M-mode PISA quantitative method for atrioventricular valve regurgitation volume and electronic device

Abstract

The present invention discloses a method for quantitatively measuring the volume of atrioventricular valve regurgitation by M-mode PISA. Three target cardiac ultrasound images are received, the first image is the continuous wave Doppler spectral image of the atrioventricular valve regurgitation jet, the second image is the M-mode PISA color image, and the third image is the M-mode PISA grayscale image. The time curve of the PISA radius is obtained from the second and third images, and the value r of the PISA radius at each time point is calculated. (t) , the velocity-time curve of the regurgitation jet is obtained from the first image, and the velocity V at each time point is calculated. (t) , the aliasing velocity V is extracted from the metadata of the target cardiac ultrasound image. a , the instantaneous flow rate, total regurgitation volume, regurgitation orifice area at each time point, and average regurgitation orifice area are calculated. By automatically tracing the M-mode PISA contour and the continuous wave Doppler boundary of the regurgitation jet and calculating the total regurgitation volume, it can assist doctors in quickly and accurately quantifying the severity of atrioventricular valve regurgitation.

Description

Technical Field

[0001] The present invention relates to the technical fields of medical image processing and cardiac diagnosis, and particularly relates to a method for quantitatively determining the mitral valve regurgitation volume by M-mode PISA and an electronic device. Background Art

[0002] Proximal Isovelocity Surface Area Method (proximal iso-velocity surface area, hereinafter referred to as the PISA method) is an important Echocardiogram method for clinically Quantification determining the severity of valvular regurgitation. For the known classical PISA method, it is assumed that the regurgitant orifice area is constant during cardiac systole, and a single-frame PISA radius (a certain frame during systole) is selected to calculate the regurgitant orifice area and regurgitation volume. However, the size of the regurgitant orifice area and the PISA radius are dynamically changing during cardiac systole. The regurgitant orifice area calculated using the single-frame PISA radius of this method cannot represent the average size of the regurgitant orifice area throughout systole. Therefore, there will be situations where the obtained regurgitation volume and regurgitant orifice area are overestimated or underestimated, and in clinical applications, it sometimes cannot accurately reflect the severity of valvular regurgitation. Summary of the Invention

[0003] One embodiment of the present invention is an echocardiograph that uses an automatic quantitative method for mitral valve regurgitation volume by M-mode PISA based on a deep learning network and has an automatic quantitative function for the regurgitation volume and regurgitant orifice area of the mitral valve or tricuspid valve in an echocardiogram.

[0004] The M-mode PISA automatic quantitative method includes the steps of:

[0005] Receiving three target cardiac ultrasound images, namely,

[0006] The first image - a continuous wave Doppler spectral image of the mitral valve regurgitant jet,

[0007] The second image - an M-mode PISA color image,

[0008] The third image - an M-mode PISA grayscale image;

[0009] Obtaining the time curve of the PISA radius from the second image and the third image, and calculating the value r of the PISA radius at each time point (t) ,

[0010] Obtaining the velocity time curve of the regurgitant jet from the first image, and calculating the velocity V at each time point (t) ,

[0011] Extracting the aliasing velocity V from the metadata of the target cardiac ultrasound images a ,

[0012] The calculation includes

[0013] The instantaneous flow rate at each time point Flow rate (t) = 2πr (t) 2 ×V a (1)

[0014] The total regurgitant volume MRSV = ∫Flow rate (t) dt = 2πV a ∫r (t) 2 dt (2)

[0015] The regurgitant orifice area EROA at each time point (t) = Flow rate (t) / V (t) (3)

[0016] The average regurgitant orifice area EROA mean = ∫EROA (t) dt / t m (4).

[0017] According to the above calculation results, draw the systolic instantaneous flow rate curve and the regurgitant orifice area curve to display the total regurgitant volume.

[0018] The M-mode PISA automatic quantification method for atrioventricular valve regurgitant volume in the embodiments of the present invention can automatically trace the M-mode PISA contour and the continuous wave Doppler boundary of the regurgitant jet through an algorithm, and calculate the systolic instantaneous flow rate curve, the regurgitant orifice area curve, and the total regurgitant volume, which can assist ultrasound doctors or clinicians to quickly and accurately quantify the severity of atrioventricular valve regurgitation and reduce the problem of poor internal and external stability of the measurer. Description of the Drawings

[0019] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understood. In the drawings, several embodiments of the present invention are shown by way of example and not limitation, wherein:

[0020] Figure 1 is a flowchart of the M-mode PISA quantification method for atrioventricular valve regurgitant volume according to one of the embodiments of the present invention.

[0021] Figure 2 is a schematic diagram of the implementation process of the M-mode PISA quantification method for atrioventricular valve regurgitant volume according to one of the embodiments of the present invention. Detailed Embodiments

[0022] M-mode echocardiogram uses a single ultrasound beam to scan the heart and displays the motion of the heart and large blood vessels in the form of a curve formed by a group of light spots changing over time. M-mode PISA uses M-mode ultrasound to record the change in PISA radius over time in a two-dimensional PISA dynamic image. Related studies have confirmed that M-mode PISA can reveal the change patterns of the regurgitant orifice area and the instantaneous flow rate in patients with mitral regurgitation of different etiologies, and the regurgitant volume and the mean regurgitant orifice area derived from it are more reliable.

[0023] M-mode PISA records the change in the PISA radius during systole on one image using M-mode ultrasound, which contains time information and has high spatial resolution. M-mode PISA images can be used to calculate the instantaneous flow rate, regurgitant orifice area of the atrioventricular valve at each time point, as well as the total regurgitant volume and the mean regurgitant orifice area.

[0024] The application of M-mode PISA in quantitatively determining the regurgitant volume and regurgitant orifice area of the atrioventricular valve has encountered the following problems:

[0025] M-mode PISA requires calculating the PISA radius at multiple time points and the peak velocity of the regurgitant jet at the corresponding time points, which is time-consuming and laborious; the existing M-mode PISA output does not have the functions of tracing and calculation. Using PISA technology requires certain professional knowledge and a learning curve, and manual tracing and calculation may have problems with poor internal and external stability of the measurer.

[0026] According to one or more embodiments, as Figure 1 shown, the M-mode PISA quantitative method for the regurgitant volume of the atrioventricular valve includes the following steps:

[0027] S101, receiving a clinical ultrasound image;

[0028] S102, performing preparatory processing on the obtained clinical ultrasound image, including continuous wave Doppler spectral images, M-mode PISA images, and color-removed M-mode PISA grayscale images;

[0029] S103, performing image segmentation based on a deep learning network to obtain the upper boundary of the PISA radius time curve, the motion trajectory of the valve leaf, and the contour of the regurgitant jet velocity time curve;

[0030] S104, image processing, extracting the M-mode PISA contour;

[0031] S105, automatic quantification, instantaneous flow rate curve, regurgitant orifice area curve, total regurgitant volume.

[0032] In an embodiment of the present disclosure, a method for quantitatively determining the volume of atrioventricular valve regurgitation takes three cardiac ultrasound images as inputs (namely, the continuous wave Doppler spectral image of the atrioventricular valve regurgitation jet, the M-mode PISA image, and its derived grayscale image without color), and based on a deep learning network, extracts contour curves such as the continuous wave Doppler spectrum, M-mode PISA, and atrioventricular valve motion trajectory in the images, obtains the PISA radius and regurgitation jet velocity information at each time point, and finally calculates the instantaneous flow at each time point and performs integral calculation to obtain the regurgitation volume. By extracting the contour curves, the M-mode PISA image and its derived grayscale image without color are input into the deep neural network in a two-channel manner, rather than processing the two images separately, so as to effectively utilize the complementary information of the two images to extract the upper and lower contours on the M-mode PISA simultaneously. Furthermore, due to the adoption of an algorithm model based on a deep learning network, this method has good repeatability and high calculation efficiency and is suitable for popularization.

[0033] The embodiment of the present disclosure has the characteristics of being simple and easy to implement. The images required for implementation are the complete M-mode PISA images collected by a cardiac ultrasound diagnostic instrument (it is recommended to collect 2 to 3 cardiac cycles, Nyquist velocity limit 30 - 40 cm / s, image format DICOM), the color-removed images obtained from the same image on the machine (image format DICOM), and also include the continuous wave Doppler images of the regurgitation jet (it is recommended to collect 2 to 3 cardiac cycles, image format DICOM). Such images occupy a small space and contain time information and are easy to analyze.

[0034] The method of the embodiment of the present disclosure realizes the automation of quantitatively determining the regurgitation volume of the M-mode PISA, that is, the user or doctor only needs to input the images, and this method can automatically calculate the results without additional input of other information or other adjustments to the images.

[0035] According to one or more embodiments, a method for quantitatively calculating the volume of atrioventricular valve regurgitation includes the steps,

[0036] S201, obtain the complete M-mode PISA images of the target heart (it is recommended to collect 2 to 3 cardiac cycles, Nyquist velocity limit 30 - 40 cm / s, image format DICOM), the color-removed images obtained from the same image on the machine (image format DICOM), and also include the continuous wave Doppler images of the regurgitation jet (it is recommended to collect 2 to 3 cardiac cycles, image format DICOM).

[0037] S202. Train a segmentation model using a deep learning network. Train the segmentation model according to a loss function to optimize the model parameters. The network training input is the M-mode PISA image, the color-removed image, the continuous wave Doppler image of the regurgitant jet, and the corresponding manually annotated mask image in the training set. The network output is the time curve profile of the predicted PISA radius and the mask map of the continuous wave Doppler velocity profile.

[0038] S203. Use the trained deep learning network to identify the upper boundary of the PISA radius time curve on the M-mode PISA image. This boundary is the spatio-temporal position where the regurgitant jet accelerates and reaches the first aliasing velocity recorded by M-mode ultrasound. During ultrasound image acquisition, it appears as the junction between the two-dimensional grayscale (endocardial echo in the left ventricular cavity) and the PISA radius time curve on the M-mode PISA image, which is conducive to identification to improve the accuracy of the algorithm.

[0039] S204. Use the trained deep learning network to identify the movement trajectory of the valve leaf (mitral valve or tricuspid valve) on the color-removed image. This movement trajectory will serve as the lower profile of the PISA radius time curve. The echo of the valve leaf movement trajectory is easily identifiable on this type of image.

[0040] S205. Use an algorithm to fuse the images labeled in steps S203 and S204 above, and calculate the intersection of the labels to obtain the time curve profile of the PISA radius, and extract the value r of the PISA radius at each time point. (t) 。

[0041] Use the trained U-net algorithm to identify the profile of the regurgitant jet velocity time curve on the continuous wave Doppler image, and extract the velocity V at each time point. (t) 。And extract the aliasing velocity V from the DICOM metadata. a 。

[0042] Instantaneous flow rate formula at each time point: Flow rate (t) =2πr (t) 2 ×V a ,

[0043] Total regurgitant volume calculation formula: MRSV=∫Flow rate (t) dt=2πV a ∫r (t) 2 dt,

[0044] Regurgitant orifice area calculation formula at each time point: EROA (t) =Flow rate (t) / V (t) ,

[0045] The average regurgitant orifice area is calculated as: EROA mean = ∫ EROA (t) dt / t m ,

[0046] S206. Draw the systolic instantaneous flow curve and the regurgitant orifice area curve according to the calculation result of step S205 to display the total regurgitant volume.

[0047] In the embodiments of the present disclosure, based on artificial intelligence technology, an automatic M-mode PISA quantitative method for atrioventricular valve regurgitant volume is implemented. It mainly obtains the upper contour and the lower contour of the PISA time-radius curve, and the velocity-time curve of the regurgitant jet respectively according to a deep neural network, and calculates the PISA radius, the instantaneous flow, and the regurgitant orifice area at each moment, and calculates the total regurgitant volume based on time integration.

[0048] Since accurately evaluating the severity of atrioventricular valve regurgitation is crucial for determining the optimal surgical timing of surgical valve repair. The existing PISA method commonly used clinically calculates the regurgitant orifice area and the regurgitant volume based on the single-frame PISA radius, without considering the situation of the atrioventricular valve regurgitant orifice changing with time, and cannot obtain an accurate estimate of the regurgitant volume. Therefore, it may underestimate or overestimate the severity of the regurgitation.

[0049] The difference between M-mode PISA and the regurgitant volume obtained by cardiac magnetic resonance is smaller and the correlation is better than that of the traditional PISA method. M-mode PISA makes up for the defects of calculating the regurgitant orifice area and the regurgitant volume based on the single-frame PISA radius, and has high accuracy. However, when calculating at multiple time points, there are problems of complex and time-consuming operations and high requirements for the professional level of the operator.

[0050] Therefore, the beneficial effects of this invention are as follows:

[0051] 1) The method of this invention calculates the regurgitant orifice area and the regurgitant volume based on M-mode PISA, overcomes the errors brought by the single-frame PISA radius calculation method clinically, and draws the instantaneous flow curve and the change curve of the regurgitant orifice area, which can provide more information on the valve regurgitation mechanism and assist in diagnosis.

[0052] 2) The method of this invention is based on an artificial intelligence algorithm, and the whole process is fully automatic without the need to input other information additionally. Therefore, this method saves time and effort, can assist ultrasonic doctors or clinicians to quickly and accurately quantify the severity of atrioventricular valve regurgitation, and reduces the problem of poor internal and external stability of the measurer.

[0053] 3) The M-mode PISA of the present invention can be automatically calculated by an artificial intelligence algorithm, which will greatly reduce time and manpower and open up a new field for MR evaluation.

[0054] It should be understood that in the embodiments of the present invention, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0055] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0056] In addition, in each embodiment of the present invention, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0057] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present invention. The aforementioned storage medium includes various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0058] As mentioned above, the above are only the specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for M-mode PISA quantification of atrioventricular valve regurgitation volume, characterized in that, it includes the steps of: Receiving three target cardiac ultrasound images, namely, The first image - the continuous wave Doppler spectral image of the atrioventricular valve regurgitation jet, The second image - the M-mode PISA color image, The third image - the M-mode PISA grayscale image; Obtain the time curve of the PISA radius from the second image and the third image, and calculate the value r of the PISA radius at each time point (t) , Obtain the velocity-time curve of the regurgitant jet from the first image, and calculate the velocity V at each time point (t) , extract the aliasing velocity V from the metadata of the target cardiac ultrasound image a , The calculation includes, Instantaneous flow rate at each time point (t) = 2πr (t) 2 × V a (1) Total regurgitant volume MRSV = ∫ Flow rate (t) dt = 2πV a ∫r (t) 2 dt (2) EROA of the reflux orifice area at each time point (t) = Flow rate (t) / V (t) (3) Average effective regurgitant orifice area (EROA) mean = ∫ EROA (t) dt / t m (4) Plotting the systolic instantaneous flow curve and the regurgitant orifice area curve to display the total regurgitation volume, Obtaining the upper curve of the PISA radius-time curve by identifying in the second image through a learning model, Obtaining the lower curve of the PISA radius-time curve by identifying in the third image through a learning model, Fusing the upper curve and the lower curve to obtain the PISA radius-time curve, Obtaining the regurgitation jet velocity-time curve by identifying in the first image through a learning model.

2. The method according to claim 1, characterized in that, The learning model is constructed based on a deep learning network.

3. The method according to claim 1, characterized in that, The training sample set of the learning model includes manually annotated mask images.

4. The method according to claim 1, characterized in that, The cardiac ultrasound image format is DICOM.

5. The method according to claim 1, characterized in that, The sampling period of the first image, the second image or the third image is 2 to 3 cardiac cycles.

6. An electronic device for M-mode PISA quantification of atrioventricular valve regurgitation volume, the electronic device includes a memory; and A processor coupled to the memory, the processor being configured to execute instructions stored in the memory to implement the method according to any one of claims 1 to 5.

7. An echocardiograph, characterized in that, The main controller of the echocardiograph calculates the atrioventricular valve regurgitation volume of the heart by using the method according to any one of claims 1 to 5.

Images