RU2463949C2 - Method of automated remote estimation of parameters of human or animal motor activity, respiration and pulse - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: invention relates to field of automated estimation of parameters of human or animal motor activity, respiration and heart beat by means of electromagnetic radiosignals. Method includes reception of signal reflected from said live object; processing and analysis of reflected signal by determination of phase difference, conditioned by signal, reflected from object; obtaining one or several physiological parameters of object. To recognise types of object body movements probing signal with step-by-step frequency modulation is used, as receiver of reflected signals used is coherent square receiver.

EFFECT: absence of necessity of exact positioning by distance with respect to human or animal and improvement of quality of contactless analysis of patient's movements, their respiration and heart beat pattern.

3 cl, 5 dwg

Description

Technical field

The invention relates to the field of automated assessment of the parameters of motor activity, respiration and heartbeat of a person or animal using electromagnetic radio signals.

State of the art

Evaluation of the parameters of a person’s physical activity, respiration and pulse may be relevant in many cases, for example, in assessing the quality of sleep, diagnosing various types of respiratory disorders in sleep (obstructive, central apnea, etc.). In addition, it will allow you to accurately determine the number and nature of the patient’s movements for a certain period of time, make a conclusion about the need for anti-decubitus measures, and thus prevent the formation of pressure sores without the use of special anti-decubitus mattresses and pillows. An assessment of motor activity is also necessary in the diagnosis of motor disorders, such as Parkinson's disease or the consequences of a stroke. The proposed method can be used for automated assessment of the condition of animals during experiments in the field of pharmacology and zoopsychology.

Methods for the automated assessment of the parameters of motor activity, respiration and human pulse exist. All of them can be divided into contact and contactless. Some of them allow you to register these parameters at the same time. The main advantage of non-contact methods is the fact that, unlike contact methods, their use does not interfere with the patient, does not affect the sleep quality of the subjects and does not affect the results of experiments.

A contact method is proposed in which a ball placed in a hollow housing is used to record human movements, at the tops of which electrical contacts are closed by a ball, the housing has the ability to be installed on the upper part of the patient’s body (RF patent 2096995 C1, “Mobile device for recording and accumulating physiological parameters of recognition and diagnosis of sleepy apneetic syndrome ", publ. 11/27/1997). The fact that this method is essentially contact is its main disadvantage. In addition, the use of this kind of sensor will affect the sleep quality of the subject (the patient will not be able to sleep on the part of the body to which the sensor is attached), additional sensors are necessary to register respiration and heartbeat parameters.

Also, movements of human limbs can be detected by contact using accelerometers fixed on the thigh (for example, Tritrac-R3D, Hemokentics, Inc.,) or on the wrist (ActiWatch AW64, MiniMitter Co.). However, devices of this type allow you to evaluate the movement of only that part of the body on which are fixed.

For a non-contact assessment of motor activity, a method is proposed in which the sensor is made in the form of an elastic pneumatic capacity (mattress or chair), the pressure in which changes with the movement of a person (AS USSR 348203, “Pneumatic actograph”, published on 08.23.1972). The main disadvantage of this type of sensors is the design complexity and low accuracy of the method.

The movement of a person’s body, his breathing and heartbeat can also be estimated using piezoelectric sensors mounted in the base of the bed or located under the mattress, if pressure is applied to them, an electric charge is generated that can be detected, one of which is based on methods of registration of movements (US patent 5,724,990, "Device for monitoring a person", publ. 03/10/1998). The disadvantages of this device are similar to the previous one.

One of the methods of non-contact monitoring of motor activity, respiration and heartbeat is optical interferometry (US patent 6,352,517, "Device for optical monitoring of anatomical movements and the method of its application", publ. 05.03.2002). The limitations of this method include the fact that the signals of the optical frequency range are blocked by clothes and bedding, which makes it inapplicable in the case of analysis of sleep quality or monitoring of sedentary patients.

It is possible to register movements, breathing, and heartbeat using ultrasound (U.S. Patent Application Laid-Open No. 2010/0027378, Method of Detecting People, Published 04.02.2010), but the signal-to-noise ratio for this type of device is low.

It is also proposed to use electromagnetic radiation of the radio frequency range for similar purposes (U.S. Patent Application Laid-Open No. 2009/0203972, “Method, System, and Device for Monitoring Physiological Signals,” published August 13, 2009), with simultaneous recording of respiration and pulse frequencies. Separation of signals of motor activity, palpitations and respiration is supposed to be carried out using specially selected filters. This method was chosen as the closest analogue (prototype), since it is proposed to use electromagnetic radiation to monitor human movements, but the disadvantage of this method is the fact that it is not possible to recognize various types of movements, such as movements of the limbs, head, trunk, which is especially relevant when assessing the quality of patient’s sleep and monitoring sedentary patients and patients undergoing surgery. Another disadvantage of the method is the need for accurate positioning of the device in range relative to the person, otherwise distortion of the useful signal, a drop in its power, as the subject will go beyond the range of the device, are possible.

Disclosure of invention

The technical task of the invention is to improve the automated assessment of respiration, heartbeat, type and intensity of motor activity of a person or animal using electromagnetic signals of the radio frequency range and simplify the procedure for conducting research, namely: the lack of the need for accurate positioning in range relative to a person or animal and improving the quality of contactless analysis of the patient's movements, pattern of his breathing and heartbeat.

This problem is solved in that the method of automated remote assessment of the parameters of motor activity, respiration and pulse of a person or animal includes receiving a signal reflected from the specified living object; processing and analysis of the reflected signal by determining the phase difference due to the signal reflected from the object; obtaining one or more physiological parameters of the object. The aforementioned physiological parameters include one or more parameters characterizing respiration, cardiac activity, and body movements of an object; the conversion of the selected received information to the form available to the user. The method is characterized in that they further include recognition of the types of body movements of the object, for this, a probing signal with step frequency modulation is used, and a coherent quadrature receiver is used as a receiver of signals reflected from the object.

The probe signal is mainly formed in the form of bursts of N = 4 ... 64 partial frequency components of the gigahertz frequency range, and the value of the kth frequency component in the packet is determined by the formula f _k = f ₁ + k · Δf, where f ₁ is the first of N frequencies, Δf is the frequency step, k = 1 ... N is the serial number of the frequency component.

The method includes successive stages: 1) selecting a probe frequency that is optimal for the position of the object relative to the transceiver unit of the device, 2) analyzing and comparing the records obtained for each of the probe frequencies; 3) dividing the signal recording into intervals between periods corresponding to the movements of a living object; 4) determining the parameters of respiration and cardiac activity.

This leads to an increase in the information content of the non-contact analysis of the physiological parameters of a living object due to the recognition of various types of movement, while in contrast to the prototype there is no need for accurate positioning of the receiver in range relative to the person.

Each of the quadrature N frequency components in the recorded signal contains information about the movement of the reflecting surface of a living object, however, since each of the probing frequencies differs from the subsequent one by Δf, the information recorded for each of them differs in amplitude and spectral composition, which allows you to ensure reliable registration of the useful signal for any range between the antenna unit and the patient within 20 m. The preparation procedure for the study is simplified, since there is no need ti positioning device at a certain distance to the person in order to obtain a satisfactory signal quality.

List of drawings

Figure 1 - diagram of a device for implementing the proposed method.

Figure 2 - diagram of the recognition algorithm of various types of movements using a device that implements the method.

Figure 3, 4, 5 - graphs of the recorded signals.

The implementation of the invention

To implement the method, use the device (diagram of figure 1) with the following functional components. The “crystal oscillator" synchronizes the operation of the transmitter and receiver. The "transmitter" generates a signal with step frequency modulation in accordance with the specified parameters. The “receiver” is built according to a superheterodyne circuit with double frequency conversion (this ensures high sensitivity, selectivity and noise immunity), provides the reception of a signal reflected from a biological object, its conversion and the receipt of two quadrature components at the output. It is designed to receive an electromagnetic signal. The “output device” contains amplifiers and an analog multichannel band-pass filter (AMPF), designed to filter the DC component and high-frequency interference at each of the transmitter frequencies for each of the quadrature components. The ADC provides digitization of each of the two quadrature components. The "microcontroller" is designed to control the frequency synthesizer in the transmitter, the frequency synthesizer in the receiver, AMPPF, and also provides the recording of digitized values in a file that is transmitted to a personal computer (PC) for further processing: extracting data on the respiration and heartbeat parameters from the registered signal, types and intensity of body movements.

The method of automated remote assessment of the parameters of motor activity, respiration and pulse of a person or animal is as follows. The signal emitted by the transmitter is reflected from the interface between media with different dielectric properties. In this case, the signal reflected from a person or animal acquires specific modulation due to the movement of the body as a whole, its breathing and heartbeat. The reflected signal is captured by the receiver and fed to the output device, after which the signal is digitized using the ADC and transmitted to the PC. In a PC using filtering algorithms, the received signal is divided into the implementation of respiration, heartbeat, and motor activity. Then, using the recorded quadrature records for 16 frequencies, a search is made for the initial moments of individual motion sequences, comparison of recording segments corresponding to the motion for each of the probing frequencies, after which it is concluded that the selected motion sequence belongs to a certain type.

To recognize different types of motor activity, an algorithm for recognizing various types of movements, consisting of the following stages, is proposed:

- selection of the quadrature of the recorded signal having a maximum power spectral density;

- determination of the threshold level value characteristic of motor activity for the selected quadrature;

- localization in the selected quadrature using the selected threshold level of time intervals when motion artifacts are observed (t _A , A = 1.2 ... M, where M is the number of motion artifacts) and when they are absent (t _C , C = 1.2 ... M + 1);

- determining for each of the intervals t _{C the} quadrature of the recorded signal (QNum _C ) having a maximum power spectral density at a given time interval;

- a comparison of the spectral power density for the selected quadrature in adjacent sections t _C and t _{C + 1} and the spectral power density for artifacts of motion in the time interval t _A , for A = C, in order to determine the type of artifact of motion that takes place during the time t _A .

An example implementation of the method

The subject (person) was located on the bed in a supine position, at a distance of 3 m from him, the receiving and transmitting antennas of the device were installed and sent to the bed. As a probing signal, a signal formed in the form of bursts of 16 frequency components in the range from 3.6 to 4.0 GHz was used. The reflection from the test signal was recorded for 30 minutes.

During this time, the subject performed the following actions on the operator’s command:

1) lay without moving, while using the proposed method, the breathing and heartbeat signals were extracted from the initial signal received by the device (Figure 3);

2) moved his right hand, bending it in the elbow joint (Figure 4), the recording section corresponding to this type of body movements was detected by the proposed method and classified as “arm movement”;

3) changed the position of the body relative to the device, turning from the back to the right side (Figure 5), the recording section corresponding to this type of body movement was detected by the proposed method and classified as “body rotation”.

Figure 3, 4, 5 shows the signals corresponding to a quadrature having a maximum power spectral density in the time interval shown in each of the graphs. When the body position relative to the device changes, the frequency number and the quadrature having the maximum power spectral density in the selected time interval change, therefore, Fig. 5 shows graphs of two quadrature, one of which (shown by the dash-dot line) had the maximum power spectral density in the time interval, preceding the rotation of the body, and the second (shown by a solid line) - on the time interval following the rotation of the body.

Claims (3)

1. A method for automated remote assessment of the parameters of motor activity, respiration and pulse of a person or animal, including receiving a signal reflected from the specified living object; processing and analysis of the reflected signal by determining the phase difference due to the signal reflected from the object; obtaining one or more physiological parameters of the object, the aforementioned physiological parameters include one or more parameters characterizing respiration, cardiac activity and body movements of the object; the conversion of the selected received information to a form accessible to the user, characterized in that it further includes recognition of the types of body movements of the object, for this, a probing signal with step frequency modulation is used, and a coherent quadrature receiver is used as a receiver of signals reflected from the object. 2. The method according to claim 1, characterized in that the probe signal is formed in the form of bursts of N = 4..64 partial frequency components of the gigahertz frequency range, and the magnitude of the kth frequency component in the packet is determined by the formula f _k = f ₁ + k · Δf, where f ₁ is the first of N frequencies; Δf is the frequency step; k = l..N is the serial number of the frequency component. 3. The method according to claim 1, characterized in that it includes successively stages: selecting a probing frequency optimal for the position of the object relative to the transceiver unit of the device; analysis and comparison of records obtained for each of the probing frequencies; dividing the signal recording into intervals between periods corresponding to the movements of a living object; determination of respiration and cardiac activity.

Images