CN108882917A - A kind of heart volume discriminance analysis system and method - Google Patents

A kind of heart volume discriminance analysis system and method Download PDF

Info

Publication number
CN108882917A
CN108882917A CN201680082174.4A CN201680082174A CN108882917A CN 108882917 A CN108882917 A CN 108882917A CN 201680082174 A CN201680082174 A CN 201680082174A CN 108882917 A CN108882917 A CN 108882917A
Authority
CN
China
Prior art keywords
image
frame
location
left ventricle
endocardium
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201680082174.4A
Other languages
Chinese (zh)
Inventor
王勃
金蒙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Mindray Bio Medical Electronics Co Ltd
Original Assignee
Shenzhen Mindray Bio Medical Electronics Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shenzhen Mindray Bio Medical Electronics Co Ltd filed Critical Shenzhen Mindray Bio Medical Electronics Co Ltd
Publication of CN108882917A publication Critical patent/CN108882917A/en
Pending legal-status Critical Current

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/06—Measuring blood flow
    • A61B8/065—Measuring blood flow to determine blood output from the heart
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/08—Detecting organic movements or changes, e.g. tumours, cysts, swellings
    • A61B8/0883—Detecting organic movements or changes, e.g. tumours, cysts, swellings for diagnosis of the heart
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/48—Diagnostic techniques
    • A61B8/486—Diagnostic techniques involving arbitrary m-mode
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/48—Diagnostic techniques
    • A61B8/488—Diagnostic techniques involving Doppler signals
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/5223—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for extracting a diagnostic or physiological parameter from medical diagnostic data
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment

Abstract

A kind of heart volume identifying and analyzing method and system, the method includes：Obtain the multiple frames of ultrasonic image (501, S10) of the continuous acquisition on predetermined amount of time；Identify the section type (S12, S202, S203) of multiple frames of ultrasonic image data (501) cardiac；It identifies cardiac cycle (S14, S204, S205)；Identify the location and shape (S16, S206, S207, S208) of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image；According to the location and shape of the endocardium of left ventricle, the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image is calculated, is obtained ventricular volume curve (701, S18, S209)；It exports ventricular volume curve (701), and/or the clinical parameter (S19, S210, S211, S212, S213) of characterization cardiac function is calculated and exported according to the ventricular volume curve (701).This method does not need the operation of connection electrocardiosignal conducting wire and module in actual use, simplifies the workload of user, improves work efficiency.

Description

A kind of heart volume discriminance analysis system and method Technical field
The present invention relates to field of medical technology more particularly to a kind of heart volume discriminance analysis system and method.
Background technique
Ejection fraction (EF, Ejection Fraction) refers to that stroke output accounts for the percentage of end-diastolic volume amount, is one of the important clinical index for evaluating left ventricular function.The size of ejection fraction is related with the contractility of cardiac muscle, and cardiac contractility ability is stronger, then stroke output is more, and ejection fraction is bigger, and Left Ventricular Ejection Fraction is > 50% under normal circumstances.Measuring Left Ventricular Ejection Fraction can be by multiple means, wherein most of use the method based on medical imaging devices.Pass through medical imaging device such as CT (Computed Tomography first, computer tomography) equipment, MRI (Magnetic Resonance Imaging, magnetic resonance imaging) equipment and ultrasonic device be acquired cardiac image, after the image for obtaining a complete cardiac cycle, endocardium of left ventricle is split and is identified on each frame image, then ventricular volume is calculated according to endocardial shape, after ventricular volume on obtaining a cardiac cycle on each frame image, construct ventricular volume curve, then ejection fraction is calculated according to maximum value, that is, end-diastolic volume (EDV) of ventricular volume curve and minimum value, that is, end-systolic volume (ESV).In above-mentioned medical imaging devices, echocardiography is a kind of noninvasive safe diagnostic method, injection contrast agent, isotope or other dyestuffs are not needed, patient and doctor are not radiated by radioactive substance, method is easy, can be repeated several times, can carry out by bed, it can be to each chambers of the heart inspection, the anatomical structure and function of the entire heart of complete evaluation by more planes, multi-faceted ultrasonic imaging.The Echocardiographic pattern of currently used left ventricular ejection fraction measurement has M-Mode mode, B-Mode mode.Based on the left ventricular ejection fraction measurement method of M-Mode mode by, according to being imaged, the EDV and ESV of ventricular volume then being obtained by calibration left ventricle maximum inner diameter and minimum diameter, to calculate ejection fraction to the line number on beveling long axis of left ventricle section.Based on the left ventricular ejection fraction measurement method of B-Mode mode by carrying out the left ventricle two dimensional image under the different sections of imaging acquisition to left ventricle, then it is identified according to frame of the image to left ventricular contraction latter stage and diastasis, then endocardial position is demarcated manually, EDV and ESV is calculated, the calculating to ejection fraction is finally completed.Left ventricular ejection fraction measurement method based on M-Mode has biggish dependence to the position of scan line, and for the difference of the heart of Different Individual, it is very difficult to collect the left room long axial images of standard, is difficult the scan line position of the measurement left ventricular interior diameter of acquisition standard It sets, the method according to radial line estimation volume is also not accurate enough.Therefore B-Mode mode is the clinical ventricular ejection fraction measurement method recommended.Under B-Mode mode, the all automatic measurement to ventricular ejection fraction may be implemented at present, the relevant technologies position cardiac cycle by electrocardiosignal, different heart phases is identified by electrocardiosignal, the calculating to parameters such as the measurement of ventricular end systolic volume, end-diastolic volume and ventricular ejection fractions is then realized using image Segmentation Technology.Since all automatic measurement technology of current ventricular ejection fraction requires electrocardiosignal, when using the technology every time, the operation for being attached electrocardiosignal line, electrocardiosignal module is required, the workload of user is increased, reduces the working efficiency of user.
Summary of the invention
The present invention provides a kind of heart volume discriminance analysis system and method, can reduce the workload of user, and improve working efficiency.
As an aspect of of the present present invention, a kind of heart volume identifying and analyzing method is provided, wherein include:
Obtain the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time；
Identify the section type of above-mentioned multiple frames of ultrasonic image data cardiac；
Identify cardiac cycle；
Based on above-mentioned section type, the location and shape of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image are identified；
According to the location and shape of above-mentioned endocardium of left ventricle, the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image is calculated, obtains ventricular volume curve；
Ventricular volume curve is exported, and/or the clinical parameter of characterization cardiac function is calculated and exported according to the ventricular volume curve.
Accordingly, as another aspect of the present invention, a kind of heart volume discriminance analysis system is additionally provided, wherein include:
Ultrasound image acquisition module, for obtaining the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time；
Section type identification module identifies the section type of above-mentioned multiple frames of ultrasonic image data cardiac；
Cardiac cycle identification module, cardiac cycle for identification；
Profile obtains module, for identification the location and shape of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image；
Ventricular volume curve generation module, for the location and shape according to above-mentioned endocardium of left ventricle, meter The ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image is calculated, ventricular volume curve is obtained；
Clinical parameter output module for exporting ventricular volume curve, and/or calculates according to the ventricular volume curve and exports the clinical parameter of characterization cardiac function.
It is of the invention in one embodiment, additionally provide a kind of heart volume discriminance analysis system, wherein include:
Probe；
Transmit circuit, for emitting ultrasonic beam to target object；
Circuit and Beam synthesis module are received, for obtaining ultrasound echo signal；
Image processing module, for obtaining the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time according to ultrasound echo signal, identify the section type of above-mentioned ultrasound image cardiac, identify cardiac cycle, and identify the location and shape of endocardium of left ventricle in every frame ultrasound image in a cardiac cycle, and the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image is calculated, obtain ventricular volume curve；And
Display, for showing the location and shape of above-mentioned ultrasound image and ventricular volume curve, the above-mentioned endocardium of left ventricle of label, and the above-mentioned section type of display.
The present invention proposes a kind of heart volume discriminance analysis system and method.The identification for the automatic cardiac cycle that the present invention is realized instead of the prior art based on electrocardiosignal using the characteristic and image processing techniques of heart movement and the judgement of heart movement phase.The operation for not needing connection electrocardiosignal conducting wire and module in actual use, simplifies the workload of user, improves work efficiency.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, the drawings to be used in the description of the embodiments or prior art will be briefly described below, apparently, drawings in the following description are only some embodiments of the invention, for those of ordinary skill in the art, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of structural schematic diagram of supersonic imaging apparatus provided by the invention；
Fig. 2 is a kind of one embodiment flow diagram of heart volume identifying and analyzing method provided by the invention；
Fig. 3 is the more detailed flow chart of step S12 in Fig. 2；
Fig. 4 is the schematic diagram for identifying cardiac cycle in Fig. 2 in one embodiment of step S14；
Fig. 5 is to carry out the schematic diagram that edge extracting obtains endocardial contours based on characteristic point in one embodiment of step S16 in Fig. 2；
Fig. 6 is that machine learning network obtains the schematic diagram of endocardial contours in another embodiment of step S16 in figure；
Fig. 7 is the schematic diagram of the ventricular volume curve obtained in one embodiment of step S18 in Fig. 2；
Fig. 8 is the flow diagram of one of embodiment of the invention；
Fig. 9 is a kind of structural schematic diagram of one embodiment of heart volume discriminance analysis system provided by the invention；
Figure 10 is the structural schematic diagram of one embodiment of Fig. 9 the midsagittal plane type identification module；
Figure 11 is the structural schematic diagram of another embodiment of Fig. 9 the midsagittal plane type identification module；
Figure 12 is the structural schematic diagram of one embodiment of Fig. 9 heart cycles identification module；
Figure 13 is the structural schematic diagram of another embodiment of Fig. 9 heart cycles identification module；
Figure 14 is the structural schematic diagram of one embodiment that profile obtains module in Fig. 9.
Specific embodiment
The noise as present in ultrasound image, artifact, and the structural complexity of certain anatomicals, there are different section types for tangent plane picture corresponding to it, such as, for cardiac ultrasound images, usually there is different section types, such as: the various section types of apex of the heart cor biloculare, the apical four-chamber heart, in different heart sections types, all there is biggish difference in the endocardial location and shape of left ventricle.Therefore for if there are the cardiac ultrasound image datas of a variety of section types when carrying out automatic identification internal membrane of heart region using fixed shape progress image segmentation operation, certainly exist image zooming-out error, and it is difficult to take into account the difference of internal membrane of heart location and shape between a variety of heart sections in automatic identification, to cause measurement result error occur to endocardial segmentation and identification inaccuracy.
The all automatic measurement to ventricular ejection fraction is had been carried out at present, the relevant technologies position cardiac cycle by electrocardiosignal, different heart phases is identified by electrocardiosignal, is then realized using image Segmentation Technology to the measurement of ventricular end systolic volume, end-diastolic volume and the calculating of ventricular ejection fraction.By the way that the internal membrane of heart is split and is identified on the ultrasound image of ventricular end systolic and diastasis then ventricular volume calculation formula can be utilized to the measurement of ventricular end systolic volume, end-diastolic volume, such as: Simpson method, to calculate ventricular volume.It follows that directly determining the accuracy of ventricular volume measurement to the accuracy of endocardial segmentation and identification for B-Mode mode.Therefore, the accuracy that realization accurately measures lifting system endocardial segmentation and identification is most important.
Fig. 1 provides a kind of system structure diagram of acquiring ultrasound image equipment.Herein to obtain heart System structure is described in detail for ultrasound.As shown in Figure 1, the device for carrying out ultrasonic imaging to target area of the embodiment of the present invention includes: probe 1, transmit circuit 2, transmitting/reception selection switch 3, receives circuit 4, Beam synthesis module 5, signal processing module 6, image processing module 7 and display 8.Transmit circuit 2 by what is focused by delay there is certain amplitude and polar ultrasonic pulse to select switch 3 to be sent to probe 1 by transmitting/reception.Excitation of the probe 1 by ultrasonic pulse, it (is not shown in the figure to the target area of tested body tissue, such as heart tissue) transmitting ultrasonic wave, it receives after certain time-delay from the reflected ultrasonic echo with organizational information in target area, and this ultrasonic echo is converted into electric signal again.The electric signal that the conversion of circuit receiving transducer 1 generates is received, obtains ultrasound echo signal, and these ultrasound echo signals are sent into Beam synthesis module 5.Beam synthesis module 5 is focused the processing such as delay, the summation of weighted sum channel to ultrasound echo signal, and ultrasound echo signal feeding signal processing module 6 is then carried out relevant signal processing.Image processing module 7 is sent by the ultrasound echo signal that signal processing module 6 is handled.The difference of the imaging pattern according to needed for user of image processing module 7, different processing is carried out to signal, obtains the image data of different mode, then the ultrasound image of different mode is formed through processing such as log-compressed, dynamic range adjustment, digital scan conversions, such as B image, C image, D image etc..The ultrasound image that image processing module 7 generates is sent into display 8 and is shown, for example, in the display interface can be with the ultrasound image of simultaneous display ventricular end systolic and diastasis, and endocardial contours can be sketched the contours on the ultrasound image.The ultrasound image of ventricular end systolic and diastasis for simultaneous display can be ultrasound cross-section image corresponding to the ultrasound cross-section image (such as apex of the heart cor biloculare, the apical four-chamber heart) of standard or the arbitrary tangent of user's selection.
Furthermore, it further include operation control module 9 in system shown in FIG. 1, equipment user by operation control module 9 can input control order in the display interface, such as on the ultrasound image Introduced Malaria silhouette markup, comment mark text, carry out the operational orders such as pattern switching.
Based on above system structure, a kind of heart volume identifying and analyzing method and system are provided in an embodiment of the present invention.The workload of user can be reduced, and improves working efficiency.
Fig. 2 is a kind of flow diagram of one embodiment of heart volume identifying and analyzing method provided by the invention；As shown, this method comprises:
Step S10, obtain the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time, the multiple frames of ultrasonic image may include the multiple frames of ultrasonic cardiac image of continuous acquisition, and can also improve ground includes the combination from one of B-mode ultrasonic heart film, the super heart film of M-mode etc. or both.Such as in one of implementation of the invention, be length at least 3 seconds ultrasonic heart films, image frame per second is not less than 25 frames/second, therefore the predetermined time can be greater than 3 seconds.Multiple frames of ultrasonic image in the present embodiment can be with It is the ultrasound image data obtained in real time, is also possible to the ultrasound image data of caching or remote transmission acquisition.If it is the ultrasound image data obtained in real time, then before step S10 further include:
Firstly, emitting ultrasonic beam to the heart area of target object；Then, ultrasound echo signal is obtained, the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time is obtained according to the ultrasound echo signal.
Step S12 identifies the section type of above-mentioned multiple frames of ultrasonic image data cardiac.It is of the invention in one embodiment, above-mentioned section type includes the standard section in medical anatomy or ultrasonic imaging to target object, for example, including but is not limited to for heart tissue: the sections type such as the four chamber hearts, cor biloculare.Certainly, above-mentioned section type is also not necessarily limited to standard section, can also include customized section type, selects any direction to carry out the ultrasound cross-section image obtained after cutting to target object for example, customized section type can be user.Here ultrasound image can be, but not limited to obtain only with above-mentioned system shown in FIG. 1.In this step, the above-mentioned section type identified can be shown in display interface.
Step S14 identifies cardiac cycle.It is of the invention in one embodiment, to above-mentioned multiple frames of ultrasonic image data carry out analysis obtain cardiac cycle.After completing to the judgement of heart sections type, needs to analyze the multiple image in ultrasonic heart film, identify cardiac cycle；The specific method of the identification cardiac cycle will be described in detail below.Here ultrasonic film or ultrasonic movie file can be understood as one of the multiple frames of ultrasonic image of the continuous acquisition on the predetermined amount of time form of expression or storage form.
Step S16 identifies the location and shape of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image.Position includes position coordinates, the azimuth information etc. that the internal membrane of heart identified is shown in ultrasound image, may include the co-ordinate position information of one or more pixels.Shape includes the model parameter for simulating internal membrane of heart monnolithic case, the model parameter includes the basic parameter and deformation parameter for expressing the model configurations such as circle, ellipse, and deformation parameter includes warp parameters, zooming parameter, manual or automatic adjustment parameter, extensograph parameter etc..Certainly, for shape here in addition to being indicated with model parameter, the co-ordinate position information that one group of discrete or continuous pixel can also be obtained using the position identified is used to characterize the shape of the endocardium of left ventricle identified.
Step S18 calculates the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image according to the location and shape of above-mentioned endocardium of left ventricle, obtains ventricular volume curve.
Step S19 according to the clinical parameter of above-mentioned ventricular volume curve computational representation cardiac function, and is exported；And/or output ventricular volume curve.Wherein, it is the end-diastolic volume (EDV) in current cardiac cycle that maximum value is searched on ventricular volume curve, searches the end-systolic volume (ESV) that minimum value is current cardiac cycle.According to EDV and ESV, it can the ejection fraction (EF) of left ventricle is calculated, The important clinical parameter of the characterization cardiac function such as stroke output and cardiac output.These clinical parameters can be output on display and be shown, the mode of display can be in such a way that text is shown.Person can also be exported by the way of voice prompting.
It is following each step in Fig. 2 to be described in detail.
Fig. 3 is the more detailed flow chart of step S12 in Fig. 2；In this embodiment, step S12 includes:
Step S120 identifies the position of interventricular septum in ultrasound image；
Step S122 rotates ultrasound image according to the position of above-mentioned interventricular septum, keeps the long axis of left ventricle direction in ultrasound image vertical；
Step S124 translates above-mentioned ultrasound image, and the above-mentioned left ventricular location in above-mentioned ultrasound image is adjusted to the center of image.Above-mentioned steps S120 to step S124 can be regarded as the normalized process of ultrasound image.In normalized treatment process, the ultrasound image in step 120 may include a frame image, can also include to each frame in the multiple image of part, can also include every frame ultrasound image in above-mentioned multiple image.
One frame or multiple frames of ultrasonic image data are mapped to feature space by step S126, and features described above space is constructed by extracting to the feature in training set image.
It is understood that in some instances, which includes at least the corresponding cardiac ultrasound images of various section types such as cor biloculare section, Four-chamber view.Feature space can be constructed by extracting to the feature in training set image, the method that feature extraction can use principal component analysis, the HAAR feature that image can also be extracted can also be and extract anatomical tissue structure feature in heart, to construction feature space.
Step S128, the projection by above-mentioned ultrasound image in feature space are compared with projection of the training image of known section type in feature space, determine the section type of an above-mentioned frame or multiple frames of ultrasonic image.Specifically, the image to be classified of (i.e. identification of the above-mentioned steps S120 into step S124, rotation and translation) is compared in the projection of feature space with projection of the training image in feature space after having normalized, and image to be classified can be classified and be identified to image type to be sorted using arest neighbors or k nearest neighbor method.
It can be understood that, the present invention in one embodiment, the section type of a frame ultrasound image on predetermined amount of time in the multiple frames of ultrasonic image of continuous acquisition can be equal to the corresponding section type of multiple frames of ultrasonic image of the above-mentioned continuous acquisition on predetermined amount of time；Or in other examples, it is above-mentioned more can also to be selected based on the section type of every frame ultrasound image in above-mentioned multiple frames of ultrasonic image for a determination Then the corresponding section type of frame ultrasound image passes through Voting Algorithm, the final section type for determining film for example, can identify to multiple image.
Therefore, in one embodiment of the invention, by the feature of a frame or multiple frames of ultrasonic image in above-mentioned multiple frames of ultrasonic image, it is compared with the feature of the training image of known section type, to obtain the section type of the multiple frames of ultrasonic image data.Feature in the present embodiment may include the positional relationship of cut zone in image (such as anatomical tissue structure), image pixel value, image pixel value distribution situation, the shape size that region (such as anatomical tissue structure) is given in figure, can be used for extracting in the picture may include as the information of characteristics of image identification feature in the present embodiment column.
In some instances, the step of above-mentioned steps S14, identification cardiac cycle, specifically, can be realized by following method:
Firstly, extracting the characteristic value of every frame ultrasound image, Characteristi c curve of formation according to the multiple frames of ultrasonic image of the above-mentioned continuous acquisition on predetermined amount of time；
Then, periodicity analysis is carried out to features described above curve, identifies the cardiac cycle of object.
The present invention in one embodiment, the characteristic value of above-mentioned every frame image is likeness coefficient, and features described above curve is likeness coefficient curve.
Specifically, indicatrix can be image similarity curve.Similarity curve generate method be, a certain frame in the selected cardiac ultrasonic film being loaded into is as standard frame (501), the likeness coefficient of each the frame image and standard frame in the cardiac ultrasonic film being loaded into is calculated, is generated likeness coefficient curve (503).Wherein, the method for calculating the likeness coefficient of each frame image and standard frame in the cardiac ultrasonic movie file being loaded into may is that first, calculate each frame in cardiac ultrasonic movie file in each pixel and standard frame the absolute value of the difference of corresponding pixel points gray value and, using the addition and value as the likeness coefficient for describing similarity degree between two field pictures；The second, all regard each frame image in cardiac ultrasonic movie file as a matrix, element value of the value of each pixel as matrix, likeness coefficient of the positive correlation coefficient as similarity degree between description two field pictures between calculating matrix in image.Further, since the pixel number of original image is larger, more time is needed when calculating likeness coefficient, original image can be carried out down-sampled, original image is reduced to a suitable scale, while not losing image information, reduces the time required for calculating likeness coefficient.Furthermore, it is possible to which the regional area chosen in image calculates likeness coefficient, the regional area in image that can be chosen can make: the bicuspid valve region etc. in interventricular septum region, cardiac ultrasound images in cardiac ultrasound images.Choose the regional area in image can be further reduced calculate image between the time required for likeness coefficient.
The present invention in one embodiment, it is above-mentioned in the characteristic value of the every frame ultrasound image of said extracted The characteristic value of every frame image includes the image measurements such as tissue anatomical structure measured value, and features described above curve is the image measurement curve that image measurement changes over time.
Specifically, indicatrix also may include the image measurement curve that tissue anatomical structure measured value changes over time, these measured values include but is not limited to: long axis of left ventricle length, left ventricular area, the anatomical structures measured value such as left ventricular mass or right ventricle volume.
Image measurement curve is the curve that tissue anatomical structure measured value changes over time, such as the curve that left ventricular mass changes over time, the curve that left ventricular area changes over time, the curve etc. that the curve or right ventricle volume that long axis of left ventricle length changes over time change over time.
For example, firstly generating the preliminary profile of left ventricle during generating image measurement parameter curve.Preliminary profile can describe roughly the metamorphosis of left ventricle, but can not need the levels of precision for reaching Pixel-level.The method for generating preliminary profile can pass through, and according to the left room of positioning feature points substantially forms such as the apex of the heart and annulus of mitral valve, can also carry out endocardial Boundary Extraction by low-resolution image, to obtain above-mentioned measured value, such as long axis of left ventricle length, left ventricular area, left ventricular mass etc..Measured value based on acquisition changes with time, and obtains image measurement curve as indicatrix, for obtaining cardiac cycle.
In above-mentioned steps S14, after obtaining indicatrix, cardiac cycle can be identified according to this feature curve.Specifically, the identification of cardiac cycle can be carried out by following several ways:
First: choosing a frame ultrasound image as standard frame, identify the complete cycle where standard frame according to indicatrix.Wherein, the method of selection standard frame can be chooses a frame image as standard frame in the ultrasonic film of loading at random, perhaps a frame image is chosen as standard frame in the multiple frames of ultrasonic image of loading at random and also can choose a certain frame image during heart contraction or diastole as standard frame；
Local extremum will be searched in regional area near the moment where standard frame, and determine another frame ultrasound image corresponding to local extremum；As the beginning and end of time at the time of respectively corresponding with standard frame, a cardiac cycle is obtained.Specifically, a period is determined as the beginning and end of time at the time of using in ultrasonic film corresponding to standard frame and identified another frame image or multiple frames of ultrasonic image, and the heart movie file or part multiple frames of ultrasonic image in the period can be determined as a cardiac cycle.As shown in figure 4, showing an example that.Wherein, local extremum is searched near the moment where standard frame in regional area, and determines ultrasound image 402 corresponding to local extremum.A period 404 is determined as the beginning and end of time at the time of using in ultrasonic film corresponding to standard frame 401 and identified another frame image 402, and the heart movie file in the period can be determined as a cardiac cycle.
Second: the method for identifying cardiac cycle can also be the period according to indicatrix, then according to mark The position of quasi- frame and cycle length determine a complete cardiac cycle.The method for calculating the period of indicatrix, which can be, transforms to frequency domain for the similarity curve in time domain using Fourier transformation, obtains the frequency spectrum of indicatrix, peak value is searched in spectrogram, and the period of indicatrix is then calculated according to spectrum peak.The method for calculating the period of indicatrix can also be the auto-correlation coefficient curve for calculating indicatrix, according to the location determination period of auto-correlation coefficient peak of curve.
Wherein, in step s 16, the step of identifying the location and shape of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image can include:
A frame image is chosen as key frame, image segmentation is carried out based on endocardium of left ventricle of the above-mentioned section type to above-mentioned key frame, obtains standard endocardial parted pattern；
According to above-mentioned standard internal membrane of heart parted pattern, the location and shape of the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle are identified.
It is of the invention in one embodiment, carrying out image segmentation based on endocardium of left ventricle of the above-mentioned section type to above-mentioned key frame can use conventional methods based on characteristic point progress edge extracting, edge extracting is carried out based on characteristic point, generates the endocardial location and shape of key frame；The schematic diagram that edge extracting obtains endocardial contours is carried out based on characteristic point as shown in figure 5, showing in one embodiment of step S16；In the ultrasound cardiac images 501 containing the standards such as heart cor biloculare or Four-chamber view section, according to the position of the two of detected annulus of mitral valve key points 503,504, generate endocardial original shape and position 502, then the maximum of points 505 of several topography's gradients is detected near endocardial original shape, and endocardial location and shape are then generated according to the maximum of points of detected topography's gradient and two key points of annulus of mitral valve.
It is understood that it is of the invention in one embodiment, divide and identify the method that can also use machine learning to the internal membrane of heart, be identified according to overall model, and identify the endocardial location and shape of key frame.Implementation can with but be not limited to the left ventricle partitioning algorithm based on deep learning, e.g., the method for convolutional neural networks (CNN) plus linear regression, as shown in Figure 6.A frame image is inputted, CNN carries out feature extraction using the layer-by-layer convolution of convolution collecting image, finally estimates final profile in the feature of extraction using linear regression.Wherein CNN and the model parameter of linear regression have utilized training set training to obtain.
Certainly, in some embodiments of others of the invention, the combination that can also be both conventional method and machine learning is divided and identified to the internal membrane of heart.Traditional method for extracting key point can be used, left ventricle is positioned, endocardial accurate extraction is carried out by the method for machine learning on this basis.After can also identifying the internal membrane of heart according to overall model again, in conjunction with traditional boundary extraction algorithm, boundary is advanced optimized.
After obtaining the internal membrane of heart of key frame, according to above-mentioned standard internal membrane of heart parted pattern, the step of identifying the location and shape of the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle, can use following several methods:
First: after obtaining the internal membrane of heart of key frame, other ultrasound images in cardiac cycle in addition to key frame can be divided by identical partitioning algorithm frame by frame, obtain the location and shape of the endocardium of left ventricle of other each frames in cardiac cycle.
Second: based on the endocardial location and shape of key frame divided, the tracking for carrying out endocardial motion to other ultrasound images in cardiac cycle in addition to key frame is handled, and obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image.
Specifically, the method for motion tracking can be based on Block- matching.Such as the internal membrane of heart curve discrete for obtaining segmentation is at several trace points, image on present image using centered on the trace point in a certain size neighborhood is as original block, on next frame image centered on the trace point, region of search is constructed in a certain size neighborhood, according to grey similarity selection and the matched object block of original block in region of search, then the position using the center of object block as trace point on next frame image, obtaining each trace point of present frame available location and shape endocardial on next frame image behind the position on next frame image.Endocardial location and shape in a cardiac cycle on each frame image can be obtained by all carrying out aforesaid operations to each frame image in an identified cardiac cycle.What motion tracking was also possible to realize based on optical flow method or other track algorithms.
It is understood that it is of the invention in one embodiment, after obtaining the internal membrane of heart of key frame, generate cardiac cycle in other endocardial methods of frame image, can also be the joint of segmentation and motion tracking.The internal membrane of heart of each frame, which all combines, divides and tracks double result progress Intelligent Fusion, to obtain Optimal Boundary.The method of fusion both can be result and carry out linear combined, such as directly carry out being averaged for position；It is also possible to carry out the intelligent selection of regional area according to the confidence level of both sides' result, carries out the fusion of result after selection again.
In step S18, according to the location and shape of above-mentioned endocardium of left ventricle, the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image is calculated, obtains ventricular volume curve；
Specifically, the endocardial location and shape on each frame image are obtained, so that it may carry out the quantitative analysis of left ventricle accordingly.Wherein, quantitative analysis parameter includes but is not limited to left ventricular internal diameter, area, volume etc..Internal diameter may include electrical path length in long axis direction, is normally defined the apex of the heart to the distance at annulus of mitral valve midpoint, may also comprise the length of short-axis direction, such as annulus of mitral valve interventricular septum side endpoint to the distance of left room side wall side endpoint.Left ventricular area can be calculated by the way that chambers of the heart interior pixels are the methods of cumulative.Left ventricular mass For that can be calculated by the methods of Simpson method or Area length method.Volume parameter can be divided into left ventricle overall volume (abbreviation bulk of left ventricle) and left room local volume (local volumetric) again.Local volumetric is the bulking value in each segment of cardiac muscle, segment division mode can use U.S.'s echocardiogram association (American Society of Echocardiography,) and 16 segments that define of American Heart Association (American Heart Association, AHA) or 17 Segment Models ASE.According to above scheme, ventricle quantitative parameter at the time of each frame image corresponds in current cardiac cycle is calculated, can be obtained the ventricle quantitative analysis Parameters variation curve in current cardiac cycle.As shown in fig. 7, showing a kind of bulk of left ventricle curve synoptic diagram.
In step S19, according to the clinical parameter of above-mentioned ventricular volume curve computational representation cardiac function, and export.It is understood that can show moment corresponding ultrasound image locating for the maximum value and minimum value on above-mentioned ventricular volume curve simultaneously, and mark the location and shape for showing above-mentioned endocardium of left ventricle in step S19.Specifically, using EDV as the diastasis moment of current cardiac cycle at the time of corresponding in ventricular volume curve, position a frame image of diastasis in cardiac ultrasonic film, then using ESV at the time of corresponding in ventricular volume curve as the end-systole moment of current cardiac cycle, the frame image for positioning end-systole in cardiac ultrasonic film, then shows the diastasis of current cardiac cycle and end systole image and internal membrane of heart location and shape in the display interface.
It can be found in Fig. 7, the correspondence frame of the available current cardiac cycle end-diastolic volume of maximum value 702 and diastasis within cardiac cycle is searched in ventricular volume change curve 701, searches the correspondence frame of the available current cardiac cycle end-systolic volume of minimum value 703 and end-systole within cardiac cycle.According to end-diastolic and shrink last ventricular volume, it can calculate the ejection fraction of left ventricle, stroke output and cardiac output etc. characterize the important clinical index of cardiac function.
It is understood that it is of the invention in one embodiment, in step s 12, further includes: prompt user can modify to the above-mentioned section type of display；When section type after user's Introduced Malaria, above-mentioned section type is updated.
It is of the invention in one embodiment, the above method further include: the above-mentioned ventricular volume curve being calculated of display, and/or the clinical parameter of characterization cardiac function.
It is of the invention in one embodiment, after above-mentioned steps S14 further include: be switched to manual input mode after cardiac cycle recognition failures, the cardiac cycle being manually entered for obtaining user.
It is of the invention in one embodiment, include: in above-mentioned steps S16
Show above-mentioned key frame；
Prompt user can manually do the image segmentation process of endocardium of left ventricle in above-mentioned key frame In advance；
Based on adjustment of the user on above-mentioned key frame or input results, above-mentioned standard internal membrane of heart parted pattern is obtained.
It is of the invention in one embodiment, after above-mentioned steps S16, further includes:
The segmentation result for determining the location and shape of endocardium of left ventricle jumps to manual input mode, to prompt user to input the location and shape of endocardium of left ventricle on the ultrasound image when determining that segmentation result is wrong.
It is of the invention in one embodiment, the above method further include:
Prompt user that can current display and output result be confirmed and/or be corrected,
Confirmation and/or amendment based on user, output report is formed, amendment content includes at least: the location and shape of the endocardium of left ventricle marked at the time of diastasis corresponds to, at the time of end-systole corresponds to, on ultrasound image and one of the clinical parameter for characterizing cardiac function.
Correspondingly, one of embodiment of the invention is provided shown in Fig. 8.In the present embodiment, a more detailed flow chart is shown.It is first loaded into ultrasound image (step S201).After being loaded into ultrasound image, system is identified and is determined (step S202) to the heart sections type in ultrasound image, and the results are shown on display interface, if current section type decision mistake, (step S203) is modified to judgement result by user.After completing to the judgement of heart sections type, system analyzes the multiple image in ultrasonic heart film, realizes the identification (step S204) of cardiac cycle.If system automatic identification fails, cardiac cycle (step S205) is specified manually by user.After when completing the identification of cardiac cycle, system chooses a frame image automatically, carries out endocardial automatic segmentation (step S206), and determine automatic segmentation result.If the system decide that the internal membrane of heart of automatic identification is wrong, then the internal membrane of heart (step S207) is manually entered on a certain frame image by user.After automatically identifying the accurate internal membrane of heart or being manually entered the internal membrane of heart by user, system identifies the location and shape (step S208) of the endocardium of left ventricle on each frame image in a cardiac cycle using endocardial position and coronary cycle information.Ventricular volume and other quantitative analysis parameters according to the location and shape of the endocardium of left ventricle on each frame image, at the time of corresponding to each frame image of system-computed.According to it is several in a cardiac cycle when ventricular volume that engraves, obtain ventricular volume curve (step S209).It is the end-diastolic volume (EDV) in current cardiac cycle that maximum value is searched on ventricular volume curve, searches the end-systolic volume (ESV) that minimum value is current cardiac cycle.According to EDV and ESV, it can calculate the ejection fraction (EF) of left ventricle, stroke output and cardiac output etc. characterize the important clinical parameter of cardiac function.Meanwhile using EDV it is corresponding in ventricular volume curve at the time of as current cardiac The diastasis moment in period, position a frame image of diastasis in cardiac ultrasonic film, then using ESV at the time of corresponding in ventricular volume curve as the end-systole moment of current cardiac cycle, the frame image for positioning end-systole in cardiac ultrasonic film, then shows (step S210) for the diastasis of current cardiac cycle and end systole image and internal membrane of heart location and shape in the display interface.User needs to be determined (step S211) to current results, if customer acceptance current results, current results can be input in Final Report (step S213) by user.If user does not approve current results, user needs manual modification current results (step S212), at the time of the content of modification includes: diastasis or end-systole and the endocardial location and shape of end-systole or diastasis.After user modifies to result, current results can be input in Final Report (step S213) by user.
As shown in figure 9, being a kind of structural schematic diagram of one embodiment of heart volume discriminance analysis system provided by the invention；In this embodiment, which includes:
Ultrasound image acquisition module 10, for obtaining the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time；
Section type identification module 11 identifies the section type of above-mentioned multiple frames of ultrasonic image data cardiac；
Cardiac cycle identification module 12, cardiac cycle for identification；
Profile obtains module 13, for the location and shape based on the endocardium of left ventricle in every frame ultrasound image in above-mentioned one cardiac cycle of section type identification；
Ventricular volume curve generation module 14 calculates the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image for the location and shape according to above-mentioned endocardium of left ventricle, obtains ventricular volume curve；
Clinical parameter output module 15 for exporting ventricular volume curve, and/or calculates according to the ventricular volume curve and exports the clinical parameter of characterization cardiac function；
It shows mark module 16, for showing moment corresponding ultrasound image locating for the maximum value and minimum value on above-mentioned ventricular volume curve, and marks the location and shape for showing above-mentioned endocardium of left ventricle.
It as shown in Figure 10, is the structural schematic diagram of one embodiment of Fig. 9 the midsagittal plane type identification module；
Above-mentioned section type identification module 11 further comprises:
Section type display module 110, for showing the above-mentioned section type identified.
Section type modifies cue module 111, for prompting user that can modify to the above-mentioned section type of display；
Section type update module 112 updates above-mentioned section type when for section type after user's Introduced Malaria.
It as shown in figure 11, is the structural representation of another embodiment of Fig. 9 the midsagittal plane type identification module Figure；In this embodiment, above-mentioned section type identification module 11 further comprises:
Location identification module 113, for identification in ultrasound image interventricular septum position；
Rotation processing module 114 keeps the long axis of left ventricle direction in ultrasound image vertical for being rotated according to the position of above-mentioned interventricular septum to ultrasound image；
It translates processing module 115 and the above-mentioned left ventricular location in above-mentioned ultrasound image is adjusted to the center of image for translating above-mentioned ultrasound image.
Feature space mapping block 116, for a frame or multiple frames of ultrasonic image data to be mapped to feature space, features described above space is constructed by extracting to the feature in training set image；
Compare determining module 117, is compared for the projection by above-mentioned ultrasound image in feature space with projection of the training image of known section type in feature space, determines the section type of an above-mentioned frame or multiple frames of ultrasonic image.
It as shown in figure 12, is the structural schematic diagram of one embodiment of Fig. 9 heart cycles identification module；In this embodiment, above-mentioned cardiac cycle identification module 12 includes:
Indicatrix generation module 120 extracts the characteristic value of every frame image, Characteristi c curve of formation for the multiple frames of ultrasonic image according to the above-mentioned continuous acquisition on predetermined amount of time；
First identification module 121 identifies the cardiac cycle of object for carrying out periodicity analysis to features described above curve.
It as shown in figure 13, is the structural schematic diagram of another embodiment of Fig. 9 heart cycles identification module；
In this embodiment, above-mentioned cardiac cycle identification module 12 includes:
Standard frame chooses module 122, for choosing a frame ultrasound image as standard frame；
Second identification module 123 for that will search for local extremum in regional area near the moment where standard frame, and determines ultrasound image corresponding to local extremum；As the beginning and end of time at the time of respectively corresponding with standard frame, a cardiac cycle is obtained.
As shown in figure 14, the structural schematic diagram of one embodiment of module is obtained for profile in Fig. 9.In this embodiment, above-mentioned profile acquisition module 13 includes:
Key frame profile obtains module 130, for choosing a frame image as key frame, carries out image segmentation based on endocardium of left ventricle of the above-mentioned section type to above-mentioned key frame, obtains standard endocardial parted pattern；
Other frame profiles obtain module 132, for identifying the location and shape of the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle according to above-mentioned standard internal membrane of heart parted pattern.
Wherein, above-mentioned key frame profile obtains module 130 and obtains the endocardial location and shape of key frame using following either types:
Edge extracting is carried out based on characteristic point, generates the endocardial location and shape of key frame；And/or
Using the method for machine learning, the endocardial location and shape of key frame are identified according to overall model.
Wherein, other frame profiles obtain the location and shape that module 132 obtains the endocardium of left ventricle on each frame ultrasound image using following either types:
According to above-mentioned standard internal membrane of heart parted pattern, other ultrasound images in cardiac cycle in addition to key frame are divided frame by frame, obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image；Or
Based on the endocardial location and shape of the key frame heart divided, the tracking for carrying out endocardial motion to other ultrasound images in cardiac cycle in addition to key frame is handled, and obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image.
More details, reference can be made to the aforementioned description to Fig. 1 to Fig. 8, a bit without being described in detail.
Based on system structure shown in FIG. 1, it is of the invention in one embodiment, additionally provide a kind of heart volume discriminance analysis system, wherein include:
Probe；
Transmit circuit, for emitting ultrasonic beam to target object；
Circuit and Beam synthesis module are received, for obtaining ultrasound echo signal；
Image processing module, for obtaining the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time according to ultrasound echo signal, identify the section type of above-mentioned ultrasound image cardiac, identify cardiac cycle, and identify the location and shape of endocardium of left ventricle in every frame ultrasound image in a cardiac cycle, and the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image is calculated, obtain ventricular volume curve；And
Display, for showing the location and shape of above-mentioned ultrasound image and ventricular volume curve, the above-mentioned endocardium of left ventricle of label, and the above-mentioned section type of display.
The corresponding relationship of above-mentioned various components or module is referring to the explanation in earlier figures 1.Image processing module executes each step process in Fig. 2, is not repeated to tire out herein and state, can be found in the explanation of related content above for the specific explanations of each step process.Image processing module referred to herein can be constituted using a piece of processor or multiple-slice processor.
The corresponding relationship of above-mentioned various components or module is referring to the explanation in earlier figures 1.Image processing module executes each step process in Fig. 2, is not repeated to tire out herein and state, can be found in the explanation of related content above for the specific explanations of each step process.Image processing module referred to herein can use a piece of Processor or multiple-slice processor are constituted.
It is of the invention in one embodiment, aforementioned display device prompt user's identified cardiac cycle it is whether wrong, above system further include:
For receiving the operation control module of user's input control order,
When display prompts user above-mentioned cardiac cycle is wrong, image processing module switches to manual input mode, the cardiac cycle that user can be manually entered by operating control module.
It is of the invention in one embodiment, aforementioned display device shows that the recognition result of above-mentioned section type confirms for user, above system further include:
For receiving the operation control module of user's input control order,
When the control command of user's input is recognition result confirmation mistake, image processing module switches manual input mode, modifies current section type through the control command of operation control module input according to user and shows.
It is of the invention in one embodiment, aforementioned display device shows that the location and shape of the endocardium of left ventricle of every frame ultrasound image confirm for user, above system further include:
For receiving the operation control module of user's input control order,
When the control command of user's input is recognition result confirmation mistake, image processing module switches manual input mode, location and shape and display according to user through endocardium of left ventricle on the control command modification ultrasound image of operation control module input.
It is of the invention in one embodiment, aforementioned display device shows that current display and output result confirm and/or correct, above system for user further include:
For receiving the operation control module of user's input control order,
Confirmation and/or amendment of the image processing module based on user through operation control module input are ordered, output report is formed, amendment content includes at least: the location and shape of the endocardium of left ventricle marked at the time of diastasis corresponds to, at the time of end-systole corresponds to, on ultrasound image and one of the clinical parameter for characterizing cardiac function.
Aforementioned display device further displays the above-mentioned ventricular volume curve being calculated in one embodiment of the invention, and/or the clinical parameter of characterization cardiac function.
It is of the invention in one embodiment, described image processing module identifies cardiac cycle in the following manner:
According to the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time, the characteristic value of every frame image, Characteristi c curve of formation are extracted；
Periodicity analysis is carried out to the indicatrix, identifies the cardiac cycle of object.
The characteristic value of every frame image includes anatomical structure measured value in one embodiment of the invention, and the indicatrix is the curve that anatomical structure measured value changes over time；Or
The characteristic value of every frame image is likeness coefficient, and the indicatrix is likeness coefficient curve.
It is of the invention in one embodiment, described image processing module identifies cardiac cycle in the following manner:
A frame ultrasound image is chosen as standard frame；
Local extremum will be searched in regional area near the moment where standard frame, and determine ultrasound image corresponding to local extremum；As the beginning and end of time at the time of respectively corresponding with standard frame, a cardiac cycle is obtained.
It is of the invention in one embodiment, described image processing module identifies the location and shape of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image in the following manner:
A frame image is chosen as key frame, image segmentation is carried out based on endocardium of left ventricle of the section type to the key frame, obtains standard endocardial parted pattern；
According to the standard endocardial parted pattern, the location and shape of the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle are identified.
It is of the invention in one embodiment, described image processing module carrys out the location and shape that the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle is identified according to the standard endocardial parted pattern in the following manner:
According to the standard endocardial parted pattern, other ultrasound images in cardiac cycle in addition to key frame are divided frame by frame, obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image；Or
Based on the endocardial location and shape of the key frame heart divided, the tracking for carrying out endocardial motion to other ultrasound images in cardiac cycle in addition to key frame is handled, and obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image.
It is of the invention in one embodiment, described image processing module chooses a frame image as key frame automatically in the following manner, image segmentation is carried out based on endocardium of left ventricle of the section type to the key frame, obtains standard endocardial parted pattern:
Edge extracting is carried out based on characteristic point, generates the endocardial location and shape of key frame；And/or
Using the method for machine learning, the endocardial location and shape of key frame are identified according to overall model.
It is of the invention in one embodiment, described image processing module identifies the section type of the multiple frames of ultrasonic image data cardiac in the following manner:
One frame or multiple frames of ultrasonic image data are mapped to feature space, the feature space is constructed by extracting to the feature in training set image；
Projection by the ultrasound image in feature space is compared with projection of the training image of known section type in feature space, determines the section type of a frame or multiple frames of ultrasonic image.
It is of the invention in one embodiment, described image processing module is in the following manner handled ultrasound image:
Identify the position of interventricular septum in ultrasound image；
Ultrasound image is rotated according to the position of the interventricular septum, keeps the long axis of left ventricle direction in ultrasound image vertical；
The ultrasound image is translated, the left ventricular location in the ultrasound image is adjusted to the center of image.
The present invention proposes a kind of heart volume discriminance analysis system and method.The identification for the automatic cardiac cycle that the present invention is realized instead of the prior art based on electrocardiosignal using the characteristic and image processing techniques of heart movement and the judgement of heart movement phase.The operation for not needing connection electrocardiosignal conducting wire and module in actual use, simplifies the workload of user, improves work efficiency.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, it is that relevant hardware can be instructed to complete by computer program, above-mentioned program can be stored in computer-readable storage medium, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, above-mentioned storage medium can be magnetic disk, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access Memory, RAM) etc..
The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be said that specific implementation of the invention is only limited to these instructions.For those of ordinary skill in the art to which the present invention belongs, without departing from the inventive concept of the premise, a number of simple deductions or replacements can also be made, all shall be regarded as belonging to protection scope of the present invention.

Claims (34)

  1. A kind of heart volume identifying and analyzing method, wherein include:
    Obtain the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time；
    Identify the section type of the multiple frames of ultrasonic image data cardiac；
    Identify cardiac cycle；
    Based on the section type, the location and shape of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image are identified；
    According to the location and shape of the endocardium of left ventricle, the ventricular volume quantitative parameter at moment corresponding to every frame ultrasound image is calculated, obtains ventricular volume curve；
    Ventricular volume curve is exported, and/or the clinical parameter of characterization cardiac function is calculated and exported according to the ventricular volume curve.
  2. A kind of heart volume identifying and analyzing method as described in claim 1, wherein the method also includes:
    Moment corresponding ultrasound image locating for the maximum value and minimum value on the ventricular volume curve is shown simultaneously, and marks the location and shape for showing the endocardium of left ventricle；And/or
    Show the section type identified.
  3. A kind of heart volume identifying and analyzing method as claimed in claim 2, wherein in the step of the section type of the identification multiple frames of ultrasonic image data cardiac further include:
    Prompt user can modify to the section type of display；
    When section type after user's Introduced Malaria, the section type is updated.
  4. A kind of heart volume identifying and analyzing method as described in claim 1, wherein the step of the location and shape of the endocardium of left ventricle in one cardiac cycle of the identification in every frame ultrasound image, comprising:
    A frame image is chosen as key frame, image segmentation is carried out based on endocardium of left ventricle of the section type to the key frame, obtains standard endocardial parted pattern；
    According to the standard endocardial parted pattern, the location and shape of the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle are identified.
  5. A kind of heart volume identifying and analyzing method as claimed in claim 4, wherein according to the standard endocardial parted pattern, the step of identifying the location and shape of the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle includes:
    According to the standard endocardial parted pattern, to other ultrasounds in cardiac cycle in addition to key frame Image is divided frame by frame, obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image；Or
    Based on the endocardial location and shape of the key frame heart divided, the tracking for carrying out endocardial motion to other ultrasound images in cardiac cycle in addition to key frame is handled, and obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image.
  6. A kind of heart volume identifying and analyzing method as claimed in claim 4, wherein, as key frame, the step of being carried out image segmentation based on endocardium of left ventricle of the section type to the key frame, obtained standard endocardial parted pattern includes: the one frame image of automatic selection
    Edge extracting is carried out based on characteristic point, generates the endocardial location and shape of key frame；And/or
    Using the method for machine learning, the endocardial location and shape of key frame are identified according to overall model.
  7. A kind of heart volume identifying and analyzing method as described in claim 1, wherein the step of section type of the identification multiple frames of ultrasonic image data cardiac includes:
    It by the feature of a frame or multiple frames of ultrasonic image in the multiple frames of ultrasonic image, is compared with the feature of the training image of known section type, to obtain the section type of the multiple frames of ultrasonic image.
  8. A kind of heart volume identifying and analyzing method as described in claim 1, wherein the step of section type of the identification multiple frames of ultrasonic image data cardiac includes:
    One frame or multiple frames of ultrasonic image are mapped to feature space, the feature space is constructed by extracting to the feature in training set image；
    Projection by the ultrasound image in feature space is compared with projection of the training image of known section type in feature space, determines the section type of a frame or multiple frames of ultrasonic image.
  9. A kind of heart volume identifying and analyzing method as claimed in claim 7, wherein the feature of the frame or multiple frames of ultrasonic image by the multiple frames of ultrasonic image, before the step of being compared with the feature of the training image of known section type further include:
    Identify the position of interventricular septum in ultrasound image；
    Ultrasound image is rotated according to the position of the interventricular septum, keeps the long axis of left ventricle direction in ultrasound image vertical；
    The ultrasound image is translated, the left ventricular location in the ultrasound image is adjusted to the center of image.
  10. A kind of heart volume identifying and analyzing method as claimed in claim 7, wherein the step of the section type of the determination one frame or multiple frames of ultrasonic image further include:
    By the section type of the frame ultrasound image in the multiple frames of ultrasonic image, it is equal to the corresponding section type of the multiple frames of ultrasonic image；Or
    Based on the section type of every frame ultrasound image in the multiple frames of ultrasonic image, selects one and determine the corresponding section type of the multiple frames of ultrasonic image.
  11. A kind of heart volume identifying and analyzing method as described in claim 1, wherein the step of the identification cardiac cycle, comprising:
    According to the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time, the characteristic value of every frame image, Characteristi c curve of formation are extracted；
    Periodicity analysis is carried out to the indicatrix, identifies cardiac cycle.
  12. A kind of heart volume identifying and analyzing method as claimed in claim 11, wherein the characteristic value of every frame image includes anatomical structure measured value, and the indicatrix is the curve that anatomical structure measured value changes over time；Or
    The characteristic value of every frame image is likeness coefficient, and the indicatrix is likeness coefficient curve.
  13. A kind of heart volume identifying and analyzing method as described in claim 1, wherein the step of identification cardiac cycle includes:
    A frame ultrasound image is chosen as standard frame；
    Local extremum will be searched in regional area near the moment where standard frame, and determine ultrasound image corresponding to local extremum；As the beginning and end of time at the time of respectively corresponding with standard frame, a cardiac cycle is obtained.
  14. A kind of heart volume identifying and analyzing method as described in claim 1, wherein the method also includes:
    The ventricular volume curve being calculated described in display, and/or the clinical parameter of characterization cardiac function.
  15. A kind of heart volume identifying and analyzing method as described in claim 1, wherein after the step of identifying cardiac cycle further include:
    It is switched to manual input mode after cardiac cycle recognition failures, the cardiac cycle being manually entered for obtaining user.
  16. A kind of heart volume identifying and analyzing method as claimed in claim 4, wherein include: in the step of the location and shape of the endocardium of left ventricle in one cardiac cycle of the identification in every frame ultrasound image
    Show the key frame；
    Prompt user can be to the image segmentation process progress manual intervention of endocardium of left ventricle in the key frame；
    Based on adjustment of the user on the key frame or input results, the standard endocardial segmentation is obtained Model.
  17. A kind of heart volume identifying and analyzing method as described in claim 1 or 4, wherein, the step of location and shape of endocardium of left ventricle in one cardiac cycle of the identification in every frame ultrasound image, or described image segmentation is carried out to the endocardium of left ventricle of the key frame based on the section type, after the step of obtaining standard endocardial parted pattern, further includes:
    The segmentation result for determining the location and shape of the endocardium of left ventricle jumps to manual input mode, to prompt user to input the location and shape of endocardium of left ventricle on the ultrasound image when determining that segmentation result is wrong.
  18. A kind of heart volume identifying and analyzing method as claimed in claim 2, wherein the method also includes:
    Prompt user can confirm and/or correct to current display and output result；
    Confirmation and/or amendment based on user, output report is formed, amendment content includes at least: the location and shape of the endocardium of left ventricle marked at the time of diastasis corresponds to, at the time of end-systole corresponds to, on ultrasound image and one of the clinical parameter for characterizing cardiac function.
  19. A kind of heart volume discriminance analysis system, wherein include:
    Ultrasound image acquisition module, for obtaining the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time；
    Section type identification module, for identification the section type of the multiple frames of ultrasonic image data cardiac；
    Cardiac cycle identification module, cardiac cycle for identification；
    Profile obtains module, for the location and shape based on the endocardium of left ventricle in every frame ultrasound image in one cardiac cycle of section type identification；
    Ventricular volume curve generation module calculates the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image for the location and shape according to the endocardium of left ventricle, obtains ventricular volume curve；
    Clinical parameter output module for exporting ventricular volume curve, and/or calculates according to the ventricular volume curve and exports the clinical parameter of characterization cardiac function.
  20. A kind of heart volume discriminance analysis system as claimed in claim 19, wherein the system also includes:
    It shows mark module, for showing moment corresponding ultrasound image locating for the maximum value and minimum value on the ventricular volume curve, and marks the location and shape for showing the endocardium of left ventricle；And/or
    Section type display module, for showing the section type identified.
  21. A kind of heart volume discriminance analysis system, wherein include:
    Probe；
    Transmit circuit, for emitting ultrasonic beam to the heart area of target object；
    Circuit and Beam synthesis module are received, for obtaining ultrasound echo signal；
    Image processing module, for obtaining the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time according to ultrasound echo signal, identify the section type of the ultrasound image cardiac, identify cardiac cycle, and identify the location and shape of endocardium of left ventricle in every frame ultrasound image in a cardiac cycle, and the ventricular volume quantitative parameter at moment corresponding to each frame ultrasound image is calculated, obtain ventricular volume curve；And
    Display, for showing the location and shape of the ultrasound image, the label endocardium of left ventricle, and the display section type.
  22. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein whether display prompts user the identified cardiac cycle is wrong, the system also includes:
    For receiving the operation control module of user's input control order；
    When the cardiac cycle described in the display prompts user is wrong, image processing module switches to manual input mode, the cardiac cycle that user can be manually entered by operating control module.
  23. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein the display shows that the recognition result of above-mentioned section type confirms for user, the system also includes:
    For receiving the operation control module of user's input control order；
    When the control command of user's input is recognition result confirmation mistake, image processing module switches manual input mode, modifies current section type through the control command of operation control module input according to user and shows.
  24. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein the display shows that the location and shape of the endocardium of left ventricle of every frame ultrasound image confirm for user, the system also includes:
    For receiving the operation control module of user's input control order；
    When the control command of user's input is recognition result confirmation mistake, image processing module switches manual input mode, location and shape and display according to user through endocardium of left ventricle on the control command modification ultrasound image of operation control module input.
  25. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein the display shows that current display and output result confirm and/or correct for user, the system also includes:
    For receiving the operation control module of user's input control order；
    Image processing module based on user through operation control module input confirmation and/or amendment order, formed output report, amendment content include at least: at the time of diastasis corresponds to, at the time of end-systole corresponds to, The location and shape of the endocardium of left ventricle marked on ultrasound image and one of the clinical parameter for characterizing cardiac function.
  26. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein the display further displays the ventricular volume curve being calculated, and/or the clinical parameter of characterization cardiac function.
  27. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein described image processing module identifies cardiac cycle in the following manner:
    According to the multiple frames of ultrasonic image of the continuous acquisition on predetermined amount of time, the characteristic value of every frame image, Characteristi c curve of formation are extracted；
    Periodicity analysis is carried out to the indicatrix, identifies the cardiac cycle of object.
  28. A kind of heart volume discriminance analysis system as claimed in claim 27, wherein the characteristic value of every frame image includes anatomical structure measured value, and the indicatrix is the curve that anatomical structure measured value changes over time；Or
    The characteristic value of every frame image is likeness coefficient, and the indicatrix is likeness coefficient curve.
  29. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein described image processing module identifies cardiac cycle in the following manner:
    A frame ultrasound image is chosen as standard frame；
    Local extremum will be searched in regional area near the moment where standard frame, and determine ultrasound image corresponding to local extremum；As the beginning and end of time at the time of respectively corresponding with standard frame, a cardiac cycle is obtained.
  30. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein described image processing module identifies the location and shape of the endocardium of left ventricle in a cardiac cycle in every frame ultrasound image in the following manner:
    A frame image is chosen as key frame, image segmentation is carried out based on endocardium of left ventricle of the section type to the key frame, obtains standard endocardial parted pattern；
    According to the standard endocardial parted pattern, the location and shape of the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle are identified.
  31. A kind of heart volume discriminance analysis system as claimed in claim 30, wherein, described image processing module carrys out the location and shape that the endocardium of left ventricle on each frame ultrasound image in a cardiac cycle is identified according to the standard endocardial parted pattern in the following manner:
    According to the standard endocardial parted pattern, to other ultrasounds in cardiac cycle in addition to key frame Image is divided frame by frame, obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image；Or
    Based on the endocardial location and shape of the key frame heart divided, the tracking for carrying out endocardial motion to other ultrasound images in cardiac cycle in addition to key frame is handled, and obtains the location and shape of the endocardium of left ventricle on each frame ultrasound image.
  32. A kind of heart volume discriminance analysis system as claimed in claim 30, wherein, described image processing module chooses a frame image as key frame automatically in the following manner, image segmentation is carried out based on endocardium of left ventricle of the section type to the key frame, obtains standard endocardial parted pattern:
    Edge extracting is carried out based on characteristic point, generates the endocardial location and shape of key frame；And/or
    Using the method for machine learning, the endocardial location and shape of key frame are identified according to overall model.
  33. A kind of heart volume discriminance analysis system as claimed in claim 21, wherein described image processing module identifies the section type of the multiple frames of ultrasonic image data cardiac in the following manner:
    It by the feature of a frame or multiple frames of ultrasonic image in the multiple frames of ultrasonic image, is compared with the feature of the training image of known section type, to obtain the section type of the multiple frames of ultrasonic image.
  34. A kind of heart volume discriminance analysis system as claimed in claim 33, wherein, described image processing module is also handled as follows before being compared with the feature of the training image of known section type by the feature of a frame or multiple frames of ultrasonic image in the multiple frames of ultrasonic image:
    Identify the position of interventricular septum in ultrasound image；
    Ultrasound image is rotated according to the position of the interventricular septum, keeps the long axis of left ventricle direction in ultrasound image vertical；
    The ultrasound image is translated, the left ventricular location in the ultrasound image is adjusted to the center of image.
CN201680082174.4A 2016-05-30 2016-05-30 A kind of heart volume discriminance analysis system and method Pending CN108882917A (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2016/083923 WO2017206023A1 (en) 2016-05-30 2016-05-30 Cardiac volume identification analysis system and method

Publications (1)

Publication Number Publication Date
CN108882917A true CN108882917A (en) 2018-11-23

Family

ID=60479479

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201680082174.4A Pending CN108882917A (en) 2016-05-30 2016-05-30 A kind of heart volume discriminance analysis system and method

Country Status (2)

Country Link
CN (1) CN108882917A (en)
WO (1) WO2017206023A1 (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110349149A (en) * 2019-07-12 2019-10-18 广东省人民医院(广东省医学科学院) Congenital heart disease categorizing system and method based on deep neural network and form similarity
CN110543912A (en) * 2019-09-02 2019-12-06 李肯立 Method for automatically acquiring cardiac cycle video in fetal key section ultrasonic video
CN110664435A (en) * 2019-09-23 2020-01-10 东软医疗系统股份有限公司 Method and device for acquiring cardiac data and ultrasonic imaging equipment
CN111652954A (en) * 2020-07-01 2020-09-11 杭州脉流科技有限公司 Left ventricle volume automatic calculation method and device based on left ventricle segmentation picture, computer equipment and storage medium
CN112656445A (en) * 2020-12-18 2021-04-16 青岛海信医疗设备股份有限公司 Ultrasonic device, ultrasonic image processing method and storage medium
CN112842384A (en) * 2020-12-30 2021-05-28 无锡触典科技有限公司 Echocardiography myocardial envelope measuring method, device and storage medium
CN113243935A (en) * 2021-04-02 2021-08-13 中国人民解放军63919部队 Echocardiogram acquisition method and device, electronic equipment and storage medium
CN114419499A (en) * 2022-01-07 2022-04-29 乐普（北京）医疗器械股份有限公司 Cardiac ultrasound video-based ejection fraction data processing method and device
CN115587971A (en) * 2022-09-21 2023-01-10 四川大学华西医院 Method and system for monitoring body reaction and hemodynamics based on heart ultrasonic segmental motion
CN116705307A (en) * 2023-08-07 2023-09-05 天津云检医学检验所有限公司 AI model-based heart function assessment method, system and storage medium for children

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU2019268319A1 (en) * 2018-05-15 2020-12-24 New York University System and method for orientating capture of ultrasound images
CN111214254B (en) * 2018-11-26 2023-06-20 深圳迈瑞生物医疗电子股份有限公司 Ultrasonic diagnostic apparatus, method of acquiring ultrasound image of section of ultrasonic diagnostic apparatus, and readable storage medium
TW202022713A (en) * 2018-12-05 2020-06-16 宏碁股份有限公司 Method and system for evaluating cardiac status, electronic device and ultrasonic scanning device
CN109925002A (en) * 2019-01-15 2019-06-25 胡秋明 Artificial intelligence echocardiogram data collection system and its collecting method
CN110742653B (en) * 2019-10-31 2020-10-30 无锡祥生医疗科技股份有限公司 Cardiac cycle determination method and ultrasonic equipment
EP4106633B1 (en) * 2020-02-20 2023-08-23 Koninklijke Philips N.V. Methods and systems for fetal heart assessment
CN111466894B (en) * 2020-04-07 2023-03-31 上海深至信息科技有限公司 Ejection fraction calculation method and system based on deep learning
US20210330285A1 (en) * 2020-04-28 2021-10-28 EchoNous, Inc. Systems and methods for automated physiological parameter estimation from ultrasound image sequences
CN111508077B (en) * 2020-04-29 2021-01-15 中国人民解放军总医院 Method, device, equipment and medium for intelligently evaluating heart function by using double-triangle model
CN111768391B (en) * 2020-07-01 2024-04-05 杭州脉流科技有限公司 Full-automatic heart function analysis method, device, computer equipment and storage medium based on CT image
CN112336381B (en) * 2020-11-07 2022-04-22 吉林大学 Echocardiogram end systole/diastole frame automatic identification method based on deep learning
CN112489043A (en) * 2020-12-21 2021-03-12 无锡祥生医疗科技股份有限公司 Heart disease detection device, model training method, and storage medium
CN114504343B (en) * 2020-12-31 2023-08-22 深圳北芯生命科技股份有限公司 Method for calculating cardiac cycle based on intravascular ultrasound image
CN113592802B (en) * 2021-07-26 2024-02-02 东北大学 Mitral valve annular displacement automatic detection system based on ultrasonic image
CN113689441A (en) * 2021-08-30 2021-11-23 华东师范大学 DeepLabV3 network-based left ventricle ultrasonic dynamic segmentation method
CN116109626A (en) * 2023-04-04 2023-05-12 深圳鲲为科技有限公司 Processing method and device of heart ultrasonic data, storage medium and electronic equipment

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1777898A (en) * 2003-04-24 2006-05-24 皇家飞利浦电子股份有限公司 Non-invasive left ventricular volume determination
CN1934589A (en) * 2004-03-23 2007-03-21 美国西门子医疗解决公司 Systems and methods providing automated decision support for medical imaging
CN102421373A (en) * 2009-05-20 2012-04-18 株式会社日立医疗器械 Medical image diagnosis device and region-of-interest setting method therefor
CN103732134A (en) * 2010-12-29 2014-04-16 迪亚卡帝奥有限公司 Automatic left ventricular function evaluation
US20150065879A1 (en) * 2004-04-02 2015-03-05 Teratech Corporation Wall motion analyzer
CN105555198A (en) * 2014-03-20 2016-05-04 深圳迈瑞生物医疗电子股份有限公司 Method and device for automatic identification of measurement item, and ultrasound imaging apparatus

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120283567A1 (en) * 2010-01-29 2012-11-08 Hitachi Medical Corporation Ultrasonic diagnostic apparatus and measurement-point tracking method
JP5347003B2 (en) * 2011-09-30 2013-11-20 富士フイルム株式会社 Medical image processing apparatus and method, and program
JP5958931B2 (en) * 2012-04-27 2016-08-02 国立大学法人東京農工大学 Left ventricular wall information extraction device, method, and program

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1777898A (en) * 2003-04-24 2006-05-24 皇家飞利浦电子股份有限公司 Non-invasive left ventricular volume determination
CN1934589A (en) * 2004-03-23 2007-03-21 美国西门子医疗解决公司 Systems and methods providing automated decision support for medical imaging
US20150065879A1 (en) * 2004-04-02 2015-03-05 Teratech Corporation Wall motion analyzer
CN102421373A (en) * 2009-05-20 2012-04-18 株式会社日立医疗器械 Medical image diagnosis device and region-of-interest setting method therefor
CN103732134A (en) * 2010-12-29 2014-04-16 迪亚卡帝奥有限公司 Automatic left ventricular function evaluation
CN105555198A (en) * 2014-03-20 2016-05-04 深圳迈瑞生物医疗电子股份有限公司 Method and device for automatic identification of measurement item, and ultrasound imaging apparatus

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110349149A (en) * 2019-07-12 2019-10-18 广东省人民医院(广东省医学科学院) Congenital heart disease categorizing system and method based on deep neural network and form similarity
CN110543912A (en) * 2019-09-02 2019-12-06 李肯立 Method for automatically acquiring cardiac cycle video in fetal key section ultrasonic video
CN110543912B (en) * 2019-09-02 2021-10-01 李肯立 Method for automatically acquiring cardiac cycle video in fetal key section ultrasonic video
CN110664435A (en) * 2019-09-23 2020-01-10 东软医疗系统股份有限公司 Method and device for acquiring cardiac data and ultrasonic imaging equipment
CN111652954B (en) * 2020-07-01 2023-09-05 杭州脉流科技有限公司 Left ventricle volume automatic calculation method, device, computer equipment and storage medium based on left ventricle segmentation picture
CN111652954A (en) * 2020-07-01 2020-09-11 杭州脉流科技有限公司 Left ventricle volume automatic calculation method and device based on left ventricle segmentation picture, computer equipment and storage medium
CN112656445A (en) * 2020-12-18 2021-04-16 青岛海信医疗设备股份有限公司 Ultrasonic device, ultrasonic image processing method and storage medium
CN112842384A (en) * 2020-12-30 2021-05-28 无锡触典科技有限公司 Echocardiography myocardial envelope measuring method, device and storage medium
CN113243935A (en) * 2021-04-02 2021-08-13 中国人民解放军63919部队 Echocardiogram acquisition method and device, electronic equipment and storage medium
CN114419499A (en) * 2022-01-07 2022-04-29 乐普（北京）医疗器械股份有限公司 Cardiac ultrasound video-based ejection fraction data processing method and device
CN114419499B (en) * 2022-01-07 2024-03-29 乐普（北京）医疗器械股份有限公司 Method and device for processing ejection fraction data based on heart ultrasonic video
CN115587971A (en) * 2022-09-21 2023-01-10 四川大学华西医院 Method and system for monitoring body reaction and hemodynamics based on heart ultrasonic segmental motion
CN115587971B (en) * 2022-09-21 2023-10-24 四川大学华西医院 Organism reaction and hemodynamic monitoring method and system based on heart ultrasonic segment activity
CN116705307A (en) * 2023-08-07 2023-09-05 天津云检医学检验所有限公司 AI model-based heart function assessment method, system and storage medium for children

Also Published As

Publication number Publication date
WO2017206023A1 (en) 2017-12-07

Similar Documents

Publication Publication Date Title
CN108882917A (en) A kind of heart volume discriminance analysis system and method
CN108701354B (en) Method and system for identifying contour of interest region in ultrasonic image
JP5108905B2 (en) Method and apparatus for automatically identifying image views in a 3D dataset
RU2657855C2 (en) Three-dimensional ultrasound imaging system
CN102763135B (en) For the method for auto Segmentation and time tracking
US20030038802A1 (en) Automatic delineation of heart borders and surfaces from images
EP1690230B1 (en) Automatic multi-dimensional intravascular ultrasound image segmentation method
US6106466A (en) Automated delineation of heart contours from images using reconstruction-based modeling
EP2030042B1 (en) Quantification and display of cardiac chamber wall thickening
US8343053B2 (en) Detection of structure in ultrasound M-mode imaging
EP2392942B1 (en) Cardiac flow quantification with volumetric imaging data
JP6173886B2 (en) Ultrasonic diagnostic apparatus, image processing apparatus, and image processing method
US20030160786A1 (en) Automatic determination of borders of body structures
KR102251830B1 (en) Systems and methods for registration of ultrasound and ct images
US6824517B2 (en) Ultrasound quantification in real-time using acoustic data in more than two dimensions
EP2434454A2 (en) Computerized characterization of cardiac motion in medical diagnostic ultrasound
EP3363365A1 (en) Automatic imaging plane selection for echocardiography
JP2005296658A (en) Method and apparatus for detecting living body tissue structure
JP2019024925A (en) Medical imaging apparatus and image processing method
US10016180B2 (en) Ultrasonic image processing device
US11534133B2 (en) Ultrasonic detection method and ultrasonic imaging system for fetal heart
JP6498316B2 (en) Ultrasonic imaging apparatus and image processing method
Chacko et al. Three Dimensional Echocardiography: Recent Trends in Segmen tation Methods
Orderud Real-time segmentation of 3D echocardiograms using a state estimation approach with deformable models
EP4312792A1 (en) Method and apparatus with user guidance and automated image setting selection for mitral regurgitation evaluation

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20181123