WO2019091517A1

WO2019091517A1 - Method for determining blood pressure while taking a physiological parameter into consideration

Info

Publication number: WO2019091517A1
Application number: PCT/DE2018/100914
Authority: WO
Inventors: Sebastian Zaunseder; Alexander TRUMPP; Hagen Malberg; Martin Schmidt
Original assignee: Technische Universität Dresden
Priority date: 2017-11-13
Filing date: 2018-11-12
Publication date: 2019-05-16
Also published as: DE102017126551A1; DE102017126551B4

Abstract

The invention relates to a method for determining the blood pressure of an individual by using a camera to capture image series of at least one surface region of the skin of the individual, the method comprising the following steps: (a) determining, on the basis of one or more image series, photoplethysmographic signals and ballistocardiographic signals; (b) calculating a pressure intermediate value from photoplethysmographic signals and ballistocardiographic signals obtained in step (a); (c) determining the perfusion speed from photoplethysmographic signals obtained in step (a); and (d) calculating the blood pressure from the pressure intermediate value and the perfusion speed.

Description

description

Method for determining blood pressure taking into account a physiological parameter

The invention relates to a method for determining the blood pressure, taking into account a physiological parameter. So far, there is no possibility for the contactless determination of beat-to-beat blood pressure. Contacting the determination can be invasive, for example by means of catheter, or non-invasive. Non-invasive, the determination can be carried out by applanation tonometry, volume compensation method or based on the pulse transit time (PTT). While applanation tonometry and the volume compensation method are very expensive, PTT-based methods can be performed on the basis of relatively simple sensors. Electrocardiogram (ECG) and photoplethysmogram (PPG) are usually used to detect PTT. Based on reference measurements, typically by means of a humeral cuff, an individual calibration is carried out so that the blood pressure in the sequence can be determined from the PTT. PTT-based determinations are based on sensors which are attached to the body (see for example EP 0 498 885 Bl, WO 2013/109 188 AI and DE 20 2016 105 262 Ul). The detection of signals for determining the PTT by means of sensors attached to the body impede the wearer. Changes in the vascular tone also require a recalibration or ensure a larger error.

The system for optically monitoring blood pressure described in WO 2013/109188 has a first light source, which is an LED or a laser, a first sensor for receiving ballistocardiogram (BCG) signals of an individual and a first Detector for converting the ballistocardiogram signals into electronic signals. The first sensor is a fiber optic sensor, such as a wireband sensor, which is intended to receive signals from the back or head of the individual. The system further includes a second light source, which is an LED or a laser, a second sensor for receiving photoplethysmogram signals of the subject and a second detector for converting the photoplethysmogram signals into electronic signals. The second sensor has a sensor head which is attached to the finger of the individual. The system described in WO 2013/109 188 AI finally has a signal evaluation and display unit to determine time differences between predetermined parameters of the ballistocardiogram signals and predetermined points of the photoplethysmogram signals. These time differences are then used to determine outcomes that should reflect the individual's blood pressure. This requires several calibration constants and correction factors to describe the relationship between BCG, PPG and blood pressure, while taking into account the optical, biomechanical and physiological factors. This achieves an accuracy which is said to be "better than 5%" in comparison to oscillometric blood pressure measuring devices, so that the system described in WO 2013/109 188 A1 has all the disadvantages of the state of the art: firstly, the individual who is the first sensor, the second sensor or both must be obstructed, and second, continuous calibration is required when the vascular tone changes.

CA 2 952 485 A1 discloses a method for measuring arterial and venous blood pulse waveforms of a patient using the PPG. US 2016/0278 644 A1 discloses a computer-implemented method for determining the PTT from optical data obtained from a patient.

The object of the invention is to eliminate the disadvantages of the prior art. In particular, a method for determining the blood pressure, taking into account a physiological parameter is to be specified, which operates without contact, requires no continuous calibration and has improved accuracy. This object is solved by the features of claim 1. Advantageous embodiments of the invention will become apparent from the features of the dependent claims. According to the invention, there is provided a method of determining the blood pressure of an individual using a camera to acquire image sequences of at least a surface area of the individual's skin, the method comprising the steps of:

(a) determining, based on one or more image sequences, photoplethysmographic signals and ballistocardiographic signals;

(b) calculating a pressure intermediate value from photoplethymographic signals and ballistocardiographic signals obtained in step (a);

(c) determining the perfusion rate from photoplethymographic signals obtained in step (a); and (d) calculating the blood pressure from the intermediate pressure value and the perfusion rate.

The determination of the perfusion rate can be carried out by known methods. However, it may also be carried out by the method described below for determining a physiological parameter.

Accordingly, a method is provided for determining a physiological parameter of an individual using a camera that captures one or more image sequences of one or more surface areas of the individual's skin. The physiological parameter is also referred to below as the perfusion rate. Based on the image sequence (s):

at least one base point P is determined, which lies in one of the surface regions; n distance points M _n spaced from the base point P by distance d _n , where n is an integer greater than or equal to 1, are determined; photoplethysmographic signals PPGp are determined at the base point P and photoplethysmographic signals PPG _{MII are} determined at the distance points M _n ;

a predetermined characteristic is respectively determined in the photoplethysmographic signals PPGp and PPG _MII and the time interval t _n between the occurrence of this feature in the photoplethysmographic signals PPGp and PPG _{MII is} determined;

for each distance point M _n, a propagation velocity v _n is calculated from the distance d _n and the time interval At _n ; and

the perfusion velocity is calculated on the basis of the propagation velocity v _n calculated for each of the distance points M _n and.

The method described above for determining a physiological parameter of an individual is also referred to below as the first method. The inventors have found that the physiological parameter is a value that reflects vascular tone. In particular, the inventors have found that the physiological parameter correlates with vascular tone, with a change in vascular tone and / or with vessel diameter. It has therefore been found that the physiological parameter can be used as a size for vascular tone. The physiological parameter is calculated on the basis of the propagation velocity v _n . It is therefore called the perfusion rate.

The first method is a non-contact method due to the use of a camera. There is no need for a device, in particular a sensor that touches the individual. Rather, it is sufficient to record image sequences of one or more surface areas of the skin of the individual, for example a patient. Any portion of the surface of the individual's skin where photoplethysmographic signals can be determined may be a surface area that can be used to determine the physiological parameter. Particularly suitable are in particular sections of the skin which have a homogeneous light intensity. However, other sections of the skin may also be used as the surface area. For example, temporal and spatial averaging techniques may be used to improve signal quality without distorting the shape of the resulting photoplethysmograms. The surface area may be, for example, the face; one or more areas of the face, such as the forehead, one or both cheeks; one or more limbs, such as one or both arms, or one or both legs, wherein the location of the surface area (s) is not limited as long as they are surface areas of the skin of the individual. Preferably, one of the surface areas is on the forehead of the individual. The first method can be carried out as a continuous process. In this way, values for the perfusion rate are obtained continuously, ie for each heartbeat. The surface area is therefore preferably on the forehead, because this - in comparison to other surface areas of the face - has a section with only a slight curvature. Therefore, this section represents a range of homogeneous light intensity in the image sequences. In addition to the forehead, therefore, each of the cheeks is a suitable area, because the cheeks have such sections. The surface area may therefore alternatively lie on one of the cheeks of the individual. The one or more surface areas can be specified by the user of the first method. It may alternatively be provided that the one or more predetermined surface areas in the image sequence are identified by software in an automated manner.

When determining the surface area or areas, previously known anatomical features can be taken into account. Such features may be features that include all or a group of individuals. Alternatively or additionally these features are individual traits that the individual has for whom the perfusion rate is to be determined.

The camera may be a conventional camera, such as a CCD camera. The camera should allow the recording of image sequences. An image sequence is understood to mean a sequence of images, for example a video. The images can be a sequence of images with, for example, 12 frames per second. Pictures taken by the camera will be referred to as shots or camera shots below. The camera should be focused on the specified surface area (s). It can be provided that the focusing is carried out by means of software in an automated manner. In the image sequences obtained or obtained by means of the camera, the photoplethysmographic signals are included. The photoplethysmographic signals are a blood volume pulse signal which essentially reflects the arrival of the pulse wave in the periphery, ie the blood vessels in the skin. By means of software, the photoplethysmographic signals can be determined from the image sequence. The first method according to the invention can comprise the acquisition of image sequences by means of a camera. It can be provided that in a surface area a minimum distance d _m i _n a maximum distance d _max or both of one of the base points P are given. That is not necessary. A minimum distance d _m i _n may be expedient for the time interval t _{n to be} greater than the temporal resolution of the camera and / or the triggering achieved by signal processing methods, for example using spatially continuous averaging. A maximum distance d _max may be expedient to ensure that there is a functional relationship in the perfusion between the point P and the associated distance points M _n or that such a connection is at least probable. It can further be provided that at least one of the distances d _{n is} shorter or longer than another of the distances d _n and / or that at least one of the distances d _n , starting from point P, at an angle to another of the distances d _n runs. Preferably, two or more than one base point P is determined. Separate distance points M _n are then determined for each of the base points. Thus, for each base point P, propagation velocities v _n can be determined. In this way, a high number of propagation velocities can be determined so that a statistical evaluation of the determined propagation velocities is possible. Each base point P should have one or more distance points M _n . The base points and their associated distance points M _n should be in the same surface area. Different base points can be in different surface areas. If there are several basis points in a surface area, the base points should be spaced apart from each other. The distance points M _{n of} different base points can overlap, as it is only important to know the length of the distance d _n between a base point and a distance point. A distance point M _n may itself be a base point. For example, a distance point M _{n may} belong to a base point P, so that the distance d _n between the distance point M _n to a base point Pi is known. This distance point M _n , in turn, may itself be a base point P ₂ , to which the base point Pi is then assigned as the distance point M _n , but which may have further distance points M _n . It may thus be provided to use multiple base points P to calculate the perfusion rate. For this purpose, the propagation velocities obtained for each base point P or the perfusion rates calculated for each base point P can be combined into a total perfusion velocity. To determine the total perfusion rate, a statistical evaluation can be made. For example, from a frequency distribution using the frequency of propagation velocities obtained for each base point P or a superposition of the frequency distributions obtained for each base point P, the overall perfusion velocity can be determined.

The maximum number of base points is determined only by the number of pixels of the images in a sequence of images. For example, if the images are 200 by 200 pixels, the maximum number of base points is 40,000. The minimum number of base points is 1. Preferably, the minimum number of base points is greater or equal to 2, 3, 4 or 5. More preferably, as many basis points are used as possible. The expression "as many basis points as possible" describes the circumstance that as far as possible all points of the surface portion of the skin depicted in the image sequence are used as base points, provided that photoplethysmographic signals can be determined from the image sequences at these points can be determined at one point photoplethysmographic signals can be determined by software.

It can be provided that the position of the base point or points P and the position of the distance point or points M _{n are determined} by the user or by means of software in an automated manner. It can be provided that the distance point or points M _n are determined so that the distance d _{n is} equal to or greater than the minimum distance dmin and equal to or smaller than the maximum distance d _max . Preferably, n is greater than or equal to 2, 3, 4 or 5. It can be provided that n is selected so that all points on the skin of the individual on which a photoplethysmogram can be determined are distance points M _n . In this case, all of these distance points M _n or a part thereof may be base points. ,

The distance d _n is the distance between a base point P and a distance point M _n associated with this base point. This route lies in the image plane of the images of the image sequence. It is thus the surface distance between the two points. According to the findings of the inventors, in particular in the case of short distances, this surface distance can be used as a measure of the distance that covers a pulse wave in a blood vessel that could be below the surface area of the skin. The spread of the blood volume pulse within a blood vessel provides for altered light absorption of the surface area of the skin. Short distances are advantageous because the likelihood that the blood volume pulse will actually travel through one and the same blood vessel at this distance is higher than at long distances. This propagation of the blood volume pulse can be detected spatially (spatial), ie at the points P and M _n , as well as temporally (temporally), ie on the basis of the temporal sequence of the images of the image sequence. The first method is thus based on a spatio-temporal analysis of the image sequence. There- at at the base point or points P photoplethysmographic signals PPGp and at the distance points M _n each photoplethysmographic signals PPG _MII determined. This determination of the signals PPGp and PPGvi _n is preferably carried out by means of software in an automated manner.

By determining the photoplethysmographic signals PPGp and the photoplethysmographic signals PPGvi _n , photoplethysmograms are obtained at the base point (s) P and at each of the distance points M _n . These photoplethysmograms are preferably true-to-life photoplethysmograms, the term "dimensionally stable" being understood to mean a photoplethysmogram having the morphological fluctuations contained in a photoplethysmogram recorded by means of conventional contact-type sensor technology be performed by the technique known per se methods of camera-based photoplethysmography.

The photoplethysmogram at the base point or points P and the photoplethysmogram obtained at each of the distance points M _n are expediently synchronous with one another.

Each of the photoplethysmograms has features such as minima, maxima, inflection points or the phase in the frequency domain. Because of the heartbeat, the photoplethysmograms have a regularly repeating sequence of a pattern that possesses these features. The features of the photoplethysmographic signal PPGp can also be found in the photoplethysmographic signals PPG _MII , but at least at some points these features will occur at a time interval At _n . For example, a minimum in the photoplethysmographic signals PPGp determined at the base point will occur at a time interval At _n , ie sooner or later, in the photoplethysmographic signals PPGvi _n . This time interval At _n , can be used in accordance with the invention in order to determine a propagation velocity v _n with the length of the distance d _n. agree with which the feature E has apparently moved from the base point P to a distance point M _n . Alternatively or additionally, the distance between different features or the distance of the complete signal forms can be used.

However, the propagation velocity so determined describes only an apparent movement, namely a movement on the surface on a straight line, but not the progression of the pulse wave through an actual blood vessel. For this reason, it is expedient to determine photoplethysmographic signals PPGvi _n at several distance points M _n . From this distance for each point M _n obtained propagation velocities v _n can then be calculated, a value that reflects the peripheral pulse wave velocity or corresponds in the best case. This value is referred to as the perfusion rate in the present invention. The peripheral pulse wave velocity, in turn, is a measure of vascular tone and peripheral resistance. It can be provided that the propagation velocities V _n are evaluated by means of stochastic methods, for example by producing a frequency distribution. For example, it may be provided to assign the propagation velocities v _{n to} predetermined value ranges and to use the range of values with the greatest frequency of propagation velocities v _n as the perfusion velocity and thus as a measure of the vascular tone. For a statistical evaluation, it is advantageous to determine the highest possible number of propagation velocities v _n . Therefore, it is preferable to use base points P, to each of which several distance points belong. This determination of the features in the signals PPGp and PPGvi _n is preferably carried out by means of software in an automated manner. The calculation of the propagation velocities v _n and the perfusion velocity is preferably carried out by means of software in an automated manner. The evaluation of the propagation velocities v _n by means of statistical methods is preferably carried out by means of software in an automated manner.

Calibration is not required for the determination of the perfusion rate provided according to the invention. The perfusion rate according to the invention It correlates with the vascular tone as described above. The vascular tone in turn correlates, for example, with the diastolic blood pressure. Thus, a method for determining the diastolic blood pressure for calibrating the first method according to the invention can be used. Alternatively, a calibration may be performed in which the initially determined value is used as the perfusion rate for that purpose, and subsequently changes in the perfusion rate relative to the initial value may be evaluated. A method that can also be used to determine diastolic blood pressure is described below.

In accordance with the invention, there is further provided a method for determining the blood pressure of an individual using a camera to capture image sequences of at least a surface area of the individual's skin. The method comprises the steps:

(c) determining the perfusion rate from the photoplethymographic signals obtained in step (a) by means of the first method; and

(d) calculating the blood pressure from the intermediate pressure value and the perfusion rate.

The method according to the invention for determining the blood pressure of an individual is also referred to below as the second method.

The second method is a non-contact method due to the use of a camera. There is no device, in particular no sensor required, the Individual touches. Rather, it is sufficient to record image sequences of a surface area of the skin of the individual, for example a patient. The face area may be, for example, the face or an area of the face. The second method can be carried out as a continuous process. If the first method is used in step (c), the first method is also carried out continuously.

It can be provided that step (a) of the second method comprises the following substeps:

(al) identifying at least one PPG region located in a surface area of the individual's skin and in which photoplethysmographic signals are determined; and (a2) identifying at least one BCG region located in a surface region of the individual's skin and in which ballistocardiographic signals are determined.

The BCG region (s) identified in step (a2) may be in the same or the same surface regions as the one or more PPG regions identified in step (a1). However, the BCG region (s) may also be located in one or more other surface regions than the PPG regions.

Any portion of the skin surface of the subject on which photoplethysmographic signals and ballistocardiographic signals can be determined may be a surface area that may be used to determine blood pressure in accordance with the present invention. In determining the surface area or areas, previously known anatomical features can be taken into account. Such features may be features that include all or a group of individuals. Alternatively, or in addition, these features may be individual traits exhibited by the individual for whom the perfusion rate is to be determined. The surface area may be, for example, the face or one or more extremities, such as one or both arms, or one or both legs, wherein the location of the surface area (s) is not limited as long as they are surface areas of the skin of the individual. The surface area is preferably the face of the individual. The surface area can be specified by the user of the second method. The term "PPG region" refers to a region in the surface region of the second process in which photoplethysmographic signals are determined from the image sequence to obtain a photoplethysmogram. The term "BCG region" denotes a region located in the surface region of the second process. in which ballistocardiographic signals are determined from the image sequence to obtain a ballistocardiogram.

It may be provided that more than one PPG region is identified in the surface region. If the surface area is the individual's face, then a PPG area on the individual's forehead and / or a PPG area may lie on one of the two cheeks and / or a PPG area and / or on the other cheek , For example, a first PPG region may be on the individual's forehead, a second PPG region on one cheek of the individual, and a third PPG region on the other cheek of the individual. Photoplethysmographic signals can be determined in each of the PPG regions. It may be provided that the one or more surface areas provided in the first method are PPG areas of the second method. This or these PPG regions is preferably the forehead of the individual. In other words, the PPG regions can be used to determine photoplethysmographic signals, which are then used according to the first method of the invention to determine an individual's physiological parameter called perfusion rate, and those according to the second method of the invention can be used to determine the blood pressure of the same individual. The PPG region (s) are therefore preferably on the forehead and / or on the cheeks because, in comparison with other regions of the face, they each have a section with only slight curvature. This section therefore represents a range of homogeneous light intensity in the image sequences. The one or more PPG regions can be specified by the user of the second method. Alternatively, it may be provided that the PPG region (s) in the image sequence are identified by means of software in an automated manner.

It may be provided that more than one BCG area is identified in the surface area. Advantageously, any portion of the skin that has an edge or at least one point that stands out from its surroundings can be used as the BCG region. If the surface area is around the face of the individual, then a BCG area on the chin and / or a BCG area on the hairline and / or a BCG area on either eye and / or a BCG area on the hairline another eye lie. For example, a first BCG region may be at the hairline of the individual, a second BCG region may be on one of the individual's eyes, a third BCG region may be on the other eye of the individual, and a fourth BCG region may be on the individual's chin. Ballistocardiographic signals can be determined in each of the BCG areas.

The BCG region (s) are therefore preferably located on the hairline, on the chin or on the eyes, because there - in comparison to the PPG regions of the face - in each case a section with greater curvature and / or anatomically with properties that differ from its Environment take off, is present. These sections therefore represent a range of inhomogeneous light intensity in the image sequences. The BCG region (s) may be specified by the user of the second method. It may alternatively be provided that the BCG region (s) in the image sequence are identified by means of software in an automated manner. The BCG (s) should not align with the PPG (s).

This determination of the photoplethysmographic signals and the ballistocardiographic signals from an image sequence is preferably carried out by means of software in automated way. Photoplethysmographic signals are obtained for each PPG region and ballistocardiographic signals for each BCG region. The photoplethysmographic signals of all PPG regions can each separately form a photoplethysmogram for each PPG region, or alternatively they can be summed up to form a PPG region-spanning photoplethysmogram PPG. The photoplethysmograms are preferably conformal photoplethysmograms. The determination of the photoplethysmographic signals from an image sequence can be carried out according to the methods of camera-based photoplethysmography known from the prior art. The ballistocardiographic signals of all BCG areas can be used to individually form a ballistocardiogram for each BCG area or, alternatively, to summate a BCG area-spanning ballistocardiogram. The ballistocardiograms are preferably dimensionally true ballistocardiograms, the term "faithful to form" describing a ballistocardiogram which exhibits the morphological fluctuations contained in a ballistocardiogram recorded by means of conventionally contact-based sensor technology and the evaluation of the ballistocardiographic signals from a sequence of images the methods of camera-based ballistocardiography known per se from the prior art are carried out.

The one or more ballistocardiograms and the one or more photoplethysmograms are expediently synchronous with one another.

Step (b) of the second method provides for calculating a pressure intermediate value from photoplethysmographic signals and ballistocardiographic signals obtained in step (a). The term "intermediate pressure value" may in this case describe a value for pulse pressure, which is referred to below as the intermediate pulse pressure value, or a value for the blood pressure, which is referred to below as the blood pressure intermediate value. may indicate a value for the diastolic blood pressure, a value for the systolic blood pressure, or both. The term "intermediate value" is used because of this value is used to calculate the blood pressure according to the second method and therefore does not represent the process result in itself.

To calculate the pressure intermediate value, the PPG cross-sectional photoplethysmogram and the BCG cross-sectional ballistocardiogram are advantageously used. As explained above in the context of the first method, each photoplethysmogram has features, such as minima, maxima and inflection points, and a regularly repeating sequence of a pattern having these features. Similarly, each ballistocardiogram has features such as minimum, maximum, and inflection points, and a regularly repeating sequence of a pattern that possesses these features. It can now be provided that in step (b) the pressure intermediate value is calculated from the pulse transit time, wherein the time interval ΔΪ _ΒΤ between a characteristic of the ballistocardiographic signals and a feature of the photoplethysmographic signals is used to calculate the pulse transit time. For example, in the ballistocardiogram, a maximum, preferably an absolute maximum, may be located in the repeating pattern and in the photoplethysmogram a minimum, such as an absolute minimum, in the repeating pattern. The localization of features in a photoplethysmogram and in a ballistocardiogram can be done by software in an automated way. The feature localized in the ballistocardiogram may be used to characterize a time t _B at which the heartbeat has triggered a pulse wave. The feature localized in the photoplethysmogram can be used to characterize a time t _P at which the pulse wave has arrived in the PPG region (s). Thus, the time t _{P is} after the time t _B. The time difference between t _B and tp is the time interval ΔΪ _ΒΤ , which corresponds to the pulse transit time. From the pulse transit time, the blood pressure, the pulse pressure or both can be calculated in a manner known per se, for example the systole and / or the diastole of the blood pressure. In the present invention, the pulse pressure thus calculated from the pulse transit time is referred to as the intermediate pulse pressure value, and the blood pressure is referred to as the intermediate blood pressure value. Thieves- Calculation of the time interval Δΐ _Β τ and the pulse transit time can be carried out by means of software in an automated manner. Likewise, the calculation of the intermediate pressure value can be carried out by means of software in an automated manner.

As an alternative or in addition to the use of features of the ballistocardiographic signals and features of the photoplethysmographic signals, it is also possible to use the distance between different features or the spacing of the complete signal forms.

It has been found that the blood pressure calculated from the pulse transit time can be further approximated to the invasively determined values if the vessel tone is included in the calculation of the blood pressure. The determination of a value which correlates with the vascular tone - the physiological parameter referred to as the perfusion rate of the present invention - is carried out by means of the first method. For this purpose it can be provided that in step (c) one, several or all of the PPG areas are used. The blood pressure can then be calculated from the blood pressure intermediate values and the perfusion rates. This can be done by software in an automated way. The blood pressure calculated in this way can be a beat-to-beat determination of systole and diastole. The blood pressure can be continuous, d. h for each heartbeat, to be determined. This requires a continuous determination of the physiological parameter by means of the first method. Steps (b) and (c) should be performed synchronously.

For the calculation of the blood pressure from the intermediate blood pressure value and the perfusion rate, for example, based on the method described by Ding et al. "Continuous Cuffless Blood Pressure Estimation Using Pulse Transit Time and Photometry Intensity Ratio," IEEE Trans. Biomed. Eng., Vol. 63, no. 5, pp. 964-972, 2016, developed mathematical model to proceed in the following manner: (i) The calculation of the diastolic blood pressure DBP can be made according to formula 1:

DBP = DBP _{b (1)}

° cbPFG '

(ii) The systolic blood pressure SBP can be calculated according to formula 2:

SBP = DBP _b + PP _b (^ ² (2)

° cbPFG ° cbPTT) 'The formula symbols used in formulas 1 and 2 have the following meaning: cb: camera-based determination;

DBP: diastolic blood pressure;

DBP: diastolic blood pressure at the time of calibration;

SBP: systolic blood pressure;

SBP _b : systolic blood pressure at the time of calibration;

PFG: perfusion rate determined according to the first method;

PFG _b : perfusion rate, determined according to the first method, at the time of calibration;

PP _b : blood pressure amplitude at the time of calibration;

PTT: Pulse transit time;

PTT _b : Pulse transit time at the time of calibration.

For the calculation of the blood pressure provided according to the invention, a calibration can be provided. In this way, absolute values can be achieved, in particular for the values DBP _b and PP _b used in formulas 1 and 2. A one-time calibration is sufficient. Calibration may be performed using a pressure cuff or other method of determining diastolic blood pressure. The calibration is preferably carried out at the beginning of the recording of an image sequence by means of the camera. This time is the "time of calibration" used in formulas 1 and 2. By means of the calibration, the values DBP _b and PP _b used in formulas 1 and 2 are obtained are the only values which are not necessarily determined from the image sequences by means of the methods according to the invention, but in another way.

The camera may be a conventional camera, such as a CCD camera. The camera should allow the recording of image sequences. An image sequence is understood to mean a sequence of images, for example a video. The images can be a sequence of images with, for example, 12 frames per second. Images taken by the camera are also referred to as shots or camera shots. The camera should be focused on the given surface area (s). It can be provided that the focusing is carried out by means of software in an automated manner. The sequences of images obtained or obtained by means of the camera include the photoplethymographic signals and the ballistocardiographic signals. The photoplethysmographic signals are a blood volume pulse signal which is decisive for the arrival of the pulse wave in the periphery, i. H. the blood vessels in the skin, reflects. By means of software, the photoplethysmographic signals can be determined from the image sequence. The ballistocardiographic signals are a movement signal that essentially reflects the ejection of the blood into the blood vessels. By means of software, the photoplethysmographic signals and the ballistocardiographic signals can be determined from the image sequence. The second method according to the invention can include the acquisition of image sequences by means of a camera.

The same camera can be used to perform the first and second procedures. The software mentioned in connection with the first method and the second method may be a data processing program which is executed on a data processing device, for example a computer. For this purpose, the images obtained by the camera as a sequence of images can be transmitted as digital images to the data processing device and processed there by means of the software. The first and second methods allow the determination of photoplethysmograms (PPG) and ballistocardiograms (BCG) from camera images. Shaped photoplethysmograms and conformal ballistocardiograms can be determined. The pulse transit time (PTT) can be calculated by means of the photoplethysmograms and ballistocardiograms determined from the camera recordings. The first method, however, allows a contactless, ie camera-based determination of the perfusion rate. The second method enables a contactless, ie camera-based determination of the pulse transit time. The second method also allows determination of blood pressure using the pulse transit time and the perfusion rate correlated with vascular tone. The determined blood pressure has a lower error compared to the prior art. In addition, only a single calibration, if any, is required.

By means of the first method, the invention makes possible a non-contact determination of a physiological parameter correlating to vascular tone. The invention makes possible, by means of the second method, a contactless determination of the beat-to-beat blood pressure. The invention allows a more accurate determination of the blood pressure, in particular the beat-to-beat blood pressure, due to the consideration of the vascular tone. The determination of photoplethymograms and / or ballistocardiograms provided according to the invention can also be used to determine further physiological properties, for example for hemodynamic characterization.

The invention uses the basic principle of camera-based photoplethysmography and / or camera-based ballistocardiography. The invention is based on the finding that the heart activity of an individual in videos recorded by means of a camera is reflected in a multimodal manner, ie on the basis of the blood volume pulse and on the basis of ballistocardiographic effects, and that the heart activity is spatio-temporal, ie spatially and temporally resolved. can be detected. Using camera-based photoplethysmograms and camera-based ballistocardiograms, the pulse transit time (PTT) can be determined. By analyzing the spatio-temporal characteristic of blood circulation, the peripheral pulse wave lengeschwindigkeit - which is referred to here as Perfusionsgeschwindigkeit to allow a demarcation of the pulse wave velocity in the sense of 1 / PTT - be determined as a measure of the vascular tone. Pulse transit time and perfusion rate can be used to determine blood pressure. It may be provided a one-time calibration. The calibration can be done at the beginning of the camera recording. Calibration may be accomplished using a pressure cuff or otherwise determined blood pressure.

The invention allows a contactless determination of blood pressure. In this case, multimodal information can be determined by means of a camera. This information can be photoplethysmograms, in particular form-consistent photoplethysmograms, and ballistocardiograms, in particular true-to-size ballistocardiograms. The multimodal information obtained by the camera can be merged. The supplied multimodal information can be used to determine the pulse transit time. Furthermore, spatio-temporal blood flow, which is reflected in the perfusion rate, can be used to determine vascular tone.

The invention will be explained in more detail below with reference to exemplary embodiments, which are not intended to limit the invention, with reference to the drawings. Show

1 shows a schematic representation of a picture of a face obtained by means of a camera, in which exemplary PPG areas and BCG areas are identified;

FIG. 2 shows a further schematic representation of the image of a face already shown in FIG. 1, in which an exemplary surface area for the determination of the perfusion rate is identified; FIG.

3 is a diagram showing a BCG region-spanning ballistocardiogram for the exemplary BCG regions shown in FIG. shows a chronic PPG cross-sectional photoplethogram for the PPG regions shown in FIG. 1;

Fig. 4 is a diagram showing synchronous photoplethysmograms for the surface area shown in Fig. 2; and

Fig. 5 is a graph showing a frequency distribution of the propagation velocities v _n calculated from the photoplethysmograms shown in Fig. 4.

The abbreviations used have the following meaning:

BT blood pressure

GT vascular tone

BCG Balli stokardi ogramm

cbBCG he came from asi ers B al 1 i stokar di ogramm

cbPPG kamerab asi erte s Photopl ethy smogramm

M _n distance point

P base point

PFG perfusion rate

PPG Photopl ethy smogram

PTT Pulse transit time

ROI Region of Interest

The image shown in Fig. 1 was obtained with a camera. The picture shows the head of an individual, ie a human, with its face 1. The face 1 is the surface area according to the second method. By means of software, the following regions have been identified in the face 1: a first PPG region (ROIl _C bppG) lying on the forehead of the individual; a second PPG region (ROI2 _C ppG) located on a cheek of the individual;

- a third PPG region (R0I3 _C ppG) located on the other cheek of the individual;

a first BCG region (ROII _B CG) located at the hairline of the individual;

a second BCG region (ROI2 _B CG) which lies on the one eye of the individual and whose pupil comprises;

a third BCG region (ROI3 _B CG) lying on the other eye of the individual and comprising its pupil; and - a fourth BCG region (ROI4 _B CG) located on the chin of the individual.

The first BCG region (ROII _B CG) and the fourth BCG region (ROI4 _B CG) may lie on the symmetry axis of the face. On the forehead is also the surface area according to the first method. This surface area is also referred to below as the GT area. In the GT area, a base point P was determined by means of software and three distance points. In this example, n is thus equal to 3. Between the base point P and the distance Mi, the straight line di lies, between the base point P and the distance M ₂ lies the straight line d ₂ and between base point P and the distance M ₃ is even Route d ₃ . The distances di, d ₂ and d ₃ have different lengths. The distance d ₂ runs at an angle to the distance di. The route d ₃ runs at an angle to the distances di and d ₂ . The GT area and the first PPG area (ROIl _C bppG) are both on the forehead of the individual. They can overlap or overlap. A video was taken from the face 1 of the individual by means of a camera, of which the image shown in FIGS. 1 and 2 is the temporally first image. Using the software, the three PPG regions were evaluated by generating photochemistry signals. The obtained photoplethysmographic signals of all three PPG regions were combined into a PPG-spanning photoplethysmogram 2 (see FIG. 3). For this purpose, sum signals of the photoplethysmographic signals obtained on all three PPG regions were formed. The software also evaluated the four BCG areas by generating ballistocardiographic signals. The obtained ballistocardiographic signals of all four BCG regions, which are synchronous with the photoplethysmographic signals of the three PPG regions, were combined to form a BCG region-spanning ballistocardiogram 3 (see FIG. 3). To this end, sum signals of the ballistocardiographic signals obtained at all four BCG areas were formed.

The PPG cross-sectional photoplethysmogram 2 has now been used to determine a recurring characteristic - due to the heartbeat - by means of the software. In Fig. 3, this feature is an absolute minimum 4. The BCG cross-sectional photoplethysmogram 3 was also used to determine a recurrent feature by software, due to the heartbeat. In Fig. 3, this feature is an absolute maximum 5. The time interval Δΐ _Β τ between the occurrence of the absolute maximum 5 and the absolute minimum 4 correlates with the pulse transit time (PTT). From the time interval ΔΪ _ΒΤ , the pulse transit time (PTT) can thus be calculated by means of the software. The calculated pulse transit time (PTT) can in turn, according to known methods, be used to calculate the blood pressure of the individual. According to the invention, however, an inclusion of the perfusion rate in the calculation of the blood pressure is provided. The blood pressure calculated from the pulse transit time (PTT) is therefore used in the present invention only as a pressure intermediate value. To determine the perfusion velocity, the video was evaluated at points P and Mi, M ₂ and M ₃ generating photoplethysmographic signals for each of these points. These photoplethysmographic signals were recorded in synchronism with the ballistocardiographic signals from the BCG regions and the photoplethysmographic signals from the PPG regions. Some or all of these photoplethysmographic signals may be photoplethymographic signals that have already been generated in association with a PPG region. From the photoplethysmographic signals obtained for the points P and Mi, M ₂ and M ₃ , photoplethysmograms were formed, which are shown in FIG. 4. The photoplethysmogram obtained for the base point P is the photoplethysmogram PMGp, the photoplethysmograms obtained for the distances Mi, M ₂ and M ₃ are designated as photoplethysmogram PPG _MI , photoplethysmogram PPG _M2 and photoplethysmogram PPGvo, respectively. In FIG. 4, the ordinate in each case shows the amplitude of the photoplethysmograms, the photoplethysmograms being arranged one above the other in order to simplify the illustration.

A feature has now been determined in the photoplethysmogram PPGp. This may be, for example, a minimum 6. This minimum 6 is in Fig. 4 at the intersection of abscissa and ordinate. The time was then determined when the same feature 6 'appeared in the photoplethysmograms PPG _MI , PPG _M2 and PPGvo. It can be seen in FIG. 4 that the same feature 6 'appeared in the photoplethysmograms PPG _MI , PPG _VO and PPGvo at a time interval At _n at a time interval Δti in the photoplethysmogram PPG _MI , a time interval Δt ₂ in photoplethysmograph PPGvo and a time interval at ₃ in the photoplethysmograph PPGvo- _n from the thus determined time intervals at, and the corresponding distances d _{n is} the propagation velocities v _n were then calculated. The calculated propagation velocities v _n have now been subjected to a statistical evaluation. For this purpose, the propagation velocities v _{n were} assigned to speed ranges, as a result of which the frequency distribution shown in FIG. 5 was obtained. The most common propagation velocity or the region with the most common propagation velocity was used as perfusion velocity.

The blood pressure was then calculated from the intermediate blood pressure value and the perfusion rate using formulas 1 and 2. Apart from the specification of the face, the determination of the blood pressure was made by means of a software which was executed on the computer. The blood pressure can be determined continuously by means of the ancestor according to the invention, for example for each heartbeat, for a period of time determined by the start and end of the sequence of images.

Claims

claims

A method of determining the blood pressure of an individual using a camera to acquire image sequences of at least a surface area of the individual's skin, the method comprising the steps of:

Determining, based on one or more image sequences, photoplethysmographic signals and ballistocardiographic signals;

Calculating a pressure intermediate value from photoplethymographic signals and ballistocardiographic signals obtained in step (a);

(c) determining the perfusion rate from photoplethysmographic signals obtained in step (a); and

A method according to claim 1, characterized in that the intermediate pressure value is a pulse pressure intermediate value or a blood pressure intermediate value.

A method according to claim 1 or claim 2, characterized in that step (a) comprises the following substeps:

(al) identifying at least one PPG region located in a surface area of the skin of the individual and in which photoplethysmographic signals are determined; and (a2) identifying at least one BCG region located in a surface area of the skin of the individual and in which ballistocardiographic signals are determined.

4. Method according to claim 3, characterized in that the surface area of the skin of the individual is its face and the at least one PPG area lies on the forehead and / or on at least one cheek.

5. The method according to claim 3 or claim 4, characterized in that the surface area of the skin of the individual is the face and the at least one BCG area lies on the chin, on the hairline and / or on at least one eye.

6. The method according to any one of the preceding claims, characterized in that in step (b) the pressure Z wischen value from the pulse transit time is calculated, wherein for calculating the pulse transit time of the time interval between a feature of the ballistocardiographic signals and a feature of the photoplethysmographic signals is used.

7. The method according to any one of the preceding claims, characterized in that one or more or all of the PPG regions are used to determine the perfusion rate in step (c).

8. The method according to any one of the preceding claims, characterized in that the method is carried out continuously.

9. The method according to any one of the preceding claims, characterized in that in step (c):

- At least one base point P is determined, which is located in one of the surface areas; determining n distance points M _n which are spaced from the base point P by a distance d _n , where n is an integer greater than or equal to 1;

photoplethysmographic signals PPGp are determined at the base point P and photoplethysmographic signals PPG _{MII are} determined at each of the distance points M _n ;

- In the photoplethysmographic signals PPGp and PPGvin each a predetermined feature is determined and the time interval t _n between the occurrence of this feature in the photoplethysmographic signals PPGp and PPGvin is determined;

for each of the distance points M _n, a propagation velocity v _n is calculated from the distance d _n and the time interval At _n ; and

the perfusion rate is calculated on the basis of the propagation velocity v _n calculated for each of the distance points M _n .

10. The method according to claim 9, characterized in that two or more base points P are determined and for each of the base points P n distance points M _{n are} determined.

11. A method according to claim 9 or claim 10, characterized in that at least one of the surface areas is located in the face or extremity of the individual.

12. The method according to any one of claims 9 to 11, characterized in that for each of the base points P, the number n of the distance points, which are spaced from this base point P by a distance d _n , greater than or equal to 2, 3, 4 or 5 , A method according to claim 12, characterized in that, based on the distances d _n emanating from one of the base point P, at least one of the distances d _{n is} shorter or longer than another of the distances d _n and / or that at least one of the distances d _n , starting from point P, is at an angle to another of the distances d _n .

Method according to claim 12 or claim 13, characterized in that the propagation velocities v _n are evaluated by means of stochastic methods.

Method according to one of Claims 12 to 14, characterized in that the propagation velocities v _{n are} assigned to value ranges and the value range with the greatest frequency is used as perfusion velocity.