WO2023130661A1

WO2023130661A1 - Method and apparatus for processing two-dimensional spectral doppler echocardiographic image

Info

Publication number: WO2023130661A1
Application number: PCT/CN2022/097243
Authority: WO
Inventors: 马超; 张碧莹; 曹君
Original assignee: 乐普（北京）医疗器械股份有限公司
Priority date: 2022-01-07
Filing date: 2022-06-07
Publication date: 2023-07-13
Also published as: CN114376603B; CN114376603A

Abstract

A method and apparatus for processing a two-dimensional spectral Doppler echocardiographic image. The method comprises: acquiring a two-dimensional spectral Doppler echocardiographic image (1); performing region-of-interest image extraction processing to generate a second image (2); performing Gaussian blur image processing to generate a third image (3); performing binarization processing to generate a fourth image (4); performing spectral envelope identification to mark a first envelope (5); performing peak point identification on the first envelope to mark a plurality of first peak points (6); performing left and right baseline point identification on each first peak point to mark a corresponding first left baseline point and a corresponding first right baseline point (7); performing blood flow parameter measurement and calculation to generate a corresponding blood flow parameter sequence (8); and calculating an average value of parameters of the same type in the blood flow parameter sequence (9). Therefore, the problems of reduced measurement accuracy or unstable measurement quality, etc. caused by manual factors can be solved.

Description

A method and device for processing two-dimensional spectral Doppler echocardiographic images

This application claims the priority of the Chinese patent application with the application number 202210018428.3 and the title of the invention "a method and device for processing two-dimensional spectral Doppler echocardiographic images" submitted to the Chinese Patent Office on January 7, 2022.

technical field

The invention relates to the technical field of data processing, in particular to a method and device for processing two-dimensional spectrum Doppler echocardiographic images.

Background technique

Spectral Doppler echocardiography (Spectral Doppler Echocardiography) can be used to measure parameters related to blood flow velocity, such as peak flow velocity, acceleration time, deceleration time, ejection time, etc. Spectral Doppler echocardiography has longitudinal blood flow With the flow velocity scale and the horizontal time scale scale, the operator can mark the key points on the spectral Doppler echocardiogram to calculate the approximate peak flow velocity, acceleration time, deceleration time, and ejection time parameter values. Calculation of blood flow parameters in this way, on the one hand, relies too much on the experience level of manual key point marking, and its accuracy cannot be guaranteed; The pressure gradient associated with pressure changes in the direction of blood flow and the halving time of the pressure gradient.

Contents of the invention

The object of the present invention is to provide a processing method, device, electronic equipment and computer-readable storage medium for two-dimensional spectral Doppler echocardiographic images, aiming at the defects of the prior art. The cardiac image is clipped for the region of interest, Gaussian blur processing and binarization processing, the spectrum envelope is extracted from the binary image, and the Gaussian kernel weight sliding window is used to perform sliding window weight calculation on the envelope to complete the envelope The peak point identification on the top, calculate the corresponding left and right baseline points based on the amplitude difference and time interval control conditions of the peak point, and obtain the peak flow velocity, acceleration time, deceleration related to each peak point based on each peak point and its corresponding left and right baseline points Time, ejection time, speed-time integral, pressure gradient and pressure gradient halving time, and can be further converted to obtain the average value of various measurement parameters. Through the present invention, when measuring blood flow parameters based on spectral Doppler echocardiography, not only can the problems of reduced measurement accuracy or unstable measurement quality caused by artificial factors be solved, but also other problems that cannot be measured by traditional manual methods can be measured. data, expanding the parameter measurement range.

In order to achieve the above purpose, the first aspect of the embodiment of the present invention provides a method for processing two-dimensional spectral Doppler echocardiographic images, the method comprising:

acquiring a two-dimensional spectral Doppler echocardiographic image to generate a first image;

performing region-of-interest image extraction processing on the first image to generate a corresponding second image;

performing Gaussian blur image processing on the second image to generate a corresponding third image;

performing binarization processing on the third image to generate a corresponding fourth image;

performing spectrum envelope identification processing on the fourth image to mark the corresponding first envelope;

performing peak point identification processing on the first envelope to mark a plurality of first peak points;

Perform left and right baseline point identification processing on each of the first peak points to mark the corresponding first left baseline point and first right baseline point;

According to the first envelope marked with the peak point and the left and right baseline points, perform blood flow parameter calculation to generate a corresponding blood flow parameter set sequence; the blood flow parameter set sequence includes multiple blood flow parameter sets; the blood flow parameter set sequence includes multiple blood flow parameter sets; The flow parameter group includes a peak flow velocity parameter, a pressure gradient parameter, an acceleration time parameter, a deceleration time parameter, an ejection time parameter, a pressure difference halving time parameter and a velocity time integration parameter; the blood flow parameter group is related to the first peak value One-to-one correspondence;

Calculate the average value of each similar parameter in the blood flow parameter group sequence to obtain the average peak flow velocity, average pressure gradient, average acceleration time, average deceleration time, average ejection time, and average pressure difference halving time value and velocity-time-integrated average, and a measurement data set composed of all the average values is returned as the measurement data result of the two-dimensional spectral Doppler echocardiographic image.

Preferably, performing the region-of-interest image extraction process on the first image to generate a corresponding second image specifically includes:

Perform blood flow velocity zero line identification processing on the first image to mark the corresponding first zero line;

If the large peak of the spectrum image in the first image is upward, then extract the sub-image from the top of the image to the first zero line in the first image as the first sub-image; if the first image in the first image If the large peak of the spectral image is facing downward, then extract the sub-image from the first zero line to the bottom of the image in the first image, and perform image flip processing on the extracted sub-image to generate the first sub-image; The bottom of the image of the first sub-image is the first zero line;

The sum of the pixel values of each row of pixels in the first sub-image is counted to generate a corresponding first row of pixel sums; and the image row corresponding to the first row of pixel sums with the smallest numerical value is recorded as the smallest pixel row; and extracting the sub-image from the minimum pixel row to the bottom of the image in the first sub-image as an image of the region of interest to generate the second image.

Preferably, performing spectrum envelope identification processing on the fourth image to mark the corresponding first envelope specifically includes:

Rotate the fourth image to the left by 90° to generate a corresponding first transposed binary image;

Carrying out row-by-row inspection on the first transposed binary image, clustering the continuous pixel points whose pixel values in the current row are preset foreground point pixel values, and generating a corresponding sequence of continuous pixel points;

Optimum sequence screening is carried out to a plurality of said continuous pixel point sequences of the same row, and the maximum number of pixels is used as the optimal continuous pixel point sequence corresponding to the current row; Boundary pixel points are marked as row boundary points;

According to the corresponding relationship between the pixel coordinate transposition of the first transposed binary image and the fourth image, the pixel points corresponding to each of the row boundary points in the fourth image are recorded as column boundary points;

Connecting the column boundary points sequentially to obtain a first connection line; and smoothing the first connection line to obtain the first envelope; and completing the first envelope on the fourth image Marking of threads.

Preferably, performing peak point identification processing on the first envelope to mark a plurality of first peak points specifically includes:

Taking the vertical distance from each sampling point of the first envelope to the zero line at the bottom of the fourth image as the sampling value of the sampling point, performing statistics on the sampling values of each sampling point of the first envelope to generate the first The sampling value sequence is {x ₁ , x ₂ ... x _i ... x _n }, i is the sampling point index, 1≤i≤n, x _i is the sampling value of each sampling point, n is the first envelope Total number of sampling points;

Construct a Gaussian kernel weight sliding window; set the sliding window width w of the Gaussian kernel weight sliding window; set the sampling value sequence in the Gaussian kernel weight sliding window as {s ₁ ... s _j ... s _w }, j is The index of the sampling point in the sliding window, 1≤j≤w, s _j is the sampling value of each sampling point in the sliding window; according to the standard Gaussian function

Taking the maximum sampling point index j _max corresponding to the maximum sampling value s _max in the sliding window as the mean value μ, and taking a quarter of the sliding window width w/4 as the variance σ, construct the weight of each sampling point in the Gaussian kernel weight sliding window The Gaussian kernel coefficient operation function is

_kj is the Gaussian kernel coefficient of each sampling point in the Gaussian kernel weight sliding window; according to the Gaussian kernel coefficient operation function, the sliding window weight operation function of the Gaussian kernel weight sliding window is constructed as

A is the sliding window weight, and k' _j is the normalized Gaussian kernel coefficient corresponding to each described Gaussian kernel coefficient k _j in the sliding window;

In the first sample value sequence {x ₁ , x ₂ ... x _i ... x _n }, starting from the first sample value x ₁ , the sliding window is sampled with a step size of 1 and the sliding window width w The number of points, the first sampling value sequence {x ₁ , x ₂ ... x _i ... x _n } is divided into the second number of sub-sliding window sequences C _h ; the sub-sliding window sequence C _h is {xi _{= h} , x _i=h ₊₁ ... x _i=h+w-1 }, h is the sub-sliding window index, 1≤h≤the second number, the second number=n-w+1;

Use the Gaussian kernel weight sliding window to perform a sliding window weight calculation on each of the sub-sliding window sequences C _h ; during the operation, each sampling value of the current sub-sliding window sequence C _h is converted into Sampling value s _j , and taking the maximum value as the maximum sampling value s _max , and taking the sampling point index of the maximum sampling value s _max in the sliding window as the corresponding maximum sampling point index j _max ; and sampling in each sliding window The sampling point index of the value s _j and the maximum sampling point index j _max are substituted into the Gaussian kernel coefficient operation function to obtain a plurality of Gaussian kernel coefficients k _j ; and normalize all current Gaussian kernel coefficients k _j to obtain A plurality of normalized Gaussian kernel coefficients k'_j; and all current normalized Gaussian kernel coefficients k' _j and their corresponding sampling values s _j in the sliding window are substituted into the sliding window weight calculation function to obtain the corresponding The sliding window weight A _h of ;

Record the first sequence of sampled values {x ₁ , x ₂ ... x _i ... _x _n } as the current sequence; The sequence C _h is marked as the current sub-sliding window sequence; and the sampling point index corresponding to the maximum sampling value on the current sub-sliding window sequence is marked as the peak point index P; and with the peak point index P, the The current sampling value sequence is divided into left and right parts and is recorded as a left sampling value sequence and a right sampling value sequence; Perform peak point index marking processing on the sampling point index of the maximum sampling value of the sub-sliding window sequence corresponding to the maximum sliding window weight until the sequence length of the new current sampling value sequence is lower than the preset minimum sequence length;

Taking the sampling points corresponding to all the peak point indexes P on the first envelope as the first peak point.

Preferably, the identifying the left and right baseline points for each of the first peak points to mark the corresponding first left baseline point and first right baseline point specifically includes:

On the first envelope, each of the first peak points is the current peak point;

According to the preset time length threshold, a corresponding left envelope interval and a right envelope interval are respectively divided from the current peak point to the left and right;

Record the minimum envelope amplitude on the left envelope interval and the right envelope interval as the corresponding left interval minimum and right interval minimum;

On the left envelope interval, start from the current peak point to traverse the left valley point to the left; when traversing, calculate the difference between the amplitude of the current peak point and the minimum value of the left interval to generate The first difference, calculate the difference between the amplitude of the current left valley point and the minimum value of the left interval to generate the second difference, calculate the ratio of the second difference to the first difference to generate the first difference Ratio, if the first ratio is less than the preset error range, then use the current left valley point as the first left baseline point corresponding to the current peak point and stop traversing, if the first ratio If it is greater than or equal to the preset error range, go to the next left valley point and continue traversing;

On the right envelope line interval, start from the current peak point to the right to traverse the right valley point; when traversing, calculate the difference between the amplitude of the current peak point and the minimum value of the right interval to generate The third difference is to calculate the difference between the amplitude of the current right valley point and the minimum value of the right interval to generate the fourth difference, and calculate the ratio of the fourth difference to the third difference to generate the second difference. Ratio, if the second ratio is less than the preset error range, then use the current right valley point as the first right baseline point corresponding to the current peak point and stop traversing, if the second ratio If it is greater than or equal to the preset error range, go to the next valley point on the right to continue traversing.

Preferably, the blood flow parameters are measured and calculated according to the first envelope marked by the completed peak point and the left and right baseline points to generate a corresponding blood flow parameter set sequence, specifically including:

On the first envelope, take each of the first peak points as the current peak point, take the first left baseline point corresponding to the current peak point as the current left baseline point, and take the current peak point as the current left baseline point. The corresponding first right baseline point is the current right baseline point;

Taking the vertical distance from the current peak point to the bottom zero line of the fourth image as the corresponding peak distance h, and calculating the corresponding peak flow velocity parameter V according to the preset unit peak distance blood flow velocity V _s and the peak distance h _max , V _max =V _s *h;

Calculate and generate the corresponding pressure gradient parameter ΔP according to the peak flow velocity parameter V _max , wherein,

Taking the time interval from the current left baseline point to the current peak point as the corresponding acceleration time parameter T _a ;

Taking the time interval from the current peak point to the current right baseline point as the corresponding deceleration time parameter T _d ;

Taking the sum of the acceleration time parameter T _a and the deceleration time parameter T _d as the corresponding ejection time parameter T _e ;

Record the envelope segment from the current peak point to the current right baseline point on the first envelope as the current segment; and on the current segment, proceed to the right from the current peak point Sampling point traversal; when traversing, the vertical distance from the current sampling point to the bottom zero line of the fourth image is taken as the corresponding sampling point distance h _sam , and according to the sampling point distance h _sam and the unit peak distance blood flow velocity V _s Calculate and generate the corresponding sampling point flow velocity V _sam =V _s *h _sam , and calculate and generate the corresponding sampling point pressure gradient according to the sampling point flow velocity V _sam

And calculate the ratio of the pressure gradient ΔP _sam of the sampling point to the pressure gradient parameter ΔP to generate a first ratio, if the first ratio enters the preset half value ratio confirmation range, stop traversing and put the The current sampling point is used as the half-value sampling point of the pressure difference. If the first ratio has not yet entered the half-value ratio confirmation range, it will stop and go to the next sampling point to continue traversing; The time interval between value sampling points is used as the corresponding pressure difference halving time parameter T _ΔP/2 ;

performing a velocity integration operation on the first envelope segment from the current left baseline point to the current right baseline point to generate the corresponding velocity time integration parameter;

The peak flow velocity parameter V _max , the pressure gradient parameter ΔP, the acceleration time parameter T _a , the deceleration time parameter T _d , the ejection time parameter T _e , and the pressure difference halving time The parameter T _ΔP/2 and the velocity-time integral parameter form the blood flow parameter set corresponding to the current peak point; and add the blood flow parameter set to the blood flow parameter set sequence.

The second aspect of the embodiment of the present invention provides a device for implementing the method described in the first aspect above, including: an acquisition module, an image preprocessing module, an envelope processing module, and a blood flow parameter calculation module;

The acquiring module is used to acquire a two-dimensional spectral Doppler echocardiographic image to generate a first image;

The image preprocessing module is used to perform region-of-interest image extraction processing on the first image to generate a corresponding second image; and perform Gaussian blur image processing on the second image to generate a corresponding third image; and The third image is binarized to generate a corresponding fourth image;

The envelope processing module is used to perform spectrum envelope identification processing on the fourth image to mark the corresponding first envelope; and perform peak point identification processing on the first envelope to mark multiple The first peak point; and performing left and right baseline point identification processing on each of the first peak points to mark the corresponding first left baseline point and first right baseline point;

The blood flow parameter calculation module is used to perform blood flow parameter calculation and generate a corresponding blood flow parameter set sequence according to the first envelope marked with the peak point and the left and right baseline points; the blood flow parameter set sequence includes multiple a blood flow parameter group; the blood flow parameter group includes a peak flow velocity parameter, a pressure gradient parameter, an acceleration time parameter, a deceleration time parameter, an ejection time parameter, a pressure difference halving time parameter and a velocity time integral parameter; The flow parameter group is in one-to-one correspondence with the first peak point;

The blood flow parameter calculation module is also used to calculate the average value of each similar parameter in the blood flow parameter group sequence to obtain the average value of peak flow velocity, average value of pressure gradient, average value of acceleration time, average value of deceleration time, and ejection time. The time average value, the pressure difference halving time average value and the velocity time integral average value, and all the average values form a measurement data set to be returned as the measurement data results of the two-dimensional spectral Doppler echocardiographic image.

The third aspect of the embodiment of the present invention provides an electronic device, including: a memory, a processor, and a transceiver;

The processor is configured to be coupled with the memory, read and execute instructions in the memory, so as to implement the method steps described in the first aspect above;

The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.

The fourth aspect of the embodiments of the present invention provides a computer-readable storage medium, the computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a computer, the computer executes the above-mentioned first aspect. method directive.

Embodiments of the present invention provide a processing method, device, electronic equipment, and computer-readable storage medium for a two-dimensional spectral Doppler echocardiographic image. Clipping, Gaussian blur processing and binarization processing to reduce image noise and improve image recognition accuracy, and then extract the spectral envelope of the binary image to improve data recognition accuracy and increase the recognition ability of continuous data, and then pass Use the Gaussian kernel weight sliding window to perform sliding window weight calculation on the envelope to improve the recognition accuracy of the peak point of the normal signal on the envelope. After the peak point is obtained, the correspondence is calculated by the amplitude difference and time interval relationship with the peak point. Finally, based on each peak point and its corresponding left and right baseline points, not only the peak flow velocity, acceleration time, deceleration time, and ejection time related to each peak point can be obtained, but also the blood flow integral that cannot be measured by conventional methods can be obtained. It is the speed-time integral, the pressure gradient and the half-time of the pressure gradient, and at the same time, it can be further converted to obtain the average value of various measurement parameters. Through the present invention, when measuring blood flow parameters based on spectral Doppler echocardiography, not only can the problems of reduced measurement accuracy or unstable measurement quality caused by artificial factors be solved, but also other problems that cannot be measured by traditional manual methods can be measured. data, expanding the parameter measurement range.

Description of drawings

FIG. 1 is a schematic diagram of a processing method for a two-dimensional spectral Doppler echocardiographic image provided by Embodiment 1 of the present invention;

Fig. 2a is a schematic diagram of a group of first images and corresponding first sub-images provided by Embodiment 1 of the present invention;

Fig. 2b is a schematic diagram of another group of first images and corresponding first sub-images provided by Embodiment 1 of the present invention;

Fig. 2c is a schematic diagram of a set of third images and fourth images provided by Embodiment 1 of the present invention;

Fig. 2d is a schematic diagram of a group of fourth images and the first transposed binary image provided by Embodiment 1 of the present invention;

3 is a block diagram of a two-dimensional spectral Doppler echocardiographic image processing device provided by Embodiment 2 of the present invention;

FIG. 4 is a schematic structural diagram of an electronic device provided by Embodiment 3 of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments . Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

A processing method for a two-dimensional spectral Doppler echocardiographic image provided by Embodiment 1 of the present invention, as shown in FIG. 1 is a schematic diagram of a processing method for a two-dimensional spectral Doppler echocardiographic image provided by Embodiment 1 of the present invention , this method mainly includes the following steps:

Step 1, acquiring a two-dimensional spectral Doppler echocardiographic image to generate a first image.

Here, the two-dimensional spectral Doppler echocardiographic image may specifically be a two-dimensional continuous wave Doppler (continuous wave Doppler, CW) echocardiographic image.

Step 2, performing region-of-interest image extraction processing on the first image to generate a corresponding second image;

It specifically includes: step 21, performing blood flow velocity zero line identification processing on the first image to mark the corresponding first zero line;

Here, on the original two-dimensional spectral Doppler echocardiographic image, that is, the first image, the zero line of blood flow velocity is marked either by a standard scale or by a line of a special color. ways to identify and locate them;

In one of the implementation manners, performing blood flow velocity zero line identification processing on the first image and marking the corresponding first zero line specifically includes: calculating the average pixel value of each row of pixels in the first image to obtain the corresponding first row Pixel mean value, calculate the difference between the pixel mean value of the first row and the preset zero line pixel value to generate the corresponding first pixel difference value, if the first pixel difference value meets the preset zero line pixel error range, then the first pixel The line corresponding to the difference is taken as the zero line line, and the first zero line is marked on the first image according to the zero line line;

Step 22, if the large peak of the spectrum image in the first image faces upward, then extract the sub-image from the top of the image to the first zero line in the first image as the first sub-image; if the large peak of the spectrum image in the first image faces Next, extract the sub-image from the first zero line to the bottom of the image in the first image, and perform image flip processing on the extracted sub-image to generate the first sub-image; the bottom of the image of the first sub-image is the first zero line ;

Here, the large peak of the spectrum image is facing upwards and downwards corresponding to the two opposite directions of blood flow; in order to facilitate image processing, the current step processes all the spectral images as the shape of the large peak facing upwards, so in the spectral image the large peak is facing downwards In addition, for the original two-dimensional spectral Doppler echocardiographic image, if the large peak is facing upward, there will still be some interference signals with smaller peaks below the zero line of blood flow velocity. If the peak point is downward, there will generally be some interference signals with smaller peaks above the zero line of blood flow velocity, so these interference signal images will also be cut out when the image is cropped in the current step;

Take Figure 2a as an example of a group of first images and corresponding first sub-images provided by Embodiment 1 of the present invention, where the large peak of the first image faces upward, and the image is drawn from the top of the first image to the first zero line Obtain the first sub-image after clipping;

Take Fig. 2b as an example of another group of first images and corresponding first sub-images provided by Embodiment 1 of the present invention, where the large peak of the first image faces downward, first from the first zero line to the bottom of the first image performing image clipping, and then flipping the clipped image to obtain the first sub-image;

Step 23, counting the sum of the pixel values of each row of pixels in the first sub-image to generate a corresponding first row of pixel sum; and recording the image row corresponding to the first row of pixel sum with the smallest value as the minimum pixel row; and Extracting the sub-image from the minimum pixel row to the bottom of the image in the first sub-image as the image of the region of interest to generate the second image.

Here, in the image processing process, some useless backgrounds still exist on the top of the first sub-image. In order to improve image recognition, the current step is to further crop the first sub-image to delete some useless background lines on the top.

Step 3, performing Gaussian blur image processing on the second image to generate a corresponding third image.

Here, Gaussian blur processing is performed on the image to further eliminate image noise.

Step 4, performing binarization processing on the third image to generate a corresponding fourth image.

Here, FIG. 2c is a schematic diagram of a group of third images and fourth images provided by Embodiment 1 of the present invention.

Step 5, performing spectrum envelope recognition processing on the fourth image to mark the corresponding first envelope;

It specifically includes: step 51, rotating the fourth image 90° to the left to generate a corresponding first transposed binary image;

Here, the reason why the fourth image is transposed is to use a more convenient row traversal method in the subsequent steps; the fourth image before and after the transposition and the first transposed binary image, as shown in Figure 2d for the present invention A set of fourth images and a schematic diagram of the first transposed binary image provided in Embodiment 1 are shown;

Step 52, checking the first transposed binary image line by line, clustering the continuous pixel points in the current line whose pixel values are all preset foreground point pixel values, and generating a corresponding continuous pixel point sequence;

Step 53, performing optimal sequence screening on multiple consecutive pixel point sequences in the same row, taking the largest number of pixels as the optimal continuous pixel point sequence corresponding to the current row; Pixels are marked as row boundary points;

Here, selecting the optimal continuous pixel point sequence is to eliminate the interference noise on the left in Figure 2d, or to eliminate some isolated noise in the image;

Step 54, according to the corresponding relationship between the first transposed binary image and the pixel point coordinate transposition of the fourth image, the pixel points corresponding to each row boundary point in the fourth image are recorded as column boundary points;

Here, the column boundary points of the fourth image corresponding to the row boundary points of the first transposed binary image are actually envelope points;

Step 55, sequentially connect the column boundary points to obtain the first connection line; and smooth the first connection line to obtain the first envelope; and complete the marking process on the first envelope on the fourth image.

Here, in order to further eliminate the noise points in the envelope points, smoothing is performed on the connecting line of the envelope points, that is, the first connecting line. During the smoothing process, noises with too short time intervals and large amplitude changes will be eliminated. Envelope points, and finally get a relatively smooth first envelope.

Step 6, performing peak point identification processing on the first envelope to mark a plurality of first peak points;

Specifically include: step 61, taking the vertical distance from each sampling point of the first envelope to the zero line at the bottom of the fourth image as the sampling value of the sampling point, and making statistics on the sampling values of each sampling point of the first envelope to generate The first sample value sequence is {x ₁ , x ₂ ... x _i ... x _n };

Wherein, i is the sampling point index, 1≤i≤n, x _i is the sampling value of each sampling point, and n is the total number of sampling points of the first envelope;

Step 62, constructing a Gaussian kernel weight sliding window;

Specifically include: step 621, setting the sliding window width w of the Gaussian kernel weight sliding window;

Here, the total number of sliding window sampling points in the Gaussian kernel weight sliding window can be set in advance to obtain a preset total number of sampling points, and the preset total number of sampling points can be used as the sliding window width w of the Gaussian kernel weight sliding window;

Step 622, set the sampling value sequence in the Gaussian kernel weight sliding window as {s ₁ ...s _j ...s _w };

Among them, j is the sampling point index in the sliding window, 1≤j≤w, s _j is the sampling value of each sampling point in the sliding window;

Here, because the sliding window width w is the total number of sampling points in the sliding window, so 1≤j≤w;

Step 622, according to the standard Gaussian function

Taking the maximum sampling point index j _max corresponding to the maximum sampling value s _max in the sliding window as the mean value μ, and taking a quarter of the sliding window width w/4 as the variance σ, construct a Gaussian kernel weighted Gaussian kernel for each sampling point in the sliding window The coefficient operation function is

Among them, k _j is the Gaussian kernel coefficient of each sampling point in the Gaussian kernel weight sliding window;

Step 623, according to the operation function of the Gaussian kernel coefficient, construct the sliding window weight operation function of the Gaussian kernel weight sliding window as

Wherein, A is the sliding window weight, and k' _j is the normalized Gaussian kernel coefficient corresponding to each Gaussian kernel coefficient k _j in the sliding window;

Step 63, in the first sample value sequence {x ₁ , x ₂ ... x _i ... x _n }, start from the first sample value x ₁ , take the step size as 1, and take the sliding window width w as the sliding window sampling point number, the first sample value sequence {x ₁ , x ₂ ... x _i ... x _n } is divided into the second number of sub-sliding window sequences C _h ;

Among them, the sub-sliding window sequence C _h is {xi _=h , xi _=h+1 ...xi _=h+w-1 }, h is the sub-sliding window index, 1≤h≤the second number, the second number =n-w+1;

For example, the first sample value sequence is {x ₁ =d ₁ , x ₂ =d ₂ , x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ , x ₆ =d ₆ }, then n=6 , the sliding window width w=4, the second quantity=6-4+1=3, there are 3 sub-sliding window sequences respectively:

C _h=1 {x ₁ =d ₁ , x ₂ =d ₂ , x ₃ =d ₃ , x ₄ =d ₄ },

Ch ₌₂ {x ₂ =d ₂ , x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ }

Ch ₌₃ {x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ , x ₆ =d ₆ };

Step 64, use the Gaussian kernel weight sliding window to perform sliding window weight calculation on each sub-sliding window sequence C _h to obtain the corresponding sliding window weight A _h ;

It specifically includes: converting each sampling value of the current sub-sliding window sequence C _h into the corresponding sampling value s _j in the sliding window, and using the maximum value as the maximum sampling value s _max , and setting the maximum sampling value s _max in the sliding window The index of the sampling point within is taken as the corresponding index of the maximum sampling point j _max ; and the index of the sampling point of the sampling value s _j in each sliding window and the index of the maximum sampling point j _max are substituted into the Gaussian kernel coefficient operation function to obtain multiple Gaussian kernels Coefficient k _j ; and normalize all current Gaussian kernel coefficients k _j to obtain multiple normalized Gaussian kernel coefficients k'_j; and all current normalized Gaussian kernel coefficients k' _j and their corresponding The sampling value s _j in the sliding window is substituted into the sliding window weight calculation function to obtain the corresponding sliding window weight A _h ;

For example, the first sample value sequence is {x ₁ =d ₁ , x ₂ =d ₂ , x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ , x ₆ =d ₆ }, n=6, w=4, the sub-sliding window sequence includes: C _h=1 {x ₁ =d ₁ , x ₂ =d ₂ , x ₃ =d ₃ , x ₄ =d ₄ }, C _h=2 {x ₂ =d ₂ , x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ } and Ch ₌₃ {x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ , x ₆ =d ₆ };

When performing sliding window weight calculation on C _h=1 {x ₁ ＝d ₁ , x ₂ ＝d ₂ , x ₃ ＝d ₃ , x ₄ ＝d ₄ }, 1≤j≤4; the current sub-sliding window Each sampling value of the sequence C _h=1 is converted into the corresponding sampling value s _j in the sliding window, and it is obtained: s ₁ =x ₁ =d ₁ , s ₂ =x ₂ =d ₂ , s ₃ =x ₃ =d ₃ , s ₄ =x ₄ =d ₄ ; if the maximum sampling value is d ₂ , then s _max =s ₂ , and the corresponding j _max =2; index the sampling points of the sampling value s _j in each sliding window (j=1, 2, 3, 4) and the maximum sampling point index j _max = 2, which is substituted into the Gaussian kernel coefficient operation function to obtain multiple Gaussian kernel coefficients k _j :

Gaussian kernel coefficient

Gaussian kernel coefficient

Gaussian kernel coefficient

Gaussian kernel coefficient

For k ₁ , k ₂ , k ₃ and k ₄ , perform normalization processing to obtain the corresponding normalized Gaussian kernel coefficients k' ₁ , k' ₂ , k' ₃ and k'₄; then k' ₁ , k ' ₂ , k' ₃ and k' ₄ and the corresponding s ₁ , s ₂ , s ₃ and s ₄ are brought into the sliding window weight calculation function

Then the sliding window weight A _h=1 =k′ ₁ ×d ₁ +k′ ₂ ×d ₂ +k′ ₃ ×d ₃ +k′ ₄ ×d ₄ can be obtained;

When performing sliding window weight calculation on C _h=2 {x ₂ =d ₂ , x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ }, 1≤j≤4; the current sub-sliding window Each sampling value of the sequence C _h=2 is converted into the corresponding sampling value s _j in the sliding window, and it is obtained: s ₁ =x ₂ =d ₂ , s ₂ =x ₃ =d ₃ , s ₃ =x ₄ =d ₄ , s ₄ ＝x ₅ ＝d ₅ ; if the maximum sampling value is still d ₂ , then s _max =s ₁ , and the corresponding j _max =1; index the sampling points of the sampling value s _j in each sliding window (j=1 , 2, 3, 4) and the maximum sampling point index j _max =1, which is substituted into the Gaussian kernel coefficient operation function to obtain multiple Gaussian kernel coefficients k _j :

Gaussian kernel coefficient

Gaussian kernel coefficient

Gaussian kernel coefficient

Gaussian kernel coefficient

Then the sliding window weight A _h=1 =k′ ₁ ×d ₂ +k′ ₂ ×d ₃ +k′ ₃ ×d ₄ +k′ ₄ ×d ₅ can be obtained;

When performing sliding window weight calculation on C _h=3 {x ₃ =d ₃ , x ₄ =d ₄ , x ₅ =d ₅ , x ₆ =d ₆ }, 1≤j≤4; the current sub-sliding window Each sampling value of the sequence C _h=3 is converted into the corresponding sampling value s _j in the sliding window, and it is obtained: s ₁ =x ₃ =d ₃ , s ₂ =x ₄ =d ₄ , s ₃ =x ₅ =d ₅ , s ₄ =x ₆ =d ₆ ; if the maximum sampling value is d ₆ , then s _max =s ₄ , and the corresponding j _max =4; index the sampling points of the sampling value s _j in each sliding window (j=1, 2, 3, 4) and the maximum sampling point index j _max = 4, which is substituted into the Gaussian kernel coefficient operation function to obtain multiple Gaussian kernel coefficients k _j :

Gaussian kernel coefficient

Gaussian kernel coefficient

Gaussian kernel coefficient

Gaussian kernel coefficient

Then the sliding window weight A _h=1 =k′ ₁ ×d ₃ +k′ ₂ ×d ₄ +k′ ₃ ×d ₅ +k′ ₄ ×d ₆ can be obtained;

Step 65, record the first sampled value sequence {x ₁ , x ₂ ... x _i ... x _n } as the current sequence; and mark the sub-sliding window sequence C _h whose upper sliding window weight A _h is the maximum value in the current sequence, as is the current sub-sliding window sequence; and mark the sampling point index corresponding to the maximum sampling value on the current sub-sliding window sequence as the peak point index P; and use the peak point index P to divide the current sampling value sequence into left and right parts and record it as left Sampled value sequence and right sampled value sequence; and take the left and right sampled value sequence as the new current sampled value sequence respectively, and continue the maximum sampling of the sub-sliding window sequence corresponding to the maximum sliding window weight in the new current sampled value sequence The sampling point index of the value is marked with the peak point index until the sequence length of the new current sampling value sequence is lower than the preset minimum sequence length;

For example, the first sampling value sequence has 5 sub-sliding window sequences C ₁ , C ₂ , C ₃ , C ₄ and C ₅ , and the corresponding sliding window sequences of 5 sub-sliding window sequences C ₁ , C ₂ , C ₃ , C ₄ and C ₅ The size relationship of the window weight is: A ₁ _{_{_{<A 2 <A 3 , A 3 >}}} A ₄ >A ₅ ; then, the largest sliding window weight in the first sampling value sequence is C ₃ , if the sampling value in C ₃ The largest is the second sampling point, then the index of the second sampling point in C ₃ will be recorded as the peak point index; the first sampling value sequence is divided into two parts by the second sampling point of C ₃ and recorded as left, Right sample value sequence; for the left and right sample value sequences, continue to mark the peak point index in the above way until the sequence length of the separated left and right sample value sequences is lower than the minimum sequence length;

Step 66: Use the sampling points corresponding to all peak point indexes P on the first envelope as the first peak point.

Step 7, carry out left and right baseline point recognition processing to each first peak point and mark corresponding first left baseline point and first right baseline point;

It specifically includes: step 71, on the first envelope, using each first peak point as the current peak point;

Step 72, dividing a corresponding left envelope interval and a right envelope interval from the current peak point to the left and right according to the preset time length threshold;

Here, the time length threshold is generally set to half the heart cycle length, that is, the time length threshold = heart cycle length/2; there are many ways to calculate the heart cycle length, and the current peak point and the front and rear peak points The average value of the peak-to-peak distance of the first envelope can be taken as the duration of the heartbeat cycle, and the average value of the peak-to-peak distance of all adjacent peak points on the first envelope can be taken as the duration of the heartbeat cycle;

Step 73, record the minimum envelope amplitude on the left envelope interval and the right envelope interval as the corresponding left interval minimum value and right interval minimum value;

Here, under ideal conditions, when there is no drift in the baseline of the envelope, and the envelope waveform has no local maximum or minimum value caused by burrs or interference, the minimum envelope amplitude on the left and right envelope intervals The corresponding point should be a valley point; but in actual situations, the baseline of the envelope often drifts locally and the envelope waveform may also have local maximum and minimum on the rising or falling edge of the waveform due to glitches or interference value, in this case, the point corresponding to the minimum envelope amplitude on the left and right envelope intervals may be a valley point or a rising or falling edge of the left and right envelope interval boundaries The minimum value point; the reason why the minimum value of the left interval and the minimum value of the right interval are extracted here is to use the two as the reference baseline zero point of the left and right envelope intervals to weaken the baseline caused by baseline drift and envelope waveform burrs point extraction error;

Step 74, starting from the current peak point and traversing the left valley point to the left on the left envelope interval; when traversing, calculating the difference between the amplitude of the current peak point and the minimum value of the left interval to generate the first amplitude difference, Calculate the difference between the amplitude of the current left valley point and the minimum value of the left interval to generate the second amplitude difference, calculate the ratio of the second amplitude difference to the first amplitude difference to generate the first ratio, if the first ratio is less than the preset error range Then use the current left valley point as the first left baseline point corresponding to the current peak point and stop traversing, if the first ratio is greater than or equal to the preset error range, go to the next left valley point to continue traversing;

It should be noted that if there is no left valley point whose first ratio is less than the preset error range on the left envelope interval, it means that all traversed valley points may be caused by glitches or interference on the rising or falling edge of the waveform Local maximum and minimum values, at this time, set the first left baseline point as the sampling point corresponding to the minimum value of the left interval; here, the preset error range can be set by an optimal value obtained after multiple experiments;

Step 75, starting from the current peak point on the right envelope interval, traversing the right valley point to the right; when traversing, calculating the difference between the amplitude of the current peak point and the minimum value of the right interval to generate a third difference, Calculate the difference between the amplitude of the current right valley point and the minimum value of the right interval to generate the fourth difference, calculate the ratio of the fourth difference to the third difference to generate the second ratio, if the second ratio is less than the preset error range Then use the current right valley point as the first right baseline point corresponding to the current peak point and stop traversing, if the second ratio is greater than or equal to the preset error range, go to the next right valley point to continue traversing.

It should be noted that if there is no left valley point whose second ratio is smaller than the preset error range on the right envelope interval, it means that all traversed valley points may be caused by glitches or interference on the rising or falling edge of the waveform. Local maximum and minimum values, at this time, set the first right baseline point as the sampling point corresponding to the minimum value of the right interval; here, the preset error range can be set by an optimal value obtained after multiple experiments.

Step 8, according to the completed peak point and the first envelope marked by the left and right baseline points, perform blood flow parameter calculation to generate a corresponding blood flow parameter group sequence;

Wherein, the blood flow parameter group sequence includes a plurality of blood flow parameter groups; the blood flow parameter group includes a peak flow velocity parameter, a pressure gradient parameter, an acceleration time parameter, a deceleration time parameter, an ejection time parameter, a pressure difference halving time parameter and a speed Time integration parameters; the blood flow parameter group corresponds to the first peak point one by one;

It specifically includes: step 81, on the first envelope, take each first peak point as the current peak point, take the first left baseline point corresponding to the current peak point as the current left baseline point, and take the first left baseline point corresponding to the current peak point The right baseline point is the current right baseline point;

Step 82, taking the vertical distance from the current peak point to the bottom zero line of the fourth image as the corresponding peak distance h, and calculating the corresponding peak flow velocity parameters V _max according to the preset unit peak distance blood flow velocity V _s and peak distance h, Vmax _max =V _s *h;

Here, under normal circumstances, the original two-dimensional spectral Doppler echocardiographic image will have the scale information of longitudinal unit distance and flow velocity, and the fourth image is cut from the two-dimensional spectral Doppler echocardiographic image without shrinking. Zoom-in operation, so the longitudinal unit distance and flow velocity scale information on the original two-dimensional spectral Doppler echocardiography image, that is, the unit peak distance blood flow velocity V _s can be used to multiply the distance from each sampling point to the baseline to obtain the corresponding Sampling point flow rate; then if the sampling point is the peak point, the corresponding sampling point flow rate is the peak flow rate;

Step 83, calculate and generate the corresponding pressure gradient parameter ΔP according to the peak flow velocity parameter V _max , where,

Here, the well-known conversion relationship between the pressure gradient and the peak flow velocity derived from the simplified Bernoulli equation of fluid dynamics is the quadruple relationship of the square of the flow velocity, so the peak flow velocity parameter V _max can be directly brought into the conversion relationship. get the pressure gradient;

Step 84, taking the time interval from the current left baseline point to the current peak point as the corresponding acceleration time parameter T _a ;

Here, the time point corresponding to the current left baseline point can be regarded as the minimum blood flow velocity time point in the current cardiac cycle, and the time point corresponding to the current peak point can be regarded as the maximum blood flow velocity time point in the current cardiac cycle , then the acceleration time parameter T _a of the current blood flow velocity caused by the heart atrioventricular movement in the current cardiac cycle is naturally determined by the time difference between the maximum blood flow velocity time point and the minimum blood flow velocity time point before acceleration;

Step 85, taking the time interval from the current peak point to the current right baseline point as the corresponding deceleration time parameter T _d ;

Here, the time point corresponding to the current peak point can be regarded as the maximum blood flow velocity time point in the current cardiac cycle, and the time point corresponding to the current right baseline point can be regarded as another minimum blood flow velocity in the current cardiac cycle time point, then the deceleration time parameter T _a of the current blood flow velocity caused by the heart atrioventricular movement in the current cardiac cycle is naturally determined by the time difference of the minimum blood flow velocity time point after deceleration minus the maximum blood flow velocity time point Decide;

Step 86, taking the sum of the acceleration time parameter T _a and the deceleration time parameter T _d as the corresponding ejection time parameter T _e ;

Here, the ejection time can be regarded as the acceleration time parameter T _a of the current blood flow velocity from the minimum value to the maximum value and the deceleration time parameter from the maximum value to the minimum value caused by the atrioventricular movement in a single heartbeat cycle sum of _Td ;

Step 87, record the envelope segment from the current peak point to the current right baseline point on the first envelope as the current segment; and on the current segment, traverse the sampling points from the current peak point to the right; when traversing, Take the vertical distance from the current sampling point to the zero line at the bottom of the fourth image as the corresponding sampling point distance h _sam , and calculate and generate the corresponding sampling point flow velocity V _sam =V according to the sampling point distance h _sam and the unit peak distance blood flow velocity V _s _s *h _sam , and calculate and generate the corresponding sampling point pressure gradient according to the sampling point flow velocity V _sam

And calculate the ratio of the pressure gradient △P _sam of the sampling point to the pressure gradient parameter △P to generate the first ratio, if the first ratio enters the preset half-value ratio confirmation range, stop traversing and use the current sampling point as the half-value of the pressure difference Sampling point, if the first ratio has not yet entered the half-value proportional confirmation range, stop and go to the next sampling point to continue traversing; and use the time interval from the current peak point to the pressure difference half-value sampling point as the corresponding pressure difference halving time parameter T _P/2 ;

Here, the half-value sampling point of the pressure difference is actually the sampling point where the pressure gradient is halved relative to the peak point. The ideal value of the first ratio is 0.5, which is difficult to achieve in practical applications, so the embodiment of the present invention defines the ideal value as 0.5 A half-value ratio confirmation range, that is, a floating error range around 0.5, as long as the first ratio enters this range, the corresponding sampling point can be considered as the half-value sampling point of the pressure difference;

Step 88, performing velocity integration calculation on the first envelope segment from the current left baseline point to the current right baseline point to generate corresponding velocity time integration parameters;

Here, the speed-time integral parameter is often used to evaluate the cardiac function strength of the subject;

Step 89, the peak flow velocity parameter V _max , the pressure gradient parameter △P, the acceleration time parameter T _a , the deceleration time parameter T _d , the ejection time parameter T _e , the pressure difference halving time parameter T _△P/2 and the speed time Integrate parameters to form a blood flow parameter group corresponding to the current peak point; and add the blood flow parameter group to the blood flow parameter group sequence.

Step 9, calculate the average value of each similar parameter in the blood flow parameter group sequence, and obtain the average value of peak flow velocity, average value of pressure gradient, average value of acceleration time, average value of deceleration time, average value of ejection time, and halving time of pressure difference The average value and the velocity-time-integrated average value, and the measurement data set composed of all average values is returned as the measurement data result of the two-dimensional spectral Doppler echocardiography image.

FIG. 3 is a block diagram of a two-dimensional spectral Doppler echocardiographic image processing device provided in Embodiment 2 of the present invention. The device may be a terminal device or a server implementing the method of the embodiment of the present invention, or it may be the same as the above-mentioned The device connected to the terminal device or the server to implement the method of the embodiment of the present invention, for example, the device may be the device or chip system of the above-mentioned terminal device or server. As shown in FIG. 3 , the device includes: an acquisition module 201 , an image preprocessing module 202 , an envelope processing module 203 and a blood flow parameter calculation module 204 .

The acquiring module 201 is configured to acquire a two-dimensional spectral Doppler echocardiographic image to generate a first image.

The image preprocessing module 202 is used to perform region-of-interest image extraction processing on the first image to generate a corresponding second image; and perform Gaussian blur image processing on the second image to generate a corresponding third image; and perform binary processing on the third image processing to generate a corresponding fourth image.

The envelope processing module 203 is used to perform spectrum envelope identification processing on the fourth image to mark the corresponding first envelope; and perform peak point identification processing on the first envelope to mark a plurality of first peak points; And perform left and right baseline point identification processing on each first peak point to mark the corresponding first left baseline point and first right baseline point.

The blood flow parameter calculation module 204 is used to perform blood flow parameter calculation and generate a corresponding blood flow parameter set sequence according to the first envelope marked with the peak point and the left and right baseline points; the blood flow parameter set sequence includes multiple blood flow parameter sets ; The blood flow parameter group includes peak flow velocity parameters, pressure gradient parameters, acceleration time parameters, deceleration time parameters, ejection time parameters, pressure difference halving time parameters and velocity time integration parameters; the blood flow parameter group is the same as the first peak point One to one correspondence.

The blood flow parameter calculation module 204 is also used to calculate the average value of each similar parameter in the blood flow parameter group sequence to obtain the average value of peak flow velocity, average pressure gradient, average acceleration time, average deceleration time, and average ejection time , the mean value of the pressure difference halving time and the mean value of the velocity time integration, and the measurement data set composed of all the mean values is returned as the measurement data result of the two-dimensional spectral Doppler echocardiographic image.

An apparatus for processing two-dimensional spectral Doppler echocardiographic images provided by an embodiment of the present invention can execute the method steps in the above method embodiments, and its implementation principle and technical effect are similar, and will not be repeated here.

It should be noted that it should be understood that the division of each module of the above device is only a division of logical functions, and may be fully or partially integrated into one physical entity or physically separated during actual implementation. And these modules can all be implemented in the form of calling software through processing elements; they can also be implemented in the form of hardware; some modules can also be implemented in the form of calling software through processing elements, and some modules can be implemented in the form of hardware. For example, the acquisition module can be a separate processing element, or it can be integrated into a chip of the above-mentioned device. In addition, it can also be stored in the memory of the above-mentioned device in the form of program code, and a certain processing element of the above-mentioned device can Call and execute the functions of the modules identified above. The implementation of other modules is similar. In addition, all or part of these modules can be integrated together, and can also be implemented independently. The processing element described here may be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each module above can be completed by an integrated logic circuit of hardware in the processor element or an instruction in the form of software.

For example, the above modules may be one or more integrated circuits configured to implement the above method, for example: one or more specific integrated circuits (Application Specific Integrated Circuit, ASIC), or, one or more digital signal processors ( Digital Signal Processor, DSP), or, one or more Field Programmable Gate Arrays (Field Programmable Gate Array, FPGA), etc. For another example, when one of the above modules is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (Central Processing Unit, CPU) or other processors that can call program codes. For another example, these modules can be integrated together and implemented in the form of a System-on-a-chip (SOC).

In the above embodiments, all or part of them may be implemented by software, hardware, firmware or any combination thereof. When implemented using software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the processes or functions described according to the embodiments of the present invention will be generated in whole or in part. The above-mentioned computers may be general-purpose computers, special-purpose computers, computer networks, or other programmable devices. The above-mentioned computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. (such as coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (such as infrared, wireless, Bluetooth, microwave, etc.) to another website site, computer, server or data center. The above-mentioned computer-readable storage medium may be any available medium that can be accessed by a computer, or a data storage device such as a server or a data center integrated with one or more available media. The above-mentioned usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, or a magnetic tape), an optical medium (for example, DVD), or a semiconductor medium (for example, a solid state disk (solid state disk, SSD)) and the like.

FIG. 4 is a schematic structural diagram of an electronic device provided by Embodiment 3 of the present invention. The electronic device may be the aforementioned terminal device or server, or may be a terminal device or server connected to the aforementioned terminal device or server to implement the method of the embodiment of the present invention. As shown in FIG. 4 , the electronic device may include: a processor 301 (such as a CPU), a memory 302 , and a transceiver 303 ; Various instructions may be stored in the memory 302 for completing various processing functions and realizing the methods and processing procedures provided in the above-mentioned embodiments of the present invention. Preferably, the electronic device involved in this embodiment of the present invention further includes: a power supply 304 , a system bus 305 and a communication port 306 . The system bus 305 is used to realize the communication connection among the components. The above-mentioned communication port 306 is used for connection and communication between the electronic device and other peripheral devices.

The system bus mentioned in FIG. 4 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (Extended Industry Standard Architecture, EISA) bus or the like. The system bus can be divided into address bus, data bus, control bus and so on. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The communication interface is used to realize the communication between the database access device and other devices (such as client, read-write library and read-only library). The memory may include random access memory (Random Access Memory, RAM), and may also include non-volatile memory (Non-Volatile Memory), such as at least one disk memory.

Above-mentioned processor can be general-purpose processor, comprises central processing unit CPU, network processor (Network Processor, NP) etc.; Can also be digital signal processor DSP, application-specific integrated circuit ASIC, field programmable gate array FPGA or other available Program logic devices, discrete gate or transistor logic devices, discrete hardware components.

It should be noted that the embodiments of the present invention also provide a computer-readable storage medium, and instructions are stored in the storage medium, and when the storage medium is run on a computer, the computer executes the methods and processing procedures provided in the above-mentioned embodiments.

The embodiment of the present invention also provides a chip for running instructions, and the chip is used for executing the method and the processing procedure provided in the foregoing embodiments.

Professionals should further realize that the units and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the relationship between hardware and software Interchangeability. In the above description, the composition and steps of each example have been generally described according to their functions. Whether these functions are executed by hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may use different methods to implement the described functions for each specific application, but such implementation should not be regarded as exceeding the scope of the present invention.

The steps of the methods or algorithms described in connection with the embodiments disclosed herein may be implemented by hardware, software modules executed by a processor, or a combination of both. Software modules can be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other Any other known storage medium.

The specific embodiments described above have further described the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the scope of the present invention. Protection scope, within the spirit and principles of the present invention, any modification, equivalent replacement, improvement, etc., shall be included in the protection scope of the present invention.

Claims

A method for processing two-dimensional spectral Doppler echocardiographic images, characterized in that the method comprises:

acquiring a two-dimensional spectral Doppler echocardiographic image to generate a first image;

performing region-of-interest image extraction processing on the first image to generate a corresponding second image;

performing Gaussian blur image processing on the second image to generate a corresponding third image;

performing binarization processing on the third image to generate a corresponding fourth image;

performing spectrum envelope identification processing on the fourth image to mark the corresponding first envelope;

performing peak point identification processing on the first envelope to mark a plurality of first peak points;

Perform left and right baseline point identification processing on each of the first peak points to mark the corresponding first left baseline point and first right baseline point;

According to the first envelope marked with the peak point and the left and right baseline points, perform blood flow parameter calculation to generate a corresponding blood flow parameter set sequence; the blood flow parameter set sequence includes multiple blood flow parameter sets; the blood flow parameter set sequence includes multiple blood flow parameter sets; The flow parameter group includes a peak flow velocity parameter, a pressure gradient parameter, an acceleration time parameter, a deceleration time parameter, an ejection time parameter, a pressure difference halving time parameter and a velocity time integration parameter; the blood flow parameter group is related to the first peak value One-to-one correspondence;

Calculate the average value of each similar parameter in the blood flow parameter group sequence to obtain the average peak flow velocity, average pressure gradient, average acceleration time, average deceleration time, average ejection time, and average pressure difference halving time value and velocity-time-integrated average, and a measurement data set composed of all the average values is returned as the measurement data result of the two-dimensional spectral Doppler echocardiographic image.
The method for processing two-dimensional spectral Doppler echocardiographic images according to claim 1, wherein said performing region-of-interest image extraction processing on said first image to generate a corresponding second image specifically includes:

Perform blood flow velocity zero line identification processing on the first image to mark the corresponding first zero line;

If the large peak of the spectrum image in the first image is upward, then extract the sub-image from the top of the image to the first zero line in the first image as the first sub-image; if the first image in the first image If the large peak of the spectral image is facing downward, then extract the sub-image from the first zero line to the bottom of the image in the first image, and perform image flip processing on the extracted sub-image to generate the first sub-image; The bottom of the image of the first sub-image is the first zero line;

The sum of the pixel values of each row of pixels in the first sub-image is counted to generate a corresponding first row of pixel sums; and the image row corresponding to the first row of pixel sums with the smallest numerical value is recorded as the smallest pixel row; and extracting the sub-image from the minimum pixel row to the bottom of the image in the first sub-image as an image of the region of interest to generate the second image.
The method for processing two-dimensional spectral Doppler echocardiographic images according to claim 1, characterized in that, performing spectrum envelope recognition processing on the fourth image to mark the corresponding first envelope, specifically include:

Rotate the fourth image to the left by 90° to generate a corresponding first transposed binary image;

Carrying out row-by-row inspection on the first transposed binary image, clustering the continuous pixel points whose pixel values in the current row are preset foreground point pixel values, and generating a corresponding sequence of continuous pixel points;

Optimum sequence screening is carried out to a plurality of said continuous pixel point sequences of the same row, and the maximum number of pixels is used as the optimal continuous pixel point sequence corresponding to the current row; Boundary pixel points are marked as row boundary points;

According to the corresponding relationship between the pixel coordinate transposition of the first transposed binary image and the fourth image, the pixel points corresponding to each of the row boundary points in the fourth image are recorded as column boundary points;

Connecting the column boundary points sequentially to obtain a first connection line; and smoothing the first connection line to obtain the first envelope; and completing the first envelope on the fourth image Marking of threads.
The method for processing two-dimensional spectral Doppler echocardiographic images according to claim 1, characterized in that, performing peak point identification processing on the first envelope to mark a plurality of first peak points specifically includes :

Taking the vertical distance from each sampling point of the first envelope to the zero line at the bottom of the fourth image as the sampling value of the sampling point, performing statistics on the sampling values of each sampling point of the first envelope to generate the first The sampling value sequence is {x 1 , x 2 ... x i ... x n }, i is the sampling point index, 1≤i≤n, x i is the sampling value of each sampling point, n is the first envelope Total number of sampling points;

Construct a Gaussian kernel weight sliding window; set the sliding window width w of the Gaussian kernel weight sliding window; set the sampling value sequence in the Gaussian kernel weight sliding window as {s 1 ... s j ... s w }, j is The index of the sampling point in the sliding window, 1≤j≤w, s j is the sampling value of each sampling point in the sliding window; according to the standard Gaussian function
Taking the maximum sampling point index j max corresponding to the maximum sampling value s max in the sliding window as the mean value μ, and taking a quarter of the sliding window width w/4 as the variance σ, construct the weight of each sampling point in the Gaussian kernel weight sliding window The Gaussian kernel coefficient operation function is
k j is the Gaussian kernel coefficient of each sampling point in the Gaussian kernel weight sliding window; according to the Gaussian kernel coefficient operation function, the sliding window weight operation function of the Gaussian kernel weight sliding window is constructed as
A is the sliding window weight, and k' j is the normalized Gaussian kernel coefficient corresponding to each described Gaussian kernel coefficient k j in the sliding window;

In the first sample value sequence {x 1 , x 2 ... x i ... x n }, starting from the first sample value x 1 , the sliding window is sampled with a step size of 1 and the sliding window width w The number of points, the first sampling value sequence {x 1 , x 2 ... x i ... x n } is divided into the second number of sub-sliding window sequences C h ; the sub-sliding window sequence C h is {xi = h , x i=h +1 ... x i=h+w-1 }, h is the sub-sliding window index, 1≤h≤the second number, the second number=n-w+1;

Use the Gaussian kernel weight sliding window to perform a sliding window weight calculation on each of the sub-sliding window sequences C h ; during the operation, each sampling value of the current sub-sliding window sequence C h is converted into Sampling value s j , and taking the maximum value as the maximum sampling value s max , and taking the sampling point index of the maximum sampling value s max in the sliding window as the corresponding maximum sampling point index j max ; and sampling in each sliding window The sampling point index of the value s j and the maximum sampling point index j max are substituted into the Gaussian kernel coefficient operation function to obtain a plurality of Gaussian kernel coefficients k j ; and normalize all current Gaussian kernel coefficients k j to obtain A plurality of normalized Gaussian kernel coefficients k'j; and all current normalized Gaussian kernel coefficients k' j and their corresponding sampling values s j in the sliding window are substituted into the sliding window weight calculation function to obtain the corresponding The sliding window weight A h of ;

Record the first sequence of sampled values {x 1 , x 2 ... x i ... x n } as the current sequence; The sequence C h is marked as the current sub-sliding window sequence; and the sampling point index corresponding to the maximum sampling value on the current sub-sliding window sequence is marked as the peak point index P; and the peak point index P is used to set the The current sampling value sequence is divided into left and right parts and is recorded as a left sampling value sequence and a right sampling value sequence; Perform peak point index marking processing on the sampling point index of the maximum sampling value of the sub-sliding window sequence corresponding to the maximum sliding window weight until the sequence length of the new current sampling value sequence is lower than the preset minimum sequence length;

Taking the sampling points corresponding to all the peak point indexes P on the first envelope as the first peak point.
The method for processing two-dimensional spectral Doppler echocardiographic images according to claim 1, wherein the left and right baseline point identification processing is performed on each of the first peak points to mark the corresponding first left baseline point and The first right baseline point, specifically including:

On the first envelope, each of the first peak points is the current peak point;

According to the preset time length threshold, a corresponding left envelope interval and a right envelope interval are respectively divided from the current peak point to the left and right;

Record the minimum envelope amplitude on the left envelope interval and the right envelope interval as the corresponding left interval minimum and right interval minimum;

On the left envelope interval, start from the current peak point to traverse the left valley point to the left; when traversing, calculate the difference between the amplitude of the current peak point and the minimum value of the left interval to generate The first difference, calculate the difference between the amplitude of the current left valley point and the minimum value of the left interval to generate the second difference, calculate the ratio of the second difference to the first difference to generate the first difference Ratio, if the first ratio is less than the preset error range, then use the current left valley point as the first left baseline point corresponding to the current peak point and stop traversing, if the first ratio If it is greater than or equal to the preset error range, go to the next left valley point and continue traversing;

On the right envelope line interval, start from the current peak point to the right to traverse the right valley point; when traversing, calculate the difference between the amplitude of the current peak point and the minimum value of the right interval to generate The third difference is to calculate the difference between the amplitude of the current right valley point and the minimum value of the right interval to generate the fourth difference, and to calculate the ratio of the fourth difference to the third difference to generate the second difference. Ratio, if the second ratio is less than the preset error range, then use the current right valley point as the first right baseline point corresponding to the current peak point and stop traversing, if the second ratio If it is greater than or equal to the preset error range, go to the next valley point on the right to continue traversing.
The method for processing two-dimensional spectral Doppler echocardiographic images according to claim 1, wherein the blood flow parameters are calculated and generated according to the first envelope marked with the peak point and the left and right baseline points The corresponding blood flow parameter set sequence specifically includes:

On the first envelope, take each of the first peak points as the current peak point, take the first left baseline point corresponding to the current peak point as the current left baseline point, and take the current peak point as the current left baseline point. The corresponding first right baseline point is the current right baseline point;

Taking the vertical distance from the current peak point to the bottom zero line of the fourth image as the corresponding peak distance h, and calculating the corresponding peak flow velocity parameter V according to the preset unit peak distance blood flow velocity V s and the peak distance h max , V max =V s *h;

Calculate and generate the corresponding pressure gradient parameter ΔP according to the peak flow velocity parameter V max , wherein,

Taking the time interval from the current left baseline point to the current peak point as the corresponding acceleration time parameter T a ;

Taking the time interval from the current peak point to the current right baseline point as the corresponding deceleration time parameter T d ;

Taking the sum of the acceleration time parameter T a and the deceleration time parameter T d as the corresponding ejection time parameter T e ;

Record the envelope segment from the current peak point to the current right baseline point on the first envelope as the current segment; and on the current segment, proceed to the right from the current peak point Sampling point traversal; when traversing, the vertical distance from the current sampling point to the bottom zero line of the fourth image is taken as the corresponding sampling point distance h sam , and according to the sampling point distance h sam and the unit peak distance blood flow velocity V s Calculate and generate the corresponding sampling point flow velocity V sam =V s *h sam , and calculate and generate the corresponding sampling point pressure gradient according to the sampling point flow velocity V sam
And calculate the ratio of the pressure gradient ΔP sam of the sampling point to the pressure gradient parameter ΔP to generate a first ratio, if the first ratio enters the preset half value ratio confirmation range, stop traversing and put the The current sampling point is used as the half-value sampling point of the pressure difference. If the first ratio has not yet entered the half-value ratio confirmation range, it will stop and go to the next sampling point to continue traversing; The time interval between value sampling points is used as the corresponding pressure difference halving time parameter T ΔP/2 ;

performing a velocity integration operation on the first envelope segment from the current left baseline point to the current right baseline point to generate the corresponding velocity time integration parameter;

The peak flow velocity parameter V max , the pressure gradient parameter ΔP, the acceleration time parameter T a , the deceleration time parameter T d , the ejection time parameter T e , and the pressure difference halving time The parameter T ΔP/2 and the velocity-time integral parameter form the blood flow parameter set corresponding to the current peak point; and add the blood flow parameter set to the blood flow parameter set sequence.
A device for realizing the steps of the method for processing two-dimensional spectral Doppler echocardiographic images described in any one of claims 1-6, wherein the device comprises: an acquisition module, an image preprocessing module, a package A network line processing module and a blood flow parameter calculation module;

The acquiring module is used to acquire a two-dimensional spectral Doppler echocardiographic image to generate a first image;

The image preprocessing module is used to perform region-of-interest image extraction processing on the first image to generate a corresponding second image; and perform Gaussian blur image processing on the second image to generate a corresponding third image; and The third image is binarized to generate a corresponding fourth image;

The envelope processing module is used to perform spectrum envelope identification processing on the fourth image to mark the corresponding first envelope; and perform peak point identification processing on the first envelope to mark multiple The first peak point; and performing left and right baseline point identification processing on each of the first peak points to mark the corresponding first left baseline point and first right baseline point;

The blood flow parameter calculation module is used to perform blood flow parameter calculation and generate a corresponding blood flow parameter set sequence according to the first envelope marked with the peak point and the left and right baseline points; the blood flow parameter set sequence includes multiple A blood flow parameter group; the blood flow parameter group includes a peak flow velocity parameter, a pressure gradient parameter, an acceleration time parameter, a deceleration time parameter, an ejection time parameter, a pressure difference halving time parameter and a velocity time integral parameter; The flow parameter group is in one-to-one correspondence with the first peak point;

The blood flow parameter calculation module is also used to calculate the average value of each similar parameter in the blood flow parameter group sequence to obtain the average value of peak flow velocity, average value of pressure gradient, average value of acceleration time, average value of deceleration time, and ejection time. The time average value, the pressure difference halving time average value and the velocity time integral average value, and all the average values form a measurement data set to be returned as the measurement data results of the two-dimensional spectral Doppler echocardiographic image.
An electronic device, characterized in that it includes: a memory, a processor, and a transceiver;

The processor is configured to be coupled to the memory, read and execute instructions in the memory, so as to implement the method steps described in any one of claims 1-6;

The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.
A computer-readable storage medium, characterized in that the computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a computer, the computer executes the method described in any one of claims 1-6. method directive.