CN114159091B - Heart sound propagation relation detection system based on wearable sensor array - Google Patents
Heart sound propagation relation detection system based on wearable sensor array Download PDFInfo
- Publication number
- CN114159091B CN114159091B CN202111548482.0A CN202111548482A CN114159091B CN 114159091 B CN114159091 B CN 114159091B CN 202111548482 A CN202111548482 A CN 202111548482A CN 114159091 B CN114159091 B CN 114159091B
- Authority
- CN
- China
- Prior art keywords
- heart sound
- image
- energy
- sensor
- circle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B7/00—Instruments for auscultation
- A61B7/02—Stethoscopes
- A61B7/04—Electric stethoscopes
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/332—Portable devices specially adapted therefor
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- Animal Behavior & Ethology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Acoustics & Sound (AREA)
- Surgery (AREA)
- Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Cardiology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
Abstract
The invention relates to a heart sound propagation relation detection system based on a wearable sensor array, and belongs to the technical field of biomedical signal detection. The system comprises an electrocardio sensor, a wearable sensor array, a signal acquisition card and a signal analysis and processing unit. The invention is based on the image registration principle, and the position of the sensor on the body surface of the human body corresponds to the skeleton structure of the human body. And combining the energy information of the heart sound signals of the human body with the image coordinates of the sensor to form three-dimensional data. And drawing a human heart sound signal energy relation diagram in the human skeleton image by utilizing a linear interpolation principle. The invention has high spatial resolution and high time resolution, and can synchronously analyze multichannel heart sound signals. The graph can intuitively show the energy relation between the heart sound signals of the human body and different positions of the body surface, and has important significance for researching the influence of the position of heart sound origin, organ tissues on heart sound transmission and attenuation change of heart sound energy.
Description
Technical Field
The invention belongs to the technical field of biomedical signal detection, and particularly relates to a heart sound propagation relation detection system based on a wearable sensor array.
Background
The heart is the source of mechanical vibrations. Under the synergistic effect of the electrocardiographic system, a series of mechanical vibrations, called heart sounds, are generated by the autonomous contraction and relaxation of the heart in order to maintain blood circulation.
Traditionally, doctors use stethoscopes to monitor the heart sounds of patients, but this approach is greatly affected by the subjective factors of the doctor. The invention of the electronic stethoscope brings technical innovation for detecting heart sounds, and the analysis and processing of heart sound signals gradually arouses the interest of researchers. In the application process of the existing heart sound sensor, heart sound signals of traditional heart sound auscultation areas of a human body are mainly collected, wherein the heart sound signals comprise a mitral valve area, a pulmonary valve area, an aortic valve second auscultation area and a tricuspid valve area, the heart sound signals outside the auscultation areas are less researched, and meanwhile research on relations among the areas is lacking. The traditional heart sound sensor has the defects of overlarge volume, low spatial resolution, incapability of being applied to a large number of body surfaces of a human body and objectively lacking in conditions for researching the propagation relationship of heart sounds of the human body. And heart sounds are generated by the heart of a human body in the working process, and sound sources of the heart sounds comprise heart valves, large blood vessels, cardiac muscles and blood. Due to the difference of the spatial positions of the sound sources, heart sound signals at different positions of the body surface of the human body have obvious differences. Currently, a heart sound propagation relationship detection system is lacking.
Disclosure of Invention
Aiming at the problems existing in the prior art, the invention designs the miniature heart sound sensor based on the Chinese patent CN110226944B (miniature heart sound sensor based on MEMS technology and application thereof), prepares a wearable sensor array and designs a heart sound propagation relation detection system. The invention regards the human heart as a multi-source signal generator and utilizes the wearable sensor array to collect human heart sound array signals. In the acquisition process, the invention provides high time resolution for the heart sound signal by utilizing high sampling frequency, and is beneficial to analyzing the propagation relationship of the heart sound signal which changes along with time. In the process of the propagation of heart sounds, energy can be attenuated to different degrees due to different media, and the heart sound energy propagation relationship diagram is drawn based on the energy information of heart sound array signals.
The invention researches the propagation relationship of heart sounds on the body surface from a brand-new angle and provides a brand-new heart sound research thought. The invention has high spatial resolution and high time resolution, can synchronously analyze multichannel heart sound signals, and lays a foundation for exploring heart sound signals of unknown areas of human bodies and relations among heart sound signals of all areas.
The technical scheme of the invention is as follows:
a heart sound propagation relation detection system based on a wearable sensor array comprises an electrocardio sensor, the wearable sensor array, a signal acquisition card and a signal analysis and processing unit in a computer end.
The electrocardiosignal is used for collecting a first lead electrocardiosignal of a human body and identifying the heart movement period of the human body.
The wearable sensor array mainly comprises a miniature heart sound sensor, a wearable waistcoat and an inflatable air bag, and the structure of the array is a network topological structure.
The wearable waistcoat is suitable for people of different sizes, and different network topological structures are arranged on the wearable waistcoat according to requirements and used for installing the miniature heart sound sensor; the miniature heart sound sensor is embedded in the inner side of the wearable waistcoat, and in order to improve the fitting degree of the sensor and the human body surface, an inflatable air bag is added in an interlayer of the wearable waistcoat, and the inflatable air bag provides balanced air pressure for each sensor.
The acquisition card is used for synchronously acquiring human heart sound array signals acquired by the wearable sensor array and electrocardiosignals acquired by the electrocardiosignal, and converting analog signals into digital signals and transmitting the digital signals to the computer end.
The signal analysis and processing unit comprises signal preprocessing, image registration and heart sound signal linear interpolation drawing.
The signal de-averaging is used for avoiding the influence of signal baseline drift on energy analysis; the signal band-pass filtering is used for carrying out Butterworth band-pass filtering and noise reduction on the heart sound signals and the electrocardiosignals, setting the passband of the heart sound signals to be 30-150 Hz and the passband of the electrocardiosignals to be 0.5-35 Hz, and carrying out bidirectional zero phase shift filtering on the heart sound signals and the electrocardiosignals to avoid the influence of phase offset on the research of propagation relations; wherein, the Butterworth band-pass filter can select 'coif 5' wavelet basis function, the number of layers of wavelet decomposition is selected as 7, and soft threshold is adopted for processing; the heart sound signal wavelet is used for reducing noise and further filtering noise in the heart sound signal.
The image registration is used for registering heart sound acquisition points and human skeleton images. In order to accurately map the spatial position of the heart sound acquisition point to the human skeleton structure, the spatial position of the miniature heart sound sensor is rigidly registered with the skeleton image, and the method specifically comprises the following steps: first, a chest image of a subject is photographed, and visual calibration points are added to the chest image and the bone image. And a visual label is added to the position of the miniature heart sound sensor in the chest image. Based on the chest image and the calibration points in the bones, the visual labels of the miniature heart sound sensor are subjected to rotation transformation, scaling transformation and translation transformation according to the rigid transformation principle, and the mapping of the miniature heart sound sensor network topology structure in the bone image is completed.
The heart sound signal linear interpolation drawing is used for drawing an energy propagation relation diagram of the heart sound signal of the human body. In order to draw the energy information of the heart sound signal on the body surface of the human body, the energy of the heart sound signal is calculated through the amplitude value of the heart sound signal. And drawing concentric circles by using each heart sound acquisition point as a circle center according to the position of the miniature heart sound sensor in the skeleton image, wherein the corresponding heart sound signal energy linearly decreases along with the increase of the radius of the concentric circles. The heart sound signal energy sequentially corresponds to red, yellow, green, cyan and blue from high to low, colors are filled into corresponding concentric circles, and an energy propagation relation diagram of the heart sound signal of the human body is drawn. The energy propagation relation diagram of the heart sound signals of the human body is analyzed correspondingly to the period of the electrocardiosignal, and the propagation trend of the heart sound signals can be analyzed from the angles of time and space. The invention has the beneficial effects that:
according to the invention, the topological structure of the sensor network can be designed autonomously according to the requirements, and the sensor network is used for acquiring heart sound array signals. By using the statistically independent heart sounds generated at different positions, the energy relationship between the heart sounds and the body surface can be studied. The invention has the characteristics of high spatial resolution and high time resolution, is easy to operate, has important significance for researching the influence of heart sound origin position and organ tissue on heart sound transmission and attenuation change of heart sound energy, and can provide a brand new thought for clinicians and scientific research and analysis of heart sound signals.
Drawings
FIG. 1 is a flow chart of the heart sound propagation relationship detection system of the present invention.
Fig. 2 is a sensor network topology in an embodiment of the invention.
Fig. 3 (a) to 3 (f) are energy relation diagrams of heart sounds on a body surface, wherein the background of the diagrams is a human skeleton image, and contour lines in the diagrams are areas with different energy signals; the numbers in the contour lines represent the energy values of the region, with the energy value being 0 at the lowest and 2 at the highest. Fig. 3 (a) to 3 (c) correspond to the early, middle and later heart sound energy relation diagrams of the first heart sound of the human body, respectively. Fig. 3 (d) to 3 (f) correspond to the heart sound energy relation diagrams of the early, middle and later stages of the second heart sound of the human body, respectively.
FIG. 3 (g) is a color schematic diagram corresponding to the energy value.
Detailed Description
The following describes the embodiments of the present invention further with reference to the drawings and technical schemes.
(1) Fig. 1 is a flowchart of the heart sound propagation relationship detection method of the present invention. The system comprises an electrocardio sensor, a wearable sensor array, a signal acquisition card and a signal analysis and processing unit. The electrocardiosignals are worn on the left wrist and the right wrist of the human body and are used for collecting first lead electrocardiosignals of the human body; the wearable sensor array mainly comprises a miniature heart sound sensor, a wearable waistcoat and an inflatable air bag, wherein the miniature heart sound sensor is a miniature heart sound sensor in a patent CN110226944B, and the sensor is small in size, has a high signal-to-noise ratio and is suitable for collecting heart sound signals of a human body. The number and the positions of the sensors in the detection process can be set independently according to the requirements.
The system is suitable for subjects of different sizes. In use, the subject wears the device and the sensor contacts the skin surface. The air bag is placed in the equipment interlayer, and is inflated appropriately according to the body shape of a wearer, so that the fitting degree of the sensor and the skin is adjusted. Analog signals acquired by the sensor are synchronously acquired by the multichannel acquisition card, input into calculation for preprocessing, and finally, a heart sound energy relation diagram of the human body is drawn. The sampling rate can be set according to the requirement by the acquisition card, and in order to study the propagation relation of the heart sound signals of the human body, the time resolution is improved by adopting a higher sampling rate, and the sampling rate is not lower than 10kHz through experimental tests.
(2) The sensor network topology is shown in fig. 2, in which circles are heart sound sensors. The topological structure of the sensor network can be designed according to the requirements and is used for collecting heart sound array signals of a human body. The invention corresponds the arrangement positions of the heart sound sensors with the skeleton structure of the human body and is used for showing the propagation relation of heart sound. In this embodiment, a rectangular network structure is adopted, the arrangement area includes the whole chest area, the distance between the sensors is about 2cm, and the number of the sensors is at least 80.
(3) Image registration
Based on the image registration principle, the invention adopts rigid transformation to convert the position of the sensor on the body surface to the position corresponding to the skeleton of the human body. Only the pixel position of the sensor in the image is changed in the conversion process, and the topological structure of the sensor is not changed.
Visual labels are added to the miniature heart sound sensor and the human skeleton calibration points, and the specific positions of the skeleton calibration points and the miniature heart sound sensor in the acquisition process are determined. The visual label adopted by the miniature heart sound sensor is a green circle, and the size of the visual label is consistent with that of the miniature heart sound sensor. The visual label adopted by the human skeleton calibration point is a red circle. After the acquisition is completed, the chest image of the tested person containing the visual tag is photographed. In the image processing process, in order to distinguish the calibration point and the miniature heart sound sensor, the color RGB space is converted into HSV space, and according to HSV parameter values corresponding to red and green, a red part and a green part are respectively extracted, so that images respectively containing the calibration point and the miniature heart sound sensor are obtained.
The image containing the calibration point and the micro heart sound sensor is subjected to binarization processing, hue and saturation information is eliminated, and each element in the image is represented by a brightness value of 0 (black) to 255 (white). The invention adopts a weighted average method to calculate the gray value of the image:
g＝0.2989*R+0.5870*G+0.1140*B (1)
where R, G, B is the red, green, blue component of the color image. 0.2989, 0.5870 and 0.1140 are weights of R, G, B, respectively, and g is a gray value of an image.
After binarization processing, a circle in the image is found based on a Circle Hough Transform (CHT). For a circle of radius r, center (a, b), it can be expressed as:
(x-a) 2 +(y-b) 2 ＝r 2 (2)
the Hough transform converts the image space into a parameter space, and the circle represented by equation (2) may become:
H＝(x,y) T ,A＝[a,b,r] T (3)
the parameter space is now a three-dimensional space. Then for a point (x, y) in the image space a cone is corresponding in the parameter space. A circle in image space corresponds to a point where a cluster of cones intersect, and the specific point has a certain three-dimensional parameter in parameter space, and at the same time, represents a circle with a specific radius and a specific center coordinate in image space.
In order to improve the calculation efficiency, the invention adopts Hough transformation of image gradient information, and a standard equation of a circle derives x:
an accumulator array calculation is first performed. The foreground pixels of high gradient are designated as candidate pixels and allowed to "vote" in the accumulator array. Candidate pixels vote in a pattern of a full circle of fixed radius around the pixel. The voting areas of candidate pixels belonging to the same image circle are accumulated in an accumulator array corresponding to the center of the circle. Thus, the center of the circle can be estimated by detecting the peak value of the accumulator array. If the same accumulator array is used for multiple radius values, the estimated radius may be displayed using the estimated center and the image information. Taking the estimated pixel position of the center of the circle in the shot image as the center position of the sensor, and marking the position as (w) according to the serial number i of the sensor i ,h i )。
With the above method, the index points in the image are extracted for image registration. The chest lock joint (x) of the subject selected as the target point in the invention 1 ,y 1 ) Tenth rib (x 2 ,y 2 ) Left chest rib (x) 3 ,y 3 ) Right chest rib (x) 4 ,y 4 ). At the same time, a calibration point (s 1 ,k 1 )，(s 2 ,k 2 )，(s 3 ,k 3 )，(s 4 ,k 4 )。
Rigid transformations include rotational transformations, scaling transformations, and translational transformations.
The rotational transformation ensures horizontal registration of the images. The position coordinates of the i-th sensor are (w i ,h i ) The new coordinates after rotation are (w ix ,h ix )：
w ix ＝cos(theta)*w i -sin(theta)*h i (5)
h ix ＝sin(theta)*w i +cos(theta)*h i (6)
Converting the formula (5) and the formula (6) into a matrix form:
the theta is the angle of clockwise rotation of the images, namely the horizontal angle difference between the left chest rib and the right chest rib in the two images:
theta＝cot((y 4 -y 3 )/(x 4 -x 3 ))-cot((k 4 -k 3 )/(s 4 -s 3 )) (8)
after rotation transformation, the coordinate of the miniature heart sound sensor is required to be scaled and transformed to obtain a new coordinate (w ixs ,h ixs ):
w ixs ＝n x *w ix (9)
h ixs ＝n y *h ix (10)
Converting the formula (9) and the formula (10) into a matrix form:
wherein n is x Is the magnification of the abscissa, n y Is the magnification of the ordinate:
n x ＝(x 4 -x 3 )/(s 4 -s 3 ) (12)
n y ＝(y 2 -y 1 )/(k 2 -k 1 ) (13)
after rotation transformation and scaling transformation, the coordinate of the miniature heart sound sensor is subjected to translation transformation to obtain a new coordinate (w ixsp ,h ixsp )：
w ixsp ＝w ixs +L x (14)
h ixsp ＝h ixs +L y (15)
Converting the formulas (14) and (15) into a matrix form:
wherein L is x Is a pixel value shifted horizontally, L y Pixel values for vertical translation:
L x ＝(x 3 –s 3 )+s 3 (17)
L y ＝(y 1 –k 1 )+s 1 (18)
through the image registration, the position coordinates (w i ,h i ) Is converted into human skeleton image position coordinates (w ixsp ,h ixsp )。
(3) As shown in fig. 3, the heart sound energy relationship diagram is that the color of the region with higher energy value is bright, the color of the region with lower energy value is darker, the energy value sequentially corresponds to blue, green, yellow and red from low to high, and the transparency is also changed. The areas within the contour lines in the figure fill in colors by energy values. The energy value of the heart sound signal and the HSL color value are inversely proportional transformed, and then the HSL is converted into RGB for drawing.
The invention is based on heart sound array signals and electrocardiosignals. After the signals are preprocessed, defining the heart sound signals acquired by the ith channel as S i (t), wherein t is a time sequence, and the invention utilizes the energy of the signals to research the propagation relationship of heart sound signals on the body surface, and the amplitude of the heart sound signals is V i (t) electrocardiosignal S ECG (t). The energy of the heart sound signal is calculated by the formula:
E i (t)＝V i (t)*V i (t) (19)
wherein E is i (t) is the energy of the ith channel heart sound signal at the moment t, V i And (t) is the amplitude of the ith channel heart sound signal at the moment t. The position coordinates (w ixsp ,h ixsp ) Energy E of heart sound signal i (t) thereby obtaining three-dimensional data H (w ixsp ,h ixsp ,E i (t)). The three-dimensional data corresponds to a pixel point of the skeleton image, and interpolation is needed to be carried out on the heart sound signals in the skeleton image in order to show the energy relation of the heart sound signals on the body surface.
The interpolated image is drawn based on linear interpolation. The invention adopts a circular drawing method to draw the heart sound signal energy diagram. Setting the radius of a pixel as R by taking the center of a miniature heart sound sensor as the center of a circle x Concentric circles are drawn with pixel 1 as a step size. The invention adopts Bresenham circle drawing algorithm, and the algorithm draws a circle by calculating octant circumference points and utilizing symmetry. Setting the area with high energy as the center of a circle, and setting the energy value E i,R (t) gradually decreases with increasing radius of the concentric circles, calculated as follows:
R g ＝R/(R x ) (20)
E i,R (t)＝E i (t)*(1-R g ) (21)
wherein R is the radius of concentric circles, R x Radius R is the radius of the concentric circle with the largest radius g Normalized radius value for a circle of radius R, E i (t) is the signal energy value of the ith channel at time t, E i,R And (t) is a signal energy value corresponding to a concentric circle of the radius R at the moment of the ith channel t.
The energy color wheel is set based on (H, S, L) color values, where H is hue, S is saturation, and L is brightness. In the present invention, the values of H were set to 240 to 0,S to 1.0 and L to 50%. H changes in inverse proportion with the change of energy value, namely the color of the high-energy area is bright, the color of the low-energy area is dark, and the calculation mode is as follows:
H g ＝E i (t)/(max(E i (t))) (22)
H i (t)＝240*(1-H g ) (23)
wherein E is i (t) is the signal energy value of the ith channel at time t, max (E i (t)) is the maximum value of the channel signal energy, H g For the normalized signal energy value at time t, H i (t) is the color converted at the time of the ith channel tAnd (3) phase (C). And converting the HSL color value into RGB color, filling the RGB color into a circular ring with a corresponding radius, and drawing a heart sound signal energy relation diagram. The energy values in the map fill in colors. The lighter colored areas in the graph represent higher energy of the sound source signal at that location, and the darker colored areas represent lower energy.
As can be seen from fig. 3 (a) to 3 (g), the energy of different sound source signals in each region of the human body is plotted. The energy relation diagram has the characteristics of high spatial resolution and high time resolution. The position of the sensor in the figure corresponds to the skeleton structure of a human body, the energy relation diagram is combined with the electrocardiosignal of the human body for analysis, and corresponds to the heart movement period of the human body, so that the sensor has important significance for researching the influence of the heart sound origin position, the organ tissue on the heart sound transmission and the attenuation change of heart sound energy.
Claims (3)
1. The heart sound propagation relation detection system based on the wearable sensor array is characterized by comprising an electrocardio sensor, the wearable sensor array, a signal acquisition card and a signal analysis and processing unit in a computer end;
the electrocardiosignal is used for collecting a first lead electrocardiosignal of a human body and identifying the heart movement period of the human body;
the wearable sensor array mainly comprises a miniature heart sound sensor, a wearable waistcoat and an inflatable air bag, and the structure of the array is a network topological structure;
the wearable waistcoat is suitable for people of different sizes, and different network topological structures are arranged on the wearable waistcoat according to requirements and used for arranging the miniature heart sound sensor; the miniature heart sound sensor is embedded in the inner side of the wearable waistcoat, and in order to improve the fitting degree of the sensor and the human body surface, an inflatable air bag is added in an interlayer of the wearable waistcoat, and the inflatable air bag provides balanced air pressure for each sensor;
the acquisition card is used for synchronously acquiring human heart sound array signals acquired by the wearable sensor array and electrocardiosignals acquired by the electrocardiosignal, converting analog signals into digital signals and transmitting the digital signals to the computer end;
the signal analysis and processing unit comprises signal preprocessing, image registration and heart sound signal linear interpolation drawing;
the signal de-averaging is used for avoiding the influence of signal baseline drift on energy analysis; the signal band-pass filtering is used for carrying out Butterworth band-pass filtering and noise reduction on the heart sound signals and the electrocardiosignals, setting the passband of the heart sound signals to be 30-150 Hz and the passband of the electrocardiosignals to be 0.5-35 Hz, and carrying out bidirectional zero phase shift filtering on the heart sound signals and the electrocardiosignals to avoid the influence of phase offset on the research of propagation relations; the heart sound signal wavelet is used for reducing noise and further filtering noise in the heart sound signal;
the image registration is used for registering heart sound acquisition points with human skeleton images, and the spatial positions of the miniature heart sound sensors are rigidly registered with the skeleton images in order to accurately map the spatial positions of the heart sound acquisition points into human skeleton structures; the method comprises the following steps: firstly, shooting a chest image of a subject, and adding visual calibration points into the chest image and a skeleton image; meanwhile, a visual tag is added to the position of the miniature heart sound sensor in the chest image; based on the chest image and the calibration points in the bones, carrying out rotation transformation, scaling transformation and translation transformation on the visual label of the miniature heart sound sensor according to the rigid transformation principle, and completing the mapping of the topological structure of the miniature heart sound sensor network in the bone image;
the heart sound signal linear interpolation drawing is used for drawing an energy propagation relation diagram of the heart sound signal of the human body; in order to draw the energy information of the heart sound signals on the body surface of the human body, the energy of the heart sound signals is calculated through the amplitude values of the heart sound signals; drawing concentric circles by the position of the miniature heart sound sensor in the skeleton image by taking each heart sound acquisition point as a circle center, and linearly decreasing the corresponding heart sound signal energy along with the increase of the radius of the concentric circles; the heart sound signal energy sequentially corresponds to red, yellow, green, cyan and blue from high to low, colors are filled into corresponding concentric circles, and an energy propagation relation diagram of the heart sound signal of the human body is drawn; and correspondingly analyzing an energy propagation relation diagram of the heart sound signals of the human body and the period of the electrocardiosignals, and simultaneously analyzing the propagation trend of the heart sound signals from the angles of time and space.
2. The system for detecting heart sound propagation relationship based on the wearable sensor array according to claim 1, wherein the image registration is specifically as follows:
adding a visual tag to the miniature heart sound sensor and the human skeleton calibration point for marking the specific positions of the skeleton calibration point and the miniature heart sound sensor in the acquisition process; the visual label adopted by the miniature heart sound sensor is a green circle, and the size of the visual label is consistent with that of the miniature heart sound sensor in the image; the visual label adopted by the human skeleton calibration point is a red circle; shooting a chest image of a tested person containing a visual tag after the acquisition is completed; in the image processing process, in order to distinguish the calibration point from the miniature heart sound sensor, converting a color RGB space into an HSV space, and respectively extracting a red part and a green part according to HSV parameter values corresponding to red and green to obtain images respectively containing the calibration point and the miniature heart sound sensor;
performing binarization processing on the image respectively comprising the calibration point and the miniature heart sound sensor, eliminating tone and saturation information, and representing each element in the image by a brightness value of 0 to 255; calculating the gray value of the image by adopting a weighted average method:
g＝0.2989*R+0.5870*G+0.1140*B (1)
wherein R, G, B is the red, green, blue component of the color image; 0.2989, 0.5870 and 0.1140 are respectively the weight values of R, G, B, and g is the gray value of the image;
after binarization processing, searching for a circle in the image based on circular Hough transformation; for a circle of radius r, center (a, b), expressed as:
(x-a) 2 +(y-b) 2 ＝r 2 (2)
the Hough transform converts the image space into a parameter space, and the circle represented by equation (2) becomes:
H＝(x,y) T ,A＝[a,b,r] T (3)
the parameter space is a three-dimensional space at this time; then for a point (x, y) in the image space a cone is corresponding in the parameter space; a circle of the image space corresponds to a point intersected by a cluster of cones, the specific point has a certain three-dimensional parameter in the parameter space, and a circle with a specific radius and a specific center coordinate is represented in the image space;
in order to improve the calculation efficiency, hough transformation of image gradient information is adopted, and a standard equation of a circle is used for deriving x:
firstly, performing accumulator array calculation; the foreground pixels of high gradient are designated as candidate pixels and allowed to "vote" in the accumulator array; candidate pixels vote in a pattern of a full circle of fixed radius around the pixel; voting areas of candidate pixels belonging to the same image circle are accumulated in an accumulator array corresponding to the center of the circle; thus, the center of the circle is estimated by detecting the peak value of the accumulator array; if the same accumulator array is used for a plurality of radius values, displaying an estimated radius by using the estimated circle center and the image information; taking the estimated pixel position of the center of the circle in the shot image as the center position of the sensor, and marking the position as (w) according to the serial number i of the sensor i ,h i )；
Thereby extracting the index points in the image for image registration; the calibration point is selected as the chest lock joint (x) 1 ,y 1 ) Tenth rib (x 2 ,y 2 ) Left chest rib (x) 3 ,y 3 ) Right chest rib (x) 4 ,y 4 ) The method comprises the steps of carrying out a first treatment on the surface of the At the same time, a calibration point (s 1 ,k 1 )，(s 2 ,k 2 )，(s 3 ,k 3 )，(s 4 ,k 4 )；
Rigid transformations include rotational transformations, scaling transformations, and translational transformations;
the rotation transformation ensures the horizontal registration of the images; the position coordinates of the i-th sensor are (w i ,h i ) The new coordinates after rotation are (w ix ,h ix )：
w ix ＝cos(theta)*w i -sin(theta)*h i (5)
h ix ＝sin(theta)*w i +cos(theta)*h i (6)
Converting the formula (5) and the formula (6) into a matrix form:
the theta is the angle of clockwise rotation of the images, namely the horizontal angle difference between the left chest rib and the right chest rib in the two images:
theta＝cot((y 4 -y 3 )/(x 4 -x 3 ))-cot((k 4 -k 3 )/(s 4 -s 3 )) (8)
after rotation transformation, the coordinate of the miniature heart sound sensor is required to be scaled and transformed to obtain a new coordinate (w ixs ,h ixs ):
w ixs ＝n x *w ix (9)
h ixs ＝n y *h ix (10)
Converting the formula (9) and the formula (10) into a matrix form:
wherein n is x Is the magnification of the abscissa, n y Is the magnification of the ordinate:
n x ＝(x 4 -x 3 )/(s 4 -s 3 ) (12)
n y ＝(y 2 -y 1 )/(k 2 -k 1 ) (13)
after rotation transformation and scaling transformation, the coordinate of the miniature heart sound sensor is subjected to translation transformation to obtain a new coordinate (w ixsp ,h ixsp )：
w ixsp ＝w ixs +L x (14)
h ixsp ＝h ixs +L y (15)
Converting the formulas (14) and (15) into a matrix form:
wherein L is x Is a pixel value shifted horizontally, L y Pixel values for vertical translation:
L x ＝(x 3 –s 3 )+s 3 (17)
L y ＝(y 1 –k 1 )+s 1 (18)
through image registration, the position coordinates (w i ,h i ) Is converted into human skeleton image position coordinates (w ixsp ,h ixsp )。
3. A system for detecting a heart sound propagation relationship based on a wearable sensor array according to claim 1 or 2, wherein the heart sound signal linear interpolation map is as follows:
after the signals are preprocessed, defining the heart sound signals acquired by the ith channel as S i (t), wherein t is a time sequence, the propagation relationship of the heart sound signal on the body surface is researched by using the energy of the signal, and the amplitude of the heart sound signal is V i (t) electrocardiosignal S ECG (t); the energy of the heart sound signal is calculated by the formula:
E i (t)＝V i (t)*V i (t) (19)
wherein E is i (t) is the energy of the ith channel heart sound signal at the moment t, V i (t) is the amplitude of the ith channel heart sound signal at time t; position coordinates (w) of the miniature heart sound sensor in the image space ixsp ,h ixsp ) Energy E of heart sound signal i (t) thereby obtaining three-dimensional data H (w ixsp ,h ixsp ,E i (t)); the three-dimensional data corresponds to a pixel point of the skeleton image, which is the exhibitionShowing the energy relation of heart sound signals on a body surface, and interpolating the heart sound signals in a skeleton image;
drawing an interpolation image based on a linear interpolation method: drawing a heart sound signal energy diagram by adopting a circular drawing method; setting the radius of a pixel as R by taking the center of a miniature heart sound sensor as the center of a circle x Drawing concentric circles by taking the pixel 1 as a step length; a Bresenham circle drawing algorithm is adopted, and the algorithm draws a circle by calculating octant circumference points and utilizing symmetry; setting the area with high energy as the center of a circle, and setting the energy value E i,R (t) gradually decreases with increasing radius of the concentric circles, calculated as follows:
R g ＝R/(R x ) (20)
E i,R (t)＝E i (t)*(1-R g ) (21)
wherein R is the radius of concentric circles, R x Radius R is the radius of the concentric circle with the largest radius g Normalized radius value for a circle of radius R, E i (t) is the signal energy value of the ith channel at time t, E i,R (t) is the signal energy value corresponding to the concentric circle of radius R at the moment of the ith channel t;
setting an energy color disc based on (H, S, L) color values, wherein H is hue, S is saturation, and L is brightness; setting the value of H to 240 to 0,S to 1.0 and L to 50%; h changes in inverse proportion with the change of energy value, namely the color of the high-energy area is bright, the color of the low-energy area is dark, and the calculation mode is as follows:
H g ＝E i (t)/(max(E i (t))) (22)
H i (t)＝240*(1-H g ) (23)
wherein E is i (t) is the signal energy value of the ith channel at time t, max (E i (t)) is the maximum value of the channel signal energy, H g For the normalized signal energy value at time t, H i (t) is the hue converted at the time of the ith channel t; converting the HSL color value into RGB color, filling the RGB color into a circular ring with a corresponding radius, and drawing a heart sound signal energy relation diagram; filling colors in the energy values in the corresponding graphs; the lighter colored areas in the graph represent darker colored areas with higher energy of the source signal at that locationThe lower the representative energy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111548482.0A CN114159091B (en) | 2021-12-17 | 2021-12-17 | Heart sound propagation relation detection system based on wearable sensor array |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111548482.0A CN114159091B (en) | 2021-12-17 | 2021-12-17 | Heart sound propagation relation detection system based on wearable sensor array |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114159091A CN114159091A (en) | 2022-03-11 |
| CN114159091B true CN114159091B (en) | 2023-10-13 |
Family
ID=80487117
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111548482.0A Active CN114159091B (en) | 2021-12-17 | 2021-12-17 | Heart sound propagation relation detection system based on wearable sensor array |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114159091B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115501007B (en) * | 2022-11-04 | 2023-03-17 | 清华大学 | Design and processing method of high polymer bone implant multi-channel sensor |
| CN116671928A (en) * | 2023-07-27 | 2023-09-01 | 中国科学技术大学 | Bimodal cardiac electromechanical physiological source imaging system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105653880A (en) * | 2016-03-03 | 2016-06-08 | 南京邮电大学 | Heart sound production field tracing method based on heart sounds |
| CN109498054A (en) * | 2019-01-02 | 2019-03-22 | 京东方科技集团股份有限公司 | Heart sound monitoring device, the method and configuration method for obtaining cardiechema signals |
| CN110226944A (en) * | 2019-06-24 | 2019-09-13 | 大连理工大学 | Miniature heart sound transducer and its application based on MEMS technology |
| CN112494000A (en) * | 2020-11-26 | 2021-03-16 | 河北德睿健康科技有限公司 | Coronary artery stenosis visualization quantification method and equipment based on multiple paths of heart sounds |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ATE535194T1 (en) * | 2006-12-11 | 2011-12-15 | Deepbreeze Ltd | METHOD AND SYSTEM FOR ANALYZING BODY NOISE |
-
2021
- 2021-12-17 CN CN202111548482.0A patent/CN114159091B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105653880A (en) * | 2016-03-03 | 2016-06-08 | 南京邮电大学 | Heart sound production field tracing method based on heart sounds |
| CN109498054A (en) * | 2019-01-02 | 2019-03-22 | 京东方科技集团股份有限公司 | Heart sound monitoring device, the method and configuration method for obtaining cardiechema signals |
| CN110226944A (en) * | 2019-06-24 | 2019-09-13 | 大连理工大学 | Miniature heart sound transducer and its application based on MEMS technology |
| CN112494000A (en) * | 2020-11-26 | 2021-03-16 | 河北德睿健康科技有限公司 | Coronary artery stenosis visualization quantification method and equipment based on multiple paths of heart sounds |
Non-Patent Citations (1)
| Title |
|---|
| 一种基于MEMS 技术的新型心音传感器及应用;果 斌 斌;硕士论文;1-63 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114159091A (en) | 2022-03-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114159091B (en) | Heart sound propagation relation detection system based on wearable sensor array | |
| CN110662480A (en) | Device, system and method for measuring and processing physiological signals of a subject | |
| CN106901741A (en) | A kind of respiratory rate detection method suitable for environment round the clock | |
| CN106108889A (en) | Electrocardiogram classification method based on degree of depth learning algorithm | |
| CN110236515A (en) | A kind of contactless heart rate detection method based on near-infrared video | |
| CA2934659A1 (en) | System and methods for measuring physiological parameters | |
| CN100488448C (en) | Three-dimensional image electrocardiograph and implementation method | |
| JP2005519685A (en) | Pixel coding method, image processing method, and image processing method for qualitative recognition of an object reproduced by one or more pixels | |
| CN101849813A (en) | Three-dimensional cardiac ultrasonic virtual endoscope system | |
| KR20120111871A (en) | Method and apparatus for creating medical image using 3d deformable model | |
| CN109770943A (en) | A kind of ultrasonic automatic optimization method positioned using computer vision | |
| CN101292864A (en) | Image expression measuring method and apparatus for human disease information hand zone imaging | |
| KR20200070062A (en) | System and method for detecting lesion in capsule endoscopic image using artificial neural network | |
| CN109124595A (en) | Intelligent sphygmus diagnostic method based on dynamic 3 D pulse wave image | |
| US8064979B2 (en) | Tempero-spatial physiological signal detection method and apparatus | |
| CN117012344B (en) | Image analysis method for 4CMOS camera acquisition | |
| KR20160061248A (en) | Apparatus for processing medical image and method for processing medical image thereof | |
| CN105498100A (en) | Method and system for monitoring physiological motion curve and body surface contour | |
| CN112294282A (en) | Self-calibration method of emotion detection device based on RPPG | |
| Schrumpf et al. | Exploiting weak head movements for camera-based respiration detection | |
| CN117558428B (en) | Imaging optimization method and system for liver MRI | |
| CN108852307B (en) | Non-contact non-invasive arteriosclerosis detection device | |
| CN110693510A (en) | Attention deficit hyperactivity disorder auxiliary diagnosis device and using method thereof | |
| US20210204857A1 (en) | Method and device for cardiac monitoring | |
| CN111466952B (en) | Real-time conversion method and system for ultrasonic endoscope and CT three-dimensional image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |