Non-inductive monitoring method for physical signs of bedridden patient based on multi-modal sensor fusion
Through multimodal sensor fusion monitoring of patients' breathing, heart rate, temperature and body posture, the inconvenience and insufficient monitoring of traditional contact equipment are solved, sensingless monitoring and multi-band alarm are achieved, data accuracy and monitoring effect are improved.
Patent Information
- Application Number
- CN202510587372.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-08
- Publication Date
- 2025-08-15
AI Technical Summary
In the long-term use of traditional contact electrocardiogram monitoring equipment, there are problems such as inconvenience, single monitoring parameters and insufficient data accuracy, making it difficult to achieve invisible monitoring and timely alarms, and cannot fully and accurately reflect the physical condition of bedridden patients.
Multimodal sensor fusion method is adopted, including millimeter wave radar, infrared thermal imaging and pressure sensors, to monitor the patient's breathing, heart rate, temperature and body posture information respectively, and through data preprocessing and analysis, a multi-band alarm method is set up to achieve insensitive monitoring and accurate alarm.
It realizes all-round vital sign monitoring without the need for active cooperation of patients, improves data accuracy and timeliness of alarms, reduces patient discomfort, and enhances monitoring effect and compliance.
Smart Images
Figure CN120477734A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of vital sign monitoring, and in particular to a method for sensing vital signs of bedridden patients based on multimodal sensor fusion. Background Art
[0002] In the field of modern medical health monitoring, traditional methods for collecting and monitoring human ECG data usually rely on traditional contact devices, such as ECG monitors. These devices require electrodes to be directly attached to the human skin, which brings many inconveniences to users, such as skin discomfort and restrictions on the user's freedom of movement. They cannot provide continuous real-time ECG monitoring for many years. The limitations of traditional contact ECG monitoring devices are particularly obvious for patients with cardiovascular and cerebrovascular diseases who require long-term continuous monitoring.
[0003] As the current medical system becomes more and more comprehensive, the demand for health monitoring of bedridden patients is increasing. Whether it is heart disease or sleep, real-time monitoring is required to monitor the patient's physical condition, so that medical staff can treat the patient or judge the patient's physical condition in a timely manner. Due to the limitations of the monitoring principle, traditional monitoring equipment has problems such as single monitoring parameters, insufficient data accuracy and reliability, and it is difficult to fully and accurately reflect the patient's physical condition. One of the main purposes of monitoring is to enable medical staff to fully and real-time connect with the patient's physical changes, and then adapt to the patient's quality. However, the traditional monitoring methods and alarm methods are relatively simple and cannot remind medical staff of the patient's accurate condition in time. Therefore, there is an urgent need for a monitoring method that can achieve non-sensing monitoring and at the same time can comprehensively and accurately obtain the patient's vital signs information and accurate alarms. Summary of the Invention
[0004] The present invention aims to provide a method for sensing vital signs of bedridden patients based on multimodal sensor fusion to solve the problems raised in the above background technology.
[0005] In order to achieve the above object, the present invention provides the following technical solutions:
[0006] A method for sensing vital signs of bedridden patients based on multimodal sensor fusion, comprising the following steps:
[0007] S1. Acquire data. Install a millimeter-wave radar sensor on the bed, and position the millimeter-wave radar sensor above the bed when in operation. Monitor the patient's respiratory information and heart rate information through the millimeter-wave radar sensor to obtain preliminary data. Install an infrared thermal imaging sensor on one side of the bed to monitor the patient's body temperature and obtain the patient's body temperature data. Set a pressure sensor on the upper surface of the mattress to detect the patient's posture while lying in bed and collect the patient's motion modal data. The pressure sensor uses multiple piezoresistive films or multiple piezoresistive fibers. The piezoresistive films or fibers convert force signals into corresponding resistance signals through the piezoresistive effect, and convert them into electrical signals for output through a circuit, thereby measuring static force.
[0008] S2. Data preprocessing: The microcontroller receives the patient's respiratory data, heart rate information, temperature data, and pressure data obtained by different sensors, converts them into electrical signals, and transmits them to the host computer. Wavelet filtering is used to remove noise in the echo signal obtained by the millimeter radar wave sensor. The infrared thermal imaging sensor obtains data to form an infrared thermal image in the host computer, and the infrared thermal image is grayscale normalized. The pressure data is median filtered to eliminate outliers.
[0009] S3. Data refinement: Further refine the pre-processed data to extract features related to the patient's physical signs and behavior. The frequency and amplitude of the respiratory and heartbeat waveforms are extracted from the data obtained by the millimeter-wave radar sensor to obtain the patient's accurate heart rate and respiratory data. The data obtained by the pressure sensor are processed and calculated to obtain the center of gravity position and pressure changes of the patient's body pressure distribution, as well as the patient's bed posture data. The patient's body temperature information is extracted from the infrared thermal imaging image and processed to obtain the patient's average body surface temperature data and body surface temperature change data.
[0010] S4, data analysis, processing and analyzing the heart rate data obtained in step S3, analyzing and monitoring the patient's heart state while in bed, analyzing and monitoring the respiratory data obtained in step S3, monitoring the patient's sleep state, processing and analyzing the bed posture data, monitoring the patient's sleep state, and using the average body surface temperature data and body surface temperature change data obtained in step S3 to monitor the patient's body temperature state;
[0011] S5, data comparison alarm, store the patient's data information from the previous day, input the patient's medication status for the day, analyze the patient's recovery trend based on the patient's medication status, use the patient's data from the previous day as a benchmark to analyze and obtain the patient's physical condition pre-analysis data for the day, compare the analysis data obtained in step S4, compare the current day's data with the previous day's data separately, and compare the current day's data with the pre-analysis data separately;
[0012] Set health thresholds, determine corresponding thresholds based on data types and perform comparisons. Based on the comparison results, determine whether there are any abnormalities in the human body's vital signs data parameters. If the data for the day exceeds the set health threshold, a red alarm signal will be issued.
[0013] Compare the patient's data on the day with the patient's data from the previous day and the analyzed data, set a comparison difference value, and when the comparison between the patient's data on the day and the previous day exceeds the comparison difference value, an orange alarm signal is issued.
[0014] Compare the daily data with the pre-analysis data to monitor the patient's physical changes, determine whether the medication and treatment on that day have achieved the expected therapeutic effect, set the pre-change difference value, and issue a green alarm signal when the daily data exceeds the pre-change difference value;
[0015] The three types of comparisons are performed separately but are interrelated. When one set of comparison data issues an alarm signal, the other two sets of comparison data automatically record the data nodes and mark them for display, facilitating subsequent overall recording and analysis of the patient's data.
[0016] S6. Data display: the data of the day, the comparison between the data of the day and the data of the previous day, and the comparison between the data of the day and the pre-analysis data are all visualized in the form of charts; the image information is displayed and the alarm signal is received through the terminal device.
[0017] S7. Data sharing, establishing an information transmission channel, using mobile phone software as the carrier medium, sending data and alarm information to the mobile phone software of patients' families, head nurses and attending doctors in real time, so that patients' families and medical staff can monitor patients' conditions in real time.
[0018] Preferably, the red alarm signal includes an uninterrupted sound alarm and SMS reminder. When the red alarm signal is issued, the terminal device emits an uninterrupted alarm sound, which needs to be manually turned off by medical staff. At the same time, an alarm SMS is sent to all related mobile phone software to facilitate rapid treatment of the patient.
[0019] The orange alarm signal includes an intermittent sound alarm and SMS reminder. When the orange alarm signal is issued, the remote device will emit an intermittent alarm sound, which will automatically turn off after a period of time. At the same time, an alarm SMS will be sent to the relevant medical staff;
[0020] The green alarm signal includes a text message reminder. When the green alarm signal is issued, you only need to send an alarm text message to the medical staff.
[0021] Preferably, in step S5, the health threshold is set as a respiratory rate lower than 10 times / minute or higher than 30 times / minute, a heart rate lower than 50 times / minute or higher than 120 times / minute, a body surface temperature higher than 38°C or lower than 35°C, and a body position that has not changed for more than 5 hours.
[0022] Preferably, in step S1, the millimeter wave radar sensor collects the patient's echo signal at a frequency of 50 times per second, the pressure sensor collects body pressure data at a frequency of 20 times per second, and the infrared thermal imaging sensor collects thermal imaging images at a frequency of 10 times per second.
[0023] Preferably, the pressure sensors are arranged horizontally and vertically to form an array with a spacing of 3-8 cm and distributed throughout the mattress. The piezoresistive film is fixed to the mattress in a distributed wiring manner, and the piezoresistive fibers are fixed to the mattress in a criss-cross manner to form cross-sensing points.
[0024] Preferably, it also includes data storage, storing the patient's data in a database, using machine learning algorithms to analyze historical data, establishing the patient's health profile and personalized monitoring model, and providing a reference basis for subsequent medical diagnosis and health management.
[0025] Compared with the existing technology, this technical solution has the following beneficial effects:
[0026] (1) This technical solution does not require patients to actively cooperate in wearing sensors, which reduces patients' discomfort, improves patients' compliance, and avoids damage to the skin caused by wearing sensors. By integrating data collected by multiple sensors, it can fully obtain information such as patients' vital signs (such as respiratory rate, heart rate, and body surface temperature) and behavioral status (such as changes in body position and abnormal sounds), and more accurately reflect the patient's physical condition. At the same time, a multi-band alarm mode is set up to generate different alarm signals for different critical conditions of patients, assisting medical staff to quickly complete the treatment of patients. At the same time, the multi-band alarms are processed and sent separately to improve the accuracy of the alarm and the accuracy of data processing and analysis, thereby improving the monitoring effect. Intercommunication transmission is also established between multi-level alarm signals, which facilitates the analysis of multiple sets of alarm signal data and improves the monitoring effect.
[0027] (2) This technical solution comprehensively analyzes and preprocesses the initial data obtained, greatly improving the accuracy of the final data, improving the monitoring effect, optimizing and limiting the data collection frequency, further improving the comprehensiveness and accuracy of data collection, and further improving the data analysis and monitoring effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 It is a schematic diagram of the process of the present invention; DETAILED DESCRIPTION
[0029] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments:
[0030] A method for sensing vital signs of bedridden patients based on multimodal sensor fusion, comprising the following steps:
[0031] S1. Acquire data. Install a millimeter-wave radar sensor on the bed, and position the millimeter-wave radar sensor above the bed when in operation. Install the millimeter-wave radar sensor by means of a mounting bracket. Monitor the patient's respiratory information and heart rate information through the millimeter-wave radar sensor to obtain preliminary data. Install an infrared thermal imaging sensor on one side of the bed to monitor the patient's body temperature and obtain the patient's body temperature data. Set a pressure sensor on the upper surface of the mattress to detect the patient's posture when lying in bed and collect the patient's motion modal data. The pressure sensor uses multiple piezoresistive films or multiple piezoresistive fibers. The piezoresistive films or piezoresistive fibers convert force signals into corresponding resistance signals through the piezoresistive effect, and convert them into electrical signals for output through a circuit, thereby measuring static force.
[0032] The pressure sensors are arranged horizontally and vertically to form an array with a spacing of 5 cm and distributed throughout the mattress. Piezoresistive films are used, that is, the piezoresistive films are fixed to the mattress in a distributed wiring manner. Piezoresistive fibers are used, that is, the piezoresistive fibers are fixed to the mattress in a criss-cross manner to form cross-sensing points.
[0033] The millimeter-wave radar sensor collects the patient's echo signal at a frequency of 50 times per second, the pressure sensor collects body pressure data at a frequency of 20 times per second, and the infrared thermal imaging sensor collects thermal imaging images at a frequency of 10 times per second.
[0034] S2. Data preprocessing: The microcontroller receives the patient's respiratory data, heart rate information, temperature data, and pressure data obtained by different sensors, converts them into electrical signals, and transmits them to the host computer. Wavelet filtering is used to remove noise in the echo signal obtained by the millimeter radar wave sensor. The infrared thermal imaging sensor obtains data to form an infrared thermal image in the host computer. The infrared thermal image is grayscale normalized, and the pressure data is median filtered to eliminate outliers, thereby obtaining accurate monitoring data.
[0035] S3. Data refinement: Further refine the pre-processed data to extract features related to the patient's physical signs and behavior. The frequency and amplitude of the respiratory and heartbeat waveforms are extracted from the data obtained by the millimeter-wave radar sensor to obtain the patient's accurate heart rate and respiratory data. The data obtained by the pressure sensor are processed and calculated to obtain the center of gravity position and pressure changes of the patient's body pressure distribution, as well as the patient's bed posture data. The patient's body temperature information is extracted from the infrared thermal imaging image and processed to obtain the patient's average body surface temperature data and body surface temperature change data.
[0036] S4, data analysis, processing and analyzing the heart rate data obtained in step S3, analyzing and monitoring the patient's heart state while in bed, analyzing and monitoring the respiratory data obtained in step S3, monitoring the patient's sleep state, processing and analyzing the bed posture data, monitoring the patient's sleep state, and using the average body surface temperature data and body surface temperature change data obtained in step S3 to monitor the patient's body temperature state;
[0037] S5, data comparison alarm, store all monitoring data of the patient, retrieve the monitoring data information of the patient from the previous day, input the patient's medication status for the day, and also input the patient's treatment status for the day, analyze the patient's recovery trend based on the patient's medication status, and use the patient's data from the previous day as a benchmark to obtain pre-analysis data of the patient's physical condition on the day. This pre-analysis data is the physical health data that the patient can theoretically achieve after drug or surgical treatment. Based on the analysis data obtained in step S4, the data of the day is compared with the data of the previous day separately, and the data of the day is compared with the pre-analysis data separately;
[0038] Set health thresholds, determine corresponding thresholds based on data types and perform comparisons. Based on the comparison results, determine whether there are any abnormalities in the human body's vital signs data parameters. A red alarm signal will be issued when the daily data exceeds the set health thresholds. The health thresholds are set as a respiratory rate below 10 beats / minute or above 30 beats / minute, a heart rate below 50 beats / minute or above 120 beats / minute, a body surface temperature above 38°C or below 35°C, and no change in body position for more than 5 hours.
[0039] The patient's data information for the day is compared with the patient's data from the previous day and the analysis data, and a comparison difference value is set. The comparison difference value is based on the data from the previous day. When the difference between the data on the day and the data from the previous day exceeds the comparison difference value, that is, when the difference between the data on the day and the data from the previous day exceeds the comparison difference value, an orange alarm signal is issued.
[0040] Compare the daily data with the pre-analysis data to monitor the patient's physical changes, determine whether the medication and treatment of the day have achieved the expected therapeutic effect, and set the pre-change difference value. The pre-change difference value is set based on the pre-analysis data to determine the gap between the patient's actual physical data and the pre-analysis data. When the comparison between the daily data and the pre-analysis data exceeds the pre-change difference value, a green alarm signal is issued;
[0041] The three types of comparisons are performed separately but are interrelated. When one set of comparison data issues an alarm signal, the other two sets of comparison data automatically record the data nodes and mark them for display, facilitating subsequent overall recording and analysis of the patient's data. For example, when a green alarm signal is issued, a red alarm signal or an orange alarm signal may not be detected at this time, but the corresponding data will still be marked at this time to facilitate subsequent analysis of the data comparison status of the red alarm signal and the orange alarm signal when the green alarm signal is issued, thereby determining whether there is an error in the corresponding data analysis.
[0042] The red alarm signal includes a continuous sound alarm and SMS reminder. When a red alarm signal is issued, the terminal device will emit a continuous alarm sound. The continuous alarm sound needs to be manually turned off by medical staff. At the same time, an alarm SMS will be sent to all related mobile phone software to facilitate timely treatment of patients;
[0043] The orange alarm signal includes an intermittent sound alarm and SMS reminder. When the orange alarm signal is issued, the remote device will emit an intermittent alarm sound, which will automatically turn off after a period of time. At the same time, an alarm SMS will be sent to the relevant medical staff;
[0044] The green alarm signal includes a text message reminder. When the green alarm signal is issued, you only need to send an alarm text message to the medical staff.
[0045] S6. Data display: the data of the day, the comparison between the data of the day and the data of the previous day, and the comparison between the data of the day and the pre-analysis data are all visualized in the form of charts; the image information is displayed and the alarm signal is received through the terminal device.
[0046] S7. Data sharing: Establishing an information transmission channel. Using mobile phone software as a carrier medium, data and alarm information are sent to the mobile phone software of the patient's family, head nurse, and attending physician in real time, allowing the patient's family and medical staff to monitor the patient's condition in real time. A separate health file is established for each patient based on the patient's information, and a file number and password are assigned. The nurse, attending physician, and family members responsible for the patient can bind the corresponding patient in the mobile phone software using the file number.
[0047] It also includes data storage, storing patient data in a database, using machine learning algorithms to analyze historical data, and establishing a personalized patient monitoring model to provide a reference basis for subsequent medical diagnosis and health management, and provide big data support for the acquisition of pre-analysis data.
[0048] 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 method for monitoring vital signs of bedridden patients without any sensation based on multimodal sensor fusion, characterized in that: The following steps are involved: S1. Acquire data. Install a millimeter-wave radar sensor on the bed, and position the millimeter-wave radar sensor above the bed when in operation. Monitor the patient's respiratory information and heart rate information through the millimeter-wave radar sensor to obtain preliminary data. Install an infrared thermal imaging sensor on one side of the bed to monitor the patient's body temperature and obtain the patient's body temperature data. Set a pressure sensor on the upper surface of the mattress to detect the patient's posture while lying in bed and collect the patient's motion modal data. The pressure sensor uses multiple piezoresistive films or multiple piezoresistive fibers. The piezoresistive films or fibers convert force signals into corresponding resistance signals through the piezoresistive effect, and convert them into electrical signals for output through a circuit, thereby measuring static force. S2. Data preprocessing: The microcontroller receives the patient's respiratory data, heart rate information, temperature data, and pressure data obtained by different sensors, converts them into electrical signals, and transmits them to the host computer. Wavelet filtering is used to remove noise in the echo signal obtained by the millimeter radar wave sensor. The infrared thermal imaging sensor obtains data to form an infrared thermal image in the host computer, and the infrared thermal image is grayscale normalized. The pressure data is median filtered to eliminate outliers. S3. Data refinement: Further refine the pre-processed data to extract features related to the patient's physical signs and behavior. The frequency and amplitude of the respiratory and heartbeat waveforms are extracted from the data obtained by the millimeter-wave radar sensor to obtain the patient's accurate heart rate and respiratory data. The data obtained by the pressure sensor are processed and calculated to obtain the center of gravity position and pressure changes of the patient's body pressure distribution, as well as the patient's bed posture data. The patient's body temperature information is extracted from the infrared thermal imaging image and processed to obtain the patient's average body surface temperature data and body surface temperature change data. S4, data analysis, processing and analyzing the heart rate data obtained in step S3, analyzing and monitoring the patient's heart state while in bed, analyzing and monitoring the respiratory data obtained in step S3, monitoring the patient's sleep state, processing and analyzing the bed posture data, monitoring the patient's sleep state, and using the average body surface temperature data and body surface temperature change data obtained in step S3 to monitor the patient's body temperature state; S5, data comparison alarm, store the patient's data information from the previous day, input the patient's medication status for the day, analyze the patient's recovery trend based on the patient's medication status, use the patient's data from the previous day as a benchmark to analyze and obtain the patient's physical condition pre-analysis data for the day, compare the analysis data obtained in step S4, compare the current day's data with the previous day's data separately, and compare the current day's data with the pre-analysis data separately; Set health thresholds, determine corresponding thresholds based on data types and perform comparisons. Based on the comparison results, determine whether there are any abnormalities in the human body's vital signs data parameters. If the data for the day exceeds the set health threshold, a red alarm signal will be issued. Compare the patient's data on the day with the patient's data from the previous day and the analyzed data, set a comparison difference value, and when the comparison between the patient's data on the day and the previous day exceeds the comparison difference value, an orange alarm signal is issued. Compare the daily data with the pre-analysis data to monitor the patient's physical changes, determine whether the medication and treatment on that day have achieved the expected therapeutic effect, set the pre-change difference value, and issue a green alarm signal when the daily data exceeds the pre-change difference value; The three types of comparisons are performed separately but are interrelated. When one set of comparison data issues an alarm signal, the other two sets of comparison data automatically record the data nodes and mark them for display, facilitating subsequent overall recording and analysis of the patient's data. S6. Data display: the data of the day, the comparison between the data of the day and the data of the previous day, and the comparison between the data of the day and the pre-analysis data are all visualized in the form of charts; the image information is displayed and the alarm signal is received through the terminal device. S7. Data sharing, establishing an information transmission channel, using mobile phone software as the carrier medium, sending data and alarm information to the mobile phone software of patients' families, head nurses and attending doctors in real time, so that patients' families and medical staff can monitor patients' conditions in real time.
2. The method for monitoring vital signs of bedridden patients without any sensation based on multimodal sensor fusion according to claim 1, wherein: The red alarm signal includes a continuous sound alarm and SMS reminder. When a red alarm signal is issued, the terminal device will emit a continuous alarm sound, which must be manually turned off by medical staff. At the same time, an alarm SMS will be sent to all related mobile phone software to facilitate timely treatment of patients; The orange alarm signal includes an intermittent sound alarm and SMS reminder. When the orange alarm signal is issued, the remote device will emit an intermittent alarm sound, which will automatically turn off after a period of time. At the same time, an alarm SMS will be sent to the relevant medical staff; The green alarm signal includes a text message reminder. When the green alarm signal is issued, you only need to send an alarm text message to the medical staff.
3. The method for monitoring vital signs of bedridden patients without any sensation based on multimodal sensor fusion according to claim 1, wherein: In step S5, the health threshold is set as a respiratory rate lower than 10 times / minute or higher than 30 times / minute, a heart rate lower than 50 times / minute or higher than 120 times / minute, a body surface temperature higher than 38°C or lower than 35°C, and a body position that has not changed for more than 5 hours.
4. The method for monitoring vital signs of bedridden patients without any sensation based on multimodal sensor fusion according to claim 1, wherein: In step S1, the millimeter wave radar sensor collects the patient's echo signal at a frequency of 50 times per second, the pressure sensor collects body pressure data at a frequency of 20 times per second, and the infrared thermal imaging sensor collects thermal imaging images at a frequency of 10 times per second.
5. The method for monitoring vital signs of bedridden patients without any sensation based on multimodal sensor fusion according to claim 1, wherein: The pressure sensors are arranged horizontally and vertically to form an array with a spacing of 3-8 cm and distributed throughout the mattress. The piezoresistive film is fixed to the mattress with a distributed wiring method, and the piezoresistive fibers are fixed to the mattress in a criss-cross manner to form cross-sensing points.
6. The method for monitoring vital signs of bedridden patients without any sensation based on multimodal sensor fusion according to claim 1, wherein: It also includes data storage, storing patient data in a database, using machine learning algorithms to analyze historical data, establishing patient health records and personalized monitoring models, and providing a reference for subsequent medical diagnosis and health management.