A wearable vital sign monitoring device for newborns according to claim 1, characterized in that: An airbag is set in the interlayer position of the sock where the PPG sensor unit is installed, and each PPG sensor unit is located on the side of the corresponding airbag that is in close contact with the skin of the newborn; by controlling the inflation and deflation of the airbag, the airbag is expanded or contracted, so that the PPG sensor unit is in close contact with the newborn's skin when the airbag is inflated, and the pressure on the newborn's skin is reduced when the airbag is contracted.

3. A wearable vital sign monitoring device for newborns according to claim 2, characterized in that: A collection point evaluation unit is set up. When the PPG sensor unit is worn for the first time, the collection point evaluation unit controls each PPG sensor unit and its corresponding airbag to work in sequence, and obtains the PPG signal collected by each PPG sensor unit respectively; the collection point evaluation unit extracts a variety of feature information from the PPG signal, including waveform features, statistical features, nonlinear features, dynamic template features and obvious interference features, and extracts multi-scale time series information. This feature information is input into the trained quality assessment model to evaluate the quality of the PPG signal. PPG signals with unqualified quality are eliminated and do not participate in the subsequent physiological information extraction and analysis.

4. A wearable vital sign monitoring device for newborns according to claim 3, characterized in that: The collection point evaluation unit sorts the PPG sensor units corresponding to the PPG signals whose quality is evaluated as qualified, and controls each PPG sensor unit and its corresponding airbag to work in sequence according to the sorting; each PPG sensor unit collects signals in sequence and transmits them to the processor module for processing.

5. The wearable vital sign monitoring device for newborns according to claim 4, characterized in that: Setting a collection point correction unit, wherein the collection point correction unit obtains the blood oxygen saturation and heart rate data of all qualified PPG sensor units calculated for the first time in the data processing unit, and assigns a weight to each qualified PPG sensor unit according to the quality level of each qualified PPG sensor unit; The blood oxygen saturation and heart rate data obtained by weighted averaging each qualified PPG sensor unit is used as the initial output. Based on the difference between the weighted average fused data and the actual data of each PPG sensor unit, a correction function is designed for each qualified PPG sensor unit. During the continuous monitoring and acquisition of the PPG sensor unit, the accumulated collected data is used to optimize the correction function. The PPG signals collected by each qualified PPG sensor unit are calculated by the data processing unit and then corrected by the correction function to obtain the final output.

6. The wearable vital sign monitoring device for newborns according to claim 5, characterized in that: A neonatal assessment unit is set up, which judges the output blood oxygen saturation and heart rate data and presets the normal range of blood oxygen saturation and heart rate; when the blood oxygen saturation or heart rate exceeds the preset range, the neonatal assessment unit drives the intervention module to perform tapping stimulation for rescue, and activates the photoelectric alarm to remind the guardian.

7. The wearable vital sign monitoring device for newborns according to claim 6, characterized in that: The intervention module includes a tapping unit and an alarm unit; The beating unit includes a housing, a silicone wall, an electromagnet, a striking hammer, a torsion spring, and a power drive circuit. The housing is disposed within the interlayer of a newborn's sock and is located in the area around the sole of the newborn's foot, and is used to house the electromagnet, striking hammer, torsion spring, and power drive circuit. The silicone wall is disposed on the contact surface between the housing and the sole of the newborn's foot. The torsion spring movably fixes the striking hammer, allowing it to rotate along the torsion direction of the torsion spring. The electromagnet is fixedly mounted within the housing and suspended relative to the striking surface of the hammer. The power drive circuit is disposed within the housing and is used to adjust the operating current of the electromagnet to control the strength with which the electromagnet attracts the hammer. The alarm unit triggers an audible and visual alarm when it detects an abnormality, notifying the guardian to take timely measures.

8. The wearable vital sign monitoring device for newborns according to claim 7, characterized in that: The waveform characteristics include amplitude standard deviation, baseline sum, rise time standard deviation and fall time standard deviation; the statistical characteristics include kurtosis and skewness; Nonlinear features include Shannon entropy; dynamic template features include calculating the distance between all peak signal segments in the sample and the template; obvious interference features include pulse signal features and random noise features. Pulse signal features are detected by setting the signal range, and random noise features are detected by calculating the coefficient of variation of the pulse signal.

Images