WO2020136302A1

WO2020136302A1 - Method for detecting ballistocardiography signals and implementation system

Info

Publication number: WO2020136302A1
Application number: PCT/ES2019/070884
Authority: WO
Inventors: Alberto PALMA LÓPEZ; Celso MARTÍNEZ BLANQUE; Encarnación CASTILLO MORALES; Antonio GARCÍA RIOS; Luis Fermín CAPITÁN VALLVEY; Antonio MARTÍNEZ OLMOS; Miguel Ángel CARVAJAL RODRÍGUEZ; Pablo ESCOBEDO ARAQUE
Original assignee: Universidad De Granada; Lo Monaco Hogar, S.L.
Priority date: 2018-12-28
Filing date: 2019-12-26
Publication date: 2020-07-02
Also published as: ES2769914A1

Abstract

The present invention relates to a method for detecting heartbeats by means of a ballistocardiography signal. By means of the increase in the signal-to-noise ratio and the increase in the power of the signal at the moment of systole, the method allows the reference points that indicate each heartbeat to be reliably identified. In particular, the method is carried out using a system to which the invention also relates.

Description

DESCRIPTION

PROCEDURE FOR THE DETECTION OF BALLISTOCARDIOGRAPHIC AND

SYSTEM THAT IMPLEMENTS IT

TECHNICAL SECTOR

The present invention can be categorized within the field of physics, specifically in the area of the measurement of physiological parameters by physical means and, in particular, from a ballistocardiogram.

Its field of application is that of medicine, and in particular that of the diagnosis associated with the measurement of movement of some part of the body, in particular, ballistocardiography, or devices that allow the simultaneous evaluation of the cardiovascular system and of different types of body conditions

STATE OF THE ART

Ballistocardiogram

A ballistocardiogram (BCG) consists of recording the mechanical movement of the heart by monitoring force or acceleration from the chest, or alternatively taking remote measurements of blood pumping activity due to the heartbeat. With this remote monitoring, the need to place electrodes or other probes on the body is avoided, allowing non-invasive measurement, without altering the subject in their daily life or during rest. The current Balistocardiography systems can be integrated into everyday objects such as beds, chairs or weighing scales, so that it is a non-invasive technique for the subject. The BCG offers good prospects in preventive medicine, for example, in the detection of physical or mental stress, in the early detection of coronary problems or in the monitoring of sleep quality. The BCG waveform is capable of providing an estimate of the blood volume of each beat and of the time intervals, making it possible to monitor the general functionality of the heart, the rhythm and the variability of the heart rhythm. It also allows us to extract useful information to evaluate certain disorders such as problems with the aortic valve or problems in the coronary arteries, which have a very direct impact on predicting the life expectancy of the subject.

Traditionally, three axes of measurement are defined for BCG records: longitudinal (head-to-foot), transverse (lateral to lateral) and dorsoventral (chest-back). Historically, Most BCG recording systems have been on the longitudinal axis because it corresponds to the main axis of blood flow, including scale measurements. Among the latter, an additional alternative proposes a methodology to detect mechanical systolic events from the BCG, in which a transfer function is applied to said signal that compensates the dynamic response of the subject's body so that the BCG reflects only the corresponding signal. to the mechanical events that take place in the heart and in the aortic root.

Recently, non-invasive recording systems have been introduced, particularly those housed in some of the elements of a bed. In these systems, the exact measurement axis is not longitudinal but rather transverse, dorsoventral or a combination of both (from now on we will call all these combinations the transversal axis), depending on the position of the subject with respect to the sensor. This fact, together with the difficulty of introducing the mechanical modeling of the elements of the bed (mattress, bed base, etc.), produce BCG much more variable than those recorded on the longitudinal axis or by electrocardiography (ECG) [O. T. Inan et al., “Ballistocardiography and Seismocardiography: Ballistocardiography and Seismocardiography: A Review of Recent Advances,” IEEE J. Biomed. Heal. Informatics, vol. 19, no. 4, pp. 1414-1427, 2015]

Heart rate detection from BCG

Recent signal processing algorithms for estimating heart rate from BCG can be divided into two groups: those that provide the heart rate averaged over several seconds, and those that detect each beat individually. In the first case, the signal is divided into time segments and for each segment the average heart rate is estimated by finding the maximum of the autocorrelation function, or the power spectral density. Other alternatives average over a number of reference points or apply empirical decomposition. However, these systems cannot provide information for analysis of cardiac variability or classification of sleep stages.

Of the algorithms that individually detect each beat, many do so by detecting slope maxima or other BCG reference points. In some cases, BCG is pre-processed using various techniques such as low-pass filtering [S. Junnila, A. Akhbardeh, A. Várri, and T. Koivistoinen, “An EMFi-film sensor based ballistocardiographic chair: Performance and cycle extraction method,” IEEE Work. Signal Process. Syst. SiPS Des. Implement., Vol. 2005, pp. 373-377, 2005 or, noise reduction by transform wavelet [X. Zhu et al. , "Real-Time Monitoring of Respiration Rhythm and Pulse Rate During Sleep," IEEE Trans. Bio ed. Eng., Vol. 53, no. 12, pp. 2553-2563, 2006.]. These algorithms that are based on the detection of reference points present unreliability for the transverse BCGs due to the great variability of these signals with the change of position of the subject and, therefore, the loss of said points.

In the last decade, another alternative has been proposed based on the detection of the closure of the heart valves during the beat. An alternative in this line estimates the time interval between beats using an algorithm based on unsupervised learning techniques that adapts to each signal individually, obtaining a set of parameters used in three independent methods to locate each beat. The main drawbacks of this technique lie in the need for a period of system training, with no useful output, and its lack of viability in case of repeated mismatched patterns in each heartbeat [C. Brüser, K. Stadlthanner, S. De Waele, and S. Leonhardt, “Adaptive beat-to-beat heart rate estimation in ballistocardiograms,” IEEE Trans. Inf. Technol. Biomed., Vol. 15, no. 5, pp. 778-786, 2011.].

Other patents that describe similar inventions are W0200907392, WO2010145009, US8262582 and WO2018020064.

For example, in WO2018020064, a methodology and system is proposed to detect mechanical systolic events from the longitudinal BCG, in which a transfer function is applied to said signal that compensates the dynamic response of the subject's body, so that the function global is flat and phase zero in the range of frequencies of interest, thus reflecting only the signal corresponding to the mechanical events that occur in the heart and the aortic root.

Another approach makes use of a sensor array by applying the fast Fourier transform along with the cepstrum analysis to the BCG to determine the intervals between beats proposed in US8262582.

The main difficulty associated with the measurement of cross-sectional BCG is their variability and which continues to be a challenge in obtaining methods and devices that have a high correlation with standardized ECG-based techniques. The faithful and non-invasive detection of the appearance of each individual heartbeat from the BCG would allow to assess the state of heart health more quickly and comfortably even over long periods of time, which would be very useful for complex analysis of cardiac variability and estimation of phases during sleep.

OBJECT OF THE INVENTION

The present invention describes a method and a device for detecting heart rhythm from the ballistocardiogram (BCG) obtained by non-invasive means.

Thus, the first object of the present invention is a method, hereinafter "method of the invention", that allows obtaining, reliably, the reference points that allow the heartbeats to be temporarily positioned.

A second object of the invention is a device, hereinafter "device of the invention", that allows the execution of the method of the invention.

Also object of the present invention are the various devices used for resting subjects that comprise the device of the invention, in particular mattresses that comprise the device of the invention.

Finally, an object of the invention is a computer program that implements the method of the invention.

The use of this invention can be especially interesting for monitoring a subject's heart rate for long periods of time for the prevention of certain cardiovascular diseases and for the evaluation of sleep quality, by estimating the various phases of sleep. Furthermore, the non-invasive and home application nature of this invention allows this monitoring in a known and comfortable environment for the subject, providing information not altered by stress due to the hospital environment. On the other hand, this type of technology allows savings in the cost of health systems by not needing hospitalization or medical personnel on site. BRIEF DESCRIPTION OF THE FIGURES

To complete the description and help in understanding the characteristics of the invention, a set of figures is included as an integral part of this description in which, by way of illustration and not limitation, the following has been represented:

Figure 1.- Schematic representation of the device of the invention. (A) represents a sensor capable of obtaining the BCG, which sends the electronic signal to an analog processing stage, in which (B) represents a stage of conversion of the sensor output to voltage, (C) represents a stage of voltage amplification, (D) represents an analog filtering stage and communicates with a digital stage, where (E) represents the analog-to-digital converter, (6) represents the processor that houses the digital processing procedure of the signal and obtain the heart rate and (G) represents the module for presenting and transmitting the system result

Figure 2.- Transformation of the digitized BCG using the procedure proposed in this invention. The following are represented from top to bottom: 1) Digitized BCG signal (step 2), 2) Differences between consecutive maximums and minimums of the signal (step 4); 3) Filtered low-pass anterior signal for both positive and negative envelope reconstruction (step 5); 4) Signal product of the square of the previous signal (step 6) in which the maximums that are considered the reference points that indicate the temporal position of each heartbeat are clearly seen.

Figure 3.- Histogram of the difference between the heart rate measured with the reference ECG and that measured with our invention with a 60-second BCG mobile processing window.

EXPLANATION OF THE INVENTION

The invention consists of a method and an apparatus for detecting the heartbeat from balistocardiographic signals, also called a balistocardiogram. The innovative solution proposed in the present invention consists in the application of a processing algorithm to the BCG that allows it to increase its amplitude in the event of events of closure of the heart valves and, in this way, isolate and detect reference points that allow identifying reliably the occurrence of a heartbeat. Also, since BCG It can be registered by sensors that are not in direct contact with the subject and integrated into everyday objects (mattress, chair or scale for people), allowing maximum freedom of movement and avoiding the discomfort of being attached to one or more electrodes with cables.

Definitions

Throughout the present invention, a balistocardiogram or "BCG" is understood to be a record of the vibrations of the body caused by the mechanical activity of the heart.

"Low-pass filtering" is understood to mean, respectively, an analog or digital filtering, characterized by allowing the lowest frequencies to pass through and attenuating the highest frequencies. Similarly, "high-pass filtering allows higher frequencies to pass and lower frequencies to be attenuated.

When it comes to a physical element, the term "stage" refers to a subsystem or set of components, which comprises the elements necessary to carry out a certain signal processing function.

Throughout the description and claims, the word "comprises" and its variants are not intended to exclude other technical characteristics, additives, components or steps. For those skilled in the art, other objects, advantages and characteristics of the invention will emerge in part from the description and in part from the practice of the invention.

Throughout this document we will use a notation that uses the symbol as a decimal separator.

Invention procedure

In its first aspect, the present invention refers to a procedure of the invention ") to measure the heart rate from a ballistocardiogram (BCG) obtained by a sensor capable of transducing mechanical vibration in said electrical signal, comprising the following stages:

1. Analog, high-pass and low-pass filtering of the analog BCG

2. Analog to digital conversion.

3. Identification of the maximums and minimums of the digitized signal. 4. Creation of a discretized signal formed by impulses in the positions where there are maximums or minimums and whose amplitude is the difference between each consecutive maximum and minimum.

5. Construction, through digital low-pass filtering, of two continuous signals, one with the positive points of the discrete signal generated in the previous step, and the other with the negative ones.

6. Creation of a new signal product of the square of the two signals from the previous step.

7. Detection of the maximums of the signal created in the previous step.

8. Association of each maximum with a heartbeat.

Each of the stages is described in more detail below:

Stage 1.- Analog filtering, in high-pass and low-pass, of the analog BCG

This stage reduces noise and electrical interference and increases the signal-to-noise ratio for efficient subsequent digital processing.

In a particular embodiment, the width of the employee passband will be between 0.1 and 40 Hz, preferably between 0.15 and 25 Hz.

Stage 2.- Analog to digital conversion.

Digitalization of the analog signal, preferably with a sampling frequency greater than 1 KHz and a minimum resolution of 10 bits.

Stage 3.- Identification of the maximums and minimums.

In the digitized signal, the absolute maximums and minimums are acquired and stored.

Stage 4 Discretized signal.

At this stage, a signal is generated that contains, in the time positions of each maximum and minimum, a value corresponding to the difference of each consecutive maximum and minimum.

Stage 5.- Construction, through digital low-pass filtering, of two continuous signals, one with the positive points of the discrete signal generated in the previous step, and the other with the negative ones. These two signals form the envelopes of the positive and negative values respectively. In a preferred embodiment, the construction of the two signals is carried out by means of digital low-pass filtering. Preferably, the bandwidth used for filtering will be between 1 Hz and 25 Hz.

Stage 6.- Creation of a new signal product of the square of the two signals from the previous step.

This signal is proportional to the area and therefore to the power of the signal whose envelopes are those of the previous step. That is, detection consists of looking for the regions where energy is concentrated, that is, where both signals are distanced. The result is a signal with very sharp peaks in the regions where the oscillations due to the presence of each beat occur. These pulses vary in amplitude, but their characteristic is that they are very well defined with respect to the signal background in areas where there are no beats.

Stage 7.- Detection of the maximums of the signal created in the previous step.

By generating a signal in which its maxima are very well defined with respect to the signal background in areas where there are no beats, it is possible to identify these maxima with greater precision, since this signal has very sharp peaks in the regions where the oscillations caused by heartbeat.

Stage 8.- Association of each maximum with a heartbeat.

In this last stage, a beat is associated with each one of the maximums identified in the previous stage, which correspond to the regions where the oscillations caused by the heart beat occur. In this way, an estimate of the heart rate can be obtained with the necessary precision.

Invention System

Another object of the present invention is a system that allows the process of the invention to be carried out.

The system of the invention, in its most general embodiment, comprises three juxtaposed stages (or blocks of components), which execute the process steps consecutively (Figure 1):

1. Vibration transduction stage, comprising a sensor or sensor array (A) capable of transducing a mechanical vibration to an electrical signal. 2. Analog conditioning stage, comprising:

• Means for converting the sensor signal to voltage (B).

• Means for performing both high-pass and low-pass analog filtering (D).

3. Digital computing stage and presentation / transmission of results comprising:

• Means to carry out an analog to digital conversion (E)

• Computing means (F) capable of carrying out the method of the invention from the digitized data.

• Means of presentation and transmission of digital signals (G)

Stage 1 collects the vibrations produced by the heartbeat and converts them to actionable electrical signals for the following stages.

Stage 2 is used to adapt the level of the signal obtained by the sensor to an analog-digital converter with a useful signal-to-noise ratio.

Stage 3 performs the digital conversion and processing of the BCG to obtain the reference points for determining the heart rate.

In a particular embodiment, the sensor (1) is a capacitive sensor. Even more particularly, the capacitive sensor is a coaxial cable with a dielectric with a piezoelectric effect to improve shielding against electromagnetic interference.

In another particular embodiment, the capacitive sensor is replaced by an assembly comprising an LED light source and a photodetector that collects the light emitted by said LED, so that when the LED light is transmitted by the means between them (affected by vibration of heart movement) is modulated by vibration. This signal modulated by the vibration presents a similar aspect to the BCG from which to extract the information of the appearance of the beats.

In another particular embodiment, the voltage conversion means comprise a load-voltage amplifier or also called a transimpedance amplifier.

In another preferred embodiment, the analog processing stage in turn comprises a voltage gain stage (C) for those environments with a high presence of electronic noise and electromagnetic interference. In an equivalent embodiment, the processor is physically separated from the rest of the stages and the system includes means capable of communication (data exchange) with the processor.

MODES OF CARRYING OUT THE INVENTION

The following examples and drawings are provided by way of example, and are not intended to be limiting of the present invention.

System of the invention:

In a first embodiment (Figure 1), the system of the invention consists of the following three juxtaposed stages:

• Stage 1: Vibration transduction stage, formed by a capacitive sensor (A),

• Stage 2: Analog conditioning stage, consisting of:

or a charging amplifier (B)

or a high-pass filter and low-pass filter lower and upper cutoff frequencies of 0.15 Hz and 25 Hz, respectively (D)

or a voltage gain stage (C).

• Stage 3: Digital computing and presentation / transmission stage consisting of:

or a signal digitizer of at least 10 bits resolution (E). or a processor that includes the indicated procedure for digital signal processing and extraction of the time between heartbeats (F).

or an LCD screen to present the results to the user (G)

o An electronic circuit with antenna for wireless transmission (bluetooth or Wifi) of the information. (G)

o A mobile phone, which is wirelessly connected to the above circuit to receive the information and present it to the user.

The capacitive sensor generates charge by piezoelectric effect (A) due to the vibration transmitted to it by the human body and by various intervening solid means (textiles, wood, metals) caused by the mechanical movement of the heart.

Said capacitive sensor consists of a thin and flexible sheet forming a capacity of plane-parallel sheets.

In this way, direct contact of the sensor with the body surface, for greater comfort of the subject.

This generated load is converted to voltage in a load amplifier (B) to which, later and still within the analog domain, a band-pass filter (D) with lower and upper cut-off frequencies of 0.15 Hz and 25 Hz, respectively.

Then, after digitizing the signal (E), the digital procedure stated above is applied to it in a computing unit (F). The result consisting of heart rate as a function of time is displayed on a screen or is transmitted wirelessly (G) to another device (for example, a mobile device) through the corresponding data transmission and presentation modules.

Tests carried out

To carry out the acquisition of signals (BCG formation), the sensor was placed under a mattress, imperceptibly, for monitoring the subject during sleep in order to use the heart rate and its variability to assess the quality of the sleep by estimating its phases, along with other indicators.

Test results

Accompanying in Figure 3 are the results obtained by carrying out the described invention, which shows the histogram of the difference between the heart rate obtained by a standard ECG technique and our proposed invention. From this figure a correlation of more than 70% is deduced between both signals admitting an experimental error of ± 4 beats per minute.

Once the invention has been sufficiently described, as well as several preferred embodiments, it should only be added that it is possible to make modifications in its constitution, materials used, in the choice of elements and sensors used to detect BCG and in the methods to identify the reference points of the signal corresponding to the mechanical activity occurring in the heart and the aortic root, without departing from the scope of the invention, defined in the following claims.

Claims

1.- Procedure to measure heart rate from a ballistocardiogram, which includes the following stages:

• Analog filtering, in high-pass and low-pass, of the analog balistocardiogram.

• Analog to digital conversion.

• Identification of the maximums and minimums of the digitized signal

• Creation of a discretized signal formed by impulses in the positions where there are maximums or minimums and whose amplitude is the difference between each consecutive maximum and minimum.

• Construction, through digital low-pass filtering, of two continuous signals, one with the positive points of the discrete signal generated in the previous step, and the other with the negative ones.

• Creation of a new signal product of the square of the two signals from the previous step.

• Identification of the maximums of the signal generated in the previous stage as the reference points that allow the identification of each heartbeat.

2.- Procedure according to the previous claim, characterized in that the width of the passing band of interest used in the analog filtering is between 0.1 and 40 Hz, preferably between 0.15 and 25 Hz.

3. Procedure, according to claims 1 or 2, characterized in that the bandwidth used for digital low-pass filtering is between 1 Hz and 25 Hz.

4 - System for measuring heart rate from a ballistocardiogram (BCG) comprising the following juxtaposed steps, which execute the steps of the procedure according to any of the preceding claims:

1. Vibration transduction stage, comprising a sensor or sensor array (A) capable of transducing a mechanical vibration to an electrical signal.

2. Analog conditioning stage, comprising:

• Means for converting the sensor signal to voltage.

• Means for performing both high-pass and low-pass analog filtering.

3. Digital computing stage and presentation / transmission of results comprising: • Means to carry out an analog-to-digital conversion

• Computing means capable of carrying out the procedure of the invention from the digitized data.

• Means of presentation and transmission of digital signals

5.- System according to the previous claim, in which the sensor used is a capacitive sensor.

6.- System according to the previous claim, in which the capacitive sensor is a coaxial cable with a dielectric with a piezoelectric effect to improve the shielding against electromagnetic interference.

7.- System according to claim 4, in which the sensor used which comprises an LED light source and a photodetector that collects the light emitted by said LED

8. System according to any of claims 4 to 7, wherein the voltage conversion means comprise a load-voltage amplifier. 9. System according to any of claims 4 to 8, wherein the analog processing stage in turn comprises a voltage gain stage.