Radar system for simultaneous non-contact detection of vital signs of multiple people

Abstract

This invention provides a continuous wave radar system for monitoring the vital signs of multiple people simultaneously and non-contact. This system can be used for vehicle occupants, patients in hospital care rooms, and elderly people at home. Among the problems of radar systems built for this purpose are the challenge of monitoring the vital signs of multiple people simultaneously, the long time required for people to be in front of the radar, and their sensitivity to body movements. This invention is able to simultaneously focus on detecting human vital signs in a crowded environment using beamforming and eliminating other interfering frequencies. This system uses artificial intelligence and time series prediction to reduce the time required for people to be in front of a radar sensor by a third. In addition, the designed system is resistant to unconscious movements and possible jerks of people in front of the radar sensor, so that these types of movements are detected and the data related to them are not included in the main measurement.

Description

Description of the invention Title of the invention Radar system for simultaneous non-contact detection of vital signs of multiple people The technical background of Astronomy This invention is related to the application of telecommunications and radar in the manufacture of medical equipment, and specifically the monitoring of vital signs of the human body. The technical problem of the invention's objectives For a long time, there have been methods for measuring the vital signs of the body, which have often been contact-based, meaning that part of the equipment of the vital signs measuring system has been in contact with the body of the person. Among the important contact equipment for detecting vital signs, wearable sensors can be mentioned. If we want to mention the two major problems of contact measurement equipment, the first is that they are constantly connected to the human body and take away the freedom of action from the person, and the second is that connecting a device to the body of different people is not acceptable from a health perspective and in some cases it is not possible at all for people with skin problems. For this reason, non-contact equipment has recently been considered for measuring the vital signs of the body. For monitoring the vital signs of the human body in a non-contact manner, there are two general solutions: using optical equipment and radar. Here, we mean optical equipment such as cameras or equipment based on visible light. The use of the aforementioned optical equipment requires the presence of light in the environment and in many environmental conditions where sufficient light is not available, it encounters problems.In addition, using cameras is not efficient in cases where privacy is important. So the only solution is to use radar. Using radar to detect the vital signs of the human body requires that the person spend a long time without moving in front of the radar. In addition, using radar to detect the vital signs of multiple people simultaneously has its own challenges. In short, the problem with radar systems proposed for monitoring the vital signs of multiple people simultaneously is that they cannot function properly in real environments where people are scattered and there are limitations such as moving disturbing factors in the environment. In addition, the time required to monitor the vital signs of each person is relatively long, on the order of several tens of seconds. In fact, there are many cases in which people are scattered in a space and in specific locations.For example, vehicle occupants or patients on beds in hospital intensive care units or similar examples are cases where we need to continuously obtain the vital signs of several people in specific locations in a large environment and eliminate other disturbing factors such as the passage of other people, the presence of ceiling fans or table fans in the environment or any other moving factors. To do this, we have to design a system that, firstly, focuses on specific locations where people are present, and secondly, is resistant to disturbing factors in the environment such as the passage of other people or any other moving factors including the movements of the person in question. Since the respiratory rate of people is in the range of 6 to 30 breaths per minute and the heart rate is in the range of 40 to 120 beats per minute, the frequency range considered for human vital signs is between 0.1 and 2 Hz. We call this range the frequency range of vital signs.Now our goal is to, in addition to rotating the antenna beam in space alternately between individuals, simultaneously search for only the frequency range of vital signs in each area of ​​the space in question so that the system becomes resistant to other moving disturbing factors. Another very important point is that in all the cases mentioned, in order for the vital signs to be obtained correctly, a sufficient amount of radar data from each individual needs to be available. For this reason, each individual needs to be in front of the radar for a period of time without moving. In some cases, this is considered a serious limitation for which no solution has been provided. For this reason, in this invention, we intend to provide a radar system for detecting the vital signs of several people simultaneously, in which people do not have to stand in front of the radar for a long time without moving. In fact, the proposed system is able to firstly monitor several people simultaneously, secondly reduce the time required for people to stand in front of the radar, and thirdly be resistant to unconscious movements of people in front of the radar and their possible tremors. In addition to the three items mentioned, our invention is able to simultaneously calculate the heart rate and respiratory rate, as well as the variance parameter of the time interval between the peaks of the heart signal. This parameter, which is called heart rate variability, contains valuable information about the health of the people in question, and the combination of its information with the respiratory rate and heart rate will form a suitable medical information package. � Description of the state of the art and history of advances related to the invention of the common law. There are two general types of equipment available for non-contact monitoring of human vital signs (respiration rate and heart rate per minute). The first type is optical equipment and the second type is radar-based equipment. Optical equipment refers to cameras or visible light-based equipment. In patent number US20200297227 in 2020, the inventors have presented a method for finding human vital signs using an optical sensor (including a camera or any other type of optical sensor). The use of optical equipment requires the presence of light in the environment and is problematic in many environmental conditions where sufficient light is not available. Further, in patent number US20220257143 in 2022, the inventors have obtained and displayed the vital signs of the patient's body by using a camera and filtering successive frames of images of his face. However, using a camera to monitor the body's vital signs, in addition to being possible, like other optical equipment, only under certain lighting and environmental conditions, has other disadvantages, including storing private images of individuals and, as a result, the possibility of violating their privacy. For the reasons mentioned, using radar can be a very good solution. First, it operates non-contact, second, it can work in any environmental conditions (whether there is light or not), third, it protects the confidentiality of individuals' information, and finally, it has very good accuracy. In the domestic patent registration number 100167 in 2020, the inventors have presented a broadband pulse radar for monitoring the vital signs of a human. The patent number US20210325509 in 2021 also presented a continuous wave radar for detecting the vital signs of a human. In the patent number US20220218224 in 2022, a continuous wave radar is presented for detecting the vital signs of a human during sleep, so that it can analyze the depth or lack of sleep of the person using its data. In the patent number US20210076971 in 2021, a radar system is presented for detecting the vital signs of a car passenger. All of the above and similar cases have only examined the vital signs of a single target. However, in many cases, it is necessary for the vital signs of multiple targets, or in other words, multiple people, to be monitored simultaneously. There are two general ideas here. First, all humans whose vital signs are to be checked in an environment are placed in a single antenna beam.In patent number US20200268257 in 2020, the inventors used this idea. But this idea cannot be used in a large and crowded environment. This is because the transmitter and receiver use a wide beam to cover the entire environment, which causes the effects of disturbing factors such as other people passing by, a ceiling fan in the environment, or any other moving factor to be received by the radar and interfere with the operation of extracting the vital signs of the desired people. For this reason, the second idea, which is the use of several narrow antenna beams, has been proposed, so that each person is placed in one antenna beam. For this purpose, the inventors proposed an idea in patent number US20230092182 in 2023, in which a camera first detects the location of people in the environment and then tells the radar in which direction to direct its narrow beam. But as we mentioned before, in addition to not respecting the principle of confidentiality of personal information, using a camera is not possible when there is insufficient light. In addition to the above, even a narrow beam design for each individual is not sufficient to properly detect the vital signs of individuals in a real environment. There are many cases where humans are scattered in a space and in specific locations, and we need to continuously acquire their vital signs at specific positions in a large environment and eliminate other interference. Therefore, in addition to rotating the antenna beam in space alternately between individuals, we need to simultaneously search for only the frequency range of vital signs in each area of ​​the space in question, so that the system is robust to other moving interference factors. Providing a solution to an existing technical problem accompanied by an accurate, sufficient, and integrated invention In this invention, we have designed a radar system for detecting the vital signs of multiple people in such a way that the radar transmitter, by designing an appropriate space-time code, rotates its transmitted beam periodically in time between people and focuses on one of them each time and repeats this continuously. In addition, each time for each person, it focuses only on the vital sign frequency and discards the rest of the frequencies. In addition, to solve the last problem mentioned, namely the time required to measure the vital signs of each person, we have presented an artificial intelligence-based network to reduce the time required for people to be in front of the radar. A schematic of the system in question is shown in Figure (1). The system consists of a Frequency Modulated Continuous Wave (FMCW) radar (101) whose transmitter (102) and receiver (103) have more than one antenna. In Figure (2), the structure of the operation before sending is seen in the transmitter. In order for the transmitter antenna beam to rotate intermittently on the people, a space-time code is used. Therefore, first, using the space-time code, a code is assigned to each person at any time and this code is modulated and sent as a waveform transmitted from the transmitter antenna in such a way that the transmitter antenna beam is focused towards the angle where that person is located. Then this is done for the second person to focus the transmitter antenna beam on that person and this is done until the last person. After that, this is repeated again from the beginning and this operation is repeated intermittently until the end. In fact, up to this point, in the considered space, the radar sees only the people we want. Next, with the processing operations that we will perform in the radar receiver, we can eliminate the disturbing movements that exist in the radar view and only keep and amplify the frequency caused by the vital signs of the people. Figure (3) shows the structure of the processing operation in the receiver. Since continuous wave radar is used, the received signal is taken in two dimensions after demodulation and passing through the receiver filter, Fast Fourier Transform (FFT). So far, we have a three-dimensional matrix of received data, the first dimension of which is the distance cells, the second dimension of which is the angle cells, and the third dimension of which is the so-called slow-time cells. Next, using range detection, the cell in which the target is located is identified and its information vector is extracted in the slow-time direction. Therefore, from now on until the last part of the receiver block diagram, we will work with this vector, which is a vector with elements consisting of complex numbers. First, by removing the mean from the vector, we make its DC offset value zero. In the next step, by extracting the angle for each element of the vector, the chest displacement value for each individual is obtained.Since the subject may have made unconscious movements or been shaken, the next step is to discard the data for the times when the subject was shaken. This is done by checking two factors: if the subject left the distance / angle cell in which they were located or if their movement frequency exceeded the vital signs frequency range, the data for that moment is removed from their data vector. In the next step, data processing is followed in two parallel paths: detection of respiratory rate and detection of heart rate. First of all, the data passes through the low-pass filter related to the respiratory interval and the heart rate interval. So far, we have data for detecting respiratory rate (301) and heart rate (302). At this stage, we have the heart and respiratory signals, and first, by calculating the variance of the time interval between the peaks of the heart signal, we report heart rate variability. Next, we will calculate the heart rate and respiratory rate. If the total measurement time interval is T seconds and our sampling frequency is f Hz, we now have the amount of samples. In the final step, the Fourier transform is taken from the signals (301) and (302) and the respiratory rate and heart rate are extracted by using the detection of the location of its maximum value. Since the accuracy of the output of the aforementioned Fourier transform is directly proportional to the number of samples, the number of samples must be increased somehow before the Fourier transform.This is why all the methods presented so far consider a significant amount for the total time of receiving the signal T, which means that the person in question must be in front of the radar sensor for a significant amount of time (on the order of tens of seconds) without moving. Therefore, we have presented the idea of ​​time series prediction for this part, in which, using artificial intelligence, we predict some time before and after the chest movement signal, thereby increasing the number of samples and, as a result, reducing the time required for the person to be in front of the radar sensor. For this purpose, we have obtained a lot of data from the vital signs of different people and used them to train the aforementioned artificial intelligence networks. As you can see in the block diagram of Figure (3), the signals (301) and (302) first pass through the Long Short Term Memory (LSTM) time series prediction blocks (which have previously been trained with a large amount of real data).Next, the Fourier transform is taken from signals (303) and (304), whose number of samples is three times that of signals (301) and (302), and by detecting the maximum value of the Fourier transform, the vital signs of the individuals are obtained. In fact, with this idea, the time required for the individual to be in front of the radar is reduced to one third, which increases the possibility of implementing the system in real environments. Next, we will describe the aforementioned artificial intelligence-based network. The structure of the network can be seen in Figure (4). The aforementioned network is two five-layer LSTM networks for predicting previous and subsequent samples, and we will explain its parameters in the following training process. The network for predicting previous data is called backward and the subsequent data is called forward. If the number of real samples we want is N, the training data is prepared in such a way that it is sampled from a large number of different people. The first N samples and the last N samples are set aside as labels for the backward and forward LSTM networks, respectively, and the middle N samples are entered into the network as real data from the path (401) in Figure (4). First, the input data (401) is entered into the preprocessing block. In this block, the data is normalized and entered into the five-layer LSTM network, with the aim of having the least difference from the label data. All data obtained from different individuals is entered into the LSTM networks as described, and thus the network training stage is completed. After the training phase is completed, the network is used in such a way that N samples received from each individual enter the network through path (401) and finally exit with dimensions through path (402). Explanation of shapes, patterns, and patterns Figure 1: General diagram of a multi-human vital signs measurement system Part (101): FMCW continuous wave radar Part (102): Multi-antenna transmitter Part (103): Multi-antenna receiver Part (104): Transmitter processing block Part (105): Receiver processing block Part (106): Antenna beam at the first time intended for the first person Episode (107): The antenna beam at the end time intended for the last person Part (108): The space where the first person was present and the antenna beam covers that area at the first moment. Episode (109): The space where the last person was present and the antenna beam covers that space at the last moment. Figure (2): Transmitter processing block diagram Figure (3): Receiver processing block diagram Part (301): Data signal related to the number of breaths before delivery to the AI ​​network Part (302): Data signal related to the heart rate before delivery to the AI ​​network Part (303): Data signal related to the number of breaths after passing through the artificial intelligence network whose number of samples is three times the original samples Part (304): Data signal related to the number of heartbeats after passing through the artificial intelligence network, the number of samples of which is three times the original samples. Figure (4): Block diagram of LSTM AI-based network for time series prediction Part (401): Input data to the LSTM artificial intelligence network Part (402): Output data of an LSTM artificial intelligence network whose dimensions are three times the input data A clear and precise statement of the advantages of the claimed invention over prior inventions. There are contact and non-contact methods for measuring the vital signs of the body. Contact methods mean that part of the equipment of the vital signs measuring system is in contact with the body of the person. Among these types of equipment, we can mention wearable sensors. If we want to mention the two major problems of contact measurement equipment, the first is that they are constantly connected to the human body and take away the freedom of action from the person, and the second is that connecting a device to the body of different people is not acceptable from a health perspective and in some cases it is not possible at all for people with skin problems. For this reason, non-contact equipment is much more efficient for measuring the vital signs of the body. For monitoring the vital signs of the human body in a non-contact manner, there are two general solutions: using optical and radar equipment. We mean optical camera equipment or equipment based on visible light. The use of the aforementioned optical equipment requires the presence of light in the environment and in many environmental conditions where sufficient light is not available, they encounter problems.In addition, using a camera is not effective in cases where personal privacy is important. For the reasons mentioned, using radar can be a very good solution. First, it operates non-contact, second, it can work in any environmental conditions (whether there is light or not), third, it protects the principle of confidentiality of personal information, and finally, it has very good accuracy. There are many inventions that only examine the vital signs of a target using radar. This is while in many cases it is necessary to monitor the vital signs of multiple targets or in other words, multiple people simultaneously. The precise advantage of our invention over other radar inventions for detecting the vital signs of multiple people simultaneously is that firstly, by designing the space-time code and rotating the transmitter antenna beam, it eliminates moving interfering factors in the environment in such a way that it can distinguish humans from other objects in an environment and only amplify the frequency related to human vital signs and weaken other interfering frequencies. Secondly, a very important point is that in all existing inventions, in order for vital signs to be obtained correctly, a sufficient amount of radar data from each person needs to be available. For this reason, each person needs to be in front of the radar for a long time without moving. In some cases, this is considered a serious limitation for which no solution has been provided.Therefore, our system design is such that, using artificial intelligence-based algorithms, it has reduced the time required for people to be in front of the radar sensor by one third. Thirdly, the system design is done in such a way that it is resistant to unconscious movements and possible shaking of people in front of the radar sensor. In addition, our invention is able to simultaneously provide the variance parameter of the time interval of the peaks of the heart signal, in addition to calculating the heart rate and respiratory rate. This parameter, which is called heart rate variability, contains valuable information about the health of the people in question, and the combination of its information with the respiratory rate and heart rate will form a suitable medical information package. A description of the minimum steps required to implement the invention. One of the important applications of this system is its use to detect the vital signs of the vehicle occupants. In fact, a radar sensor is installed in the front of the vehicle and continuously monitors the heart rate and respiratory rate of the vehicle occupants. Then, the raw radar data is sent to a processor via a USB link, and after performing the necessary calculations, the processor displays the final result, which is the number of heart beats and respirations per minute, on an in-car display. In addition, the aforementioned data can be continuously sent to another center to monitor the health status of the vehicle occupants online. Express mention of the industrial application of the invention In fact, this is a human vital signs monitoring system that can provide the number of breaths per minute and the number of heart beats per minute for several people simultaneously in a short time and while being resistant to their unconscious movements and jerks. According to the description, this system can be used in a car for continuous monitoring of its occupants, in a hospital patient care room for monitoring the vital signs of patients in bed, at home for monitoring the vital signs of the elderly, and in many other cases. �

Claims

Claims What is claimed: Claim 1) A radar system for non-contact monitoring of vital signs, including heart rate, respiratory rate, and changes in the interval between heart rate peaks per unit time, for multiple humans and / or animals simultaneously, in which the system, using distance, Doppler, angle, and frequency spectrum information over time, detects the presence of humans in an environment from other objects and existing noises, and by eliminating frequency components outside the frequency range of vital signs and attenuating interfering frequencies related to inanimate objects and other sources of interference, only maintains and amplifies frequencies corresponding to the vital signs of humans and / or animals, such that the processed output signal allows simultaneous extraction of heart rate parameters, respiratory rate, and temporal changes in heart rate of each subject in the environment. Claim 2) According to claim 1, in addition, a network is proposed and trained that attempts to estimate and extrapolate data in a forward and backward manner. Claim 3) According to claim 1, wherein the corresponding times of the movements of the individuals are detected using the displacement of the distance cell or the exceeding of the frequency of the individuals' movement beyond the frequency range of the vital signs. Claim 4) According to claim 1, data related to the corresponding times of individuals' movements are discarded or replaced by time series prediction using data estimated by a machine learning algorithm.

Images