CN107679574A - Ultrasonoscopy processing method and system - Google Patents
Ultrasonoscopy processing method and system Download PDFInfo
- Publication number
- CN107679574A CN107679574A CN201710917423.3A CN201710917423A CN107679574A CN 107679574 A CN107679574 A CN 107679574A CN 201710917423 A CN201710917423 A CN 201710917423A CN 107679574 A CN107679574 A CN 107679574A
- Authority
- CN
- China
- Prior art keywords
- ultrasonoscopy
- image
- abnormal
- feature
- image feature
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of ultrasonoscopy processing method and system.The present invention obtains characteristics of image by carrying out feature extraction to ultrasonoscopy, template database Auto-matching is realized using characteristics of image, abnormal image information corresponding to the maximum pre-set image feature of output matching degree, aid in user quickly and accurately to carry out Treatment Analysis to ultrasonoscopy, improve efficiency and the degree of accuracy of ultrasonoscopy Treatment Analysis.
Description
Technical field
The present invention relates to technical field of medical instruments, more particularly to a kind of ultrasonoscopy processing method and system.
Background technology
At present, for advantage medical resource high concentration in big city and Hospital Affiliated with Medical College, medical model belongs to height
Centralized system.Existing ultrasonoscopy treatment technology needs sonographer to have an abundant clinical experience, and technical threshold is higher；Basic unit cures
Ultrasonoscopy in institute even without specialty handles doctor, and there is that sonographer gimmick is lack of standardization, clinical experience is few etc. is many
Problem.For rare difficult type, the typically more difficult case drawn a conclusion or had a question to conclusion, by the way of the consultation of doctors,
So that whole process cycle extends, operating efficiency is influenceed.Either because sonographer is lacking in experience error in judgement, also it is due to
The consultation of doctors causes the extension of processing time.
The content of the invention
In order to solve the above-mentioned technical problem, it is an object of the invention to provide a kind of ultrasonoscopy processing method and system, its
Efficiency and the degree of accuracy of ultrasonoscopy Treatment Analysis can be improved.
The technical solution adopted in the present invention is：A kind of ultrasonoscopy processing method, comprises the following steps：
Obtain the ultrasonoscopy of destination object；
Feature extraction is carried out to the ultrasonoscopy to obtain characteristics of image；
Described image feature is matched with the pre-set image feature in template database, obtains matching degree；The mould
Plate database purchase has a variety of abnormal image information and corresponding pre-set image feature；
When the matching degree is more than predetermined threshold value, pre-set image feature corresponding to maximum matching degree is obtained；
Export abnormal image information corresponding with the pre-set image feature that the matching degree is maximum.
Further, the template number is established by machine learning method using the ultrasound image data of multiple data centers
According to storehouse.
Further, it is described described image feature is subjected to matching with the pre-set image feature in template database to include：
Calculate the distance of described image feature and the pre-set image feature in template database；
By the distance compared with preset value, the matching degree is obtained.
Further, the feature extraction includes color characteristic and/or textural characteristics and/or shape facility and/or space
Relationship characteristic extracts；
The abnormal image information claims including exception name and/or the edge wheel of the form of abnormal image and/or abnormal image
Wide and/or Anomalous typing and/or abnormal classification.
Further, the ultrasonoscopy processing method also includes step：
When the matching degree is more than predetermined threshold value, the ultrasonoscopy is measured to obtain abnormal image feature, wherein, institute
Stating abnormal image feature includes the number and/or the size of abnormal image and/or the actual edge wheel of abnormal image of abnormal image
It is wide；
Export the abnormal image feature.
Further, contrast shows the abnormal image feature and the abnormal image information.
Another technical scheme of the present invention is：A kind of ultrasonoscopy processing system, including
Image acquisition unit, for obtaining the ultrasonoscopy of destination object；
Feature extraction unit, for carrying out feature extraction to the ultrasonoscopy to obtain characteristics of image；
Matching unit, for described image feature to be matched with the pre-set image feature in template database, obtain
Matching degree, when the matching degree is more than predetermined threshold value, pre-set image feature corresponding to maximum matching degree is obtained, wherein, it is described
Template database is stored with a variety of abnormal image information and corresponding pre-set image feature；
Output unit, for exporting abnormal image information corresponding with the pre-set image feature of matching degree maximum.
Further, the ultrasonoscopy processing system also includes：
Template database establishes unit, for passing through machine learning method using the ultrasound image data of multiple data centers
Establish the template database.
Further, the ultrasonoscopy processing system also includes：
Image measurement unit, for when the matching degree is more than predetermined threshold value, it is different to obtain to measure the ultrasonoscopy
Normal characteristics of image.
Further, the ultrasonoscopy processing system also includes：
Display unit, the abnormal image feature and the abnormal image information are shown for contrasting.
The beneficial effects of the invention are as follows：
The present invention obtains characteristics of image by carrying out feature extraction to ultrasonoscopy, and template data is realized using characteristics of image
Storehouse Auto-matching, abnormal image information corresponding to the maximum pre-set image feature of output matching degree, auxiliary user is quickly and accurately
Treatment Analysis is carried out to ultrasonoscopy, improves efficiency and the degree of accuracy of ultrasonoscopy Treatment Analysis.
Brief description of the drawings
The embodiment of the present invention is described further below in conjunction with the accompanying drawings：
Fig. 1 is ultrasonoscopy process flow figure described in first embodiment of the invention；
Fig. 2 is by the pre-set image feature progress in characteristics of image and template database described in first embodiment of the invention
The method flow diagram matched somebody with somebody；
Fig. 3 is ultrasonoscopy process flow figure described in second embodiment of the invention；
Fig. 4 is the structural representation of ultrasonoscopy processing system described in third embodiment of the invention；
Fig. 5 is the structural representation of ultrasonoscopy processing system described in fourth embodiment of the invention.
Embodiment
It should be noted that in the case where not conflicting, the feature in embodiment and embodiment in the application can phase
Mutually combination.
Embodiment one
A kind of ultrasonoscopy processing method will be described in detail for first embodiment of the invention, the present embodiment methods described
Comprise the following steps, as shown in figure 1, Fig. 1 is ultrasonoscopy process flow figure described in first embodiment of the invention：
101st, the ultrasonoscopy for including destination object is obtained；
In the present embodiment method, under real-time ultrasonoscopy or frozen state, obtain and include the super of destination object
Acoustic image, and the ultrasonoscopy of acquisition is handled.
102nd, feature extraction is carried out to ultrasonoscopy to obtain characteristics of image.
One kind that characteristics of image includes but is not limited in color characteristic, textural characteristics, shape facility, spatial relation characteristics or
It is a variety of or by the way of machine learning.Color characteristic includes statistic histogram, accumulation histogram, color moment etc., textural characteristics
Including gray scale symbiosis square, Tamura etc., shape facility includes edge orientation histogram, eccentricity etc., and spatial relation characteristics include
SIFT (Scale-invariant feature transform, Scale invariant features transform) etc..In the present embodiment, to super
Acoustic image carries out Tamura texture feature extractions.
In the present embodiment, from contrast, roughening and directive property as Tamura textural characteristics, accessed is surpassed
Acoustic image is handled to obtain corresponding contrast, roughening and directive property, and characteristics of image includes contrast, roughening and sensing
Property, they form a three dimensions point.
Specifically, contrast can be obtained by calculating the intensity profile situation of the pixel in ultrasonoscopy, contrast by
The influence of sharp edges in the grey scale change scope and image of image pixel.The calculation formula of contrast is：
Wherein, contrast represents contrast；What σ was represented is to carry out standard variance calculating to the gray scale of ultrasonoscopy and obtain
The value arrived；μ4What is represented is value obtained from carrying out fourth central square calculating to the gray scale of ultrasonoscopy；Value desirable n is 2,4,
8th, 1/2,1/4,1/8 and 1.
Roughening is for measuring the distance between texture, if the resolution ratio of image is small, texture is comparatively fine.It is coarse
The calculation formula of property is：
In above formula, coarseness is roughening；Mn is the size of pixel point detection window, takes 2k*2k；K value is
[0,5]；(i, j) is the coordinate in pixel point detection window.
Directive property is a global characteristics in the picture, and it describes direction during grain distribution in the picture.Calculating
During the directive property of image, the gradient vector △ G in image are first calculated, wherein gradient vector △ G size is | Δ G |, direction
For θ, π value is 3.1415926, and specific formula for calculation is as follows：
| Δ G |=(| Δ H |+| Δ V |)/2
θ=arctan (Δ V/ Δ H)+pi/2
Δ H, Δ V are respectively the matrix that image obtains with following formula convolution；
After calculating gradient vector △ G, gradient vector distribution histogram is obtained, the sensing of image can be obtained by following formula
Property：
Wherein, FdirRefer to tropism；P is the crest numerical value of gradient vector distribution histogram；npIt is gradient vector distribution Nogata
Crest number in figure；wpIt is the distance between wave base of crest both sides of gradient vector distribution histogram；φpIt is gradient vector
The location of crest of distribution histogram；R is normalization factor.
So far, the characteristics of image according to corresponding to above-mentioned calculation formula can obtain ultrasonoscopy, i.e. contrast, roughening and
Directional characteristics.
103rd, characteristics of image is matched with the pre-set image feature in template database, obtains matching degree, wherein, mould
Plate database purchase has a variety of abnormal image information and corresponding pre-set image feature.
In the present embodiment, the template is established by machine learning method using the ultrasound image data of multiple data centers
Database, the ultrasound image data can be distributed across the collection of the data centers such as different zones, Different hospital or clinic
Ultrasound image data.A variety of different abnormal image information and its corresponding pre-set image feature are stored with template database,
Abnormal image information corresponds with pre-set image feature.The abnormal image information claims including exception name, the shape of abnormal image
State, the edge contour of abnormal image, Anomalous typing, abnormal classification.Abnormal image refers to the normal ultrasonoscopy with destination object
Compare, the image of anomalous variation be present, such as acquired ultrasonoscopy, compared with normal ultrasonoscopy, destination object exists big
Small exception, for example, less than normal or bigger than normal etc..Anomalous typing refers to the type of anomalous variation, for example, size be anomaly divided into it is less than normal and
Two kinds of abnormal conditions bigger than normal；Abnormal classification refers to the rank of anomalous variation, i.e. abnormal degree, such as one-level is abnormal, two level is different
Often or three-level is abnormal etc..
In fact, the normal ultrasound image data of destination object is also stored with template database, including a variety of normograms
As information and its corresponding pre-set image feature.
Further, with reference to figure 2, Fig. 2 is by characteristics of image and template database described in first embodiment of the invention
The method flow diagram that pre-set image feature is matched, it is described to enter characteristics of image and the pre-set image feature in template database
Row matching includes：
201st, characteristics of image and the distance of the pre-set image feature in template database are calculated；
The method for calculating characteristics of image and the distance of the pre-set image feature in template database includes but is not limited to calculate
Euclidean distance, histogram intersection distance or included angle cosine distance between the two.
In the present embodiment, from the Euclidean distance between the pre-set image feature calculated in characteristics of image and template database
It is as follows for point set x and Y, their Euclidean distance calculation formula in n-dimensional space to carry out data processing：
Wherein, Euclidean distances (A point and B point be respectively in point set X and Y a bit) of the d (A, B) between A points, B points,
Xi、YiRespectively A points, B points i opening positions value, N be point set X and Y in element number.
In the present embodiment, the characteristics of image of ultrasonoscopy includes contrast, roughening and directive property, then correspondingly, template
Pre-set image feature in database includes default contrast, default roughening and default directive property.
The distance for calculating characteristics of image and the pre-set image feature in template database is to calculate Euclidean therebetween
Distance.
202nd, according to distance compared with preset value, the matching degree is obtained.
, can be using the ratio of the distance being calculated and preset value as matching degree, the number of matching degree in the present embodiment
For one or more.Certainly other also can be selected apart from calculation, such as histogram intersection distance or included angle cosine distance,
Here repeat no more.
104th, when matching degree is more than predetermined threshold value, pre-set image feature corresponding to maximum matching degree is obtained.
, will be big when one or more matching degree that above-mentioned steps 103 obtain has the matching degree more than predetermined threshold value
Sorted in the matching degree of predetermined threshold value, and obtain pre-set image feature corresponding to maximum matching degree.
105th, output abnormal image information corresponding with the pre-set image feature that matching degree is maximum.
Abnormal image information corresponding to the pre-set image feature that matching degree is maximum in template database is searched, and is exported aobvious
Show, to help user to analyze and process ultrasonoscopy, so as to improve the speed of analyzing and processing and the degree of accuracy.In addition, also may be used
Abnormal image information is exported in a manner of by voice broadcast.
Embodiment two
Second embodiment of the invention elaborates a kind of ultrasonoscopy processing method.Ultrasonoscopy processing side described in the present embodiment
Method comprises the following steps, as shown in figure 3, Fig. 3 is ultrasonoscopy process flow figure described in second embodiment of the invention：
301st, the ultrasonoscopy for including destination object is obtained；
302nd, feature extraction is carried out to ultrasonoscopy to obtain characteristics of image；
303rd, characteristics of image is matched with the pre-set image feature in template database, obtains matching degree, wherein, mould
Plate database purchase has a variety of abnormal image information and corresponding pre-set image feature；
304th, when matching degree is more than predetermined threshold value, pre-set image feature corresponding to maximum matching degree is obtained；
305th, output abnormal image information corresponding with the pre-set image feature that matching degree is maximum；
306th, when matching degree is more than predetermined threshold value, ultrasonoscopy is measured to obtain abnormal image feature, wherein, Abnormal Map
As feature includes the number and/or the size of abnormal image and/or the actual edge profile of abnormal image of abnormal image；
In the present embodiment, according to a first embodiment of the present invention described ultrasonoscopy processing method judge ultrasonoscopy with
After the matching degree of pre-set image feature in template database, when matching degree is more than predetermined threshold value, measurement ultrasonoscopy with
Abnormal image feature is obtained, abnormal image feature includes number, the size of abnormal image, the reality of abnormal image of abnormal image
Edge contour.The measuring method of ultrasonoscopy is the common method using ultrasonoscopy process field, is repeated no more.
307th, output abnormality characteristics of image.
In the present embodiment, after getting abnormal image feature, shown, to aid in user to carry out ultrasonoscopy
Analysis, improve the degree of accuracy of ultrasonoscopy processing.Equally, abnormal image feature can also be exported by way of voice broadcast
Learnt to user.
Further, in one embodiment, it is special also to include the contrast display abnormal image for ultrasonoscopy processing method
Seek peace the abnormal image information.
The information that the actual measured value (i.e. abnormal image feature) of ultrasonoscopy and matching are obtained (i.e. believe by abnormal image
Breath) contrast, improves the degree of accuracy that ultrasonoscopy is handled.
Embodiment three
A kind of ultrasonoscopy processing system will be described in detail for third embodiment of the invention, the ultrasonoscopy processing
The concrete structure of system refers to Fig. 4, and Fig. 4 is the structural representation of ultrasonoscopy processing system described in third embodiment of the invention
Figure, including image acquisition unit 401, feature extraction unit 402, matching unit 403 and output unit 404.
Image acquisition unit 401, for obtaining the ultrasonoscopy of destination object.
In the present embodiment, the collection of ultrasonoscopy is carried out using ultrasonic image-forming system.The mode of collection can use common
Probe, such as common convex array probe and linear array probe, carry out manual scanning by user, obtain and a series of includes destination object
Ultrasonoscopy.Volume probe or probe matrix can also be used, gathers a three-dimensional or the view data of the four-dimension.
Feature extraction unit 402, for carrying out feature extraction to ultrasonoscopy to obtain characteristics of image.
One kind that characteristics of image includes but is not limited in color characteristic, textural characteristics, shape facility, spatial relation characteristics or
It is a variety of or by the way of machine learning.In the present embodiment, Tamura texture feature extractions are carried out to ultrasonoscopy.
In the present embodiment, from contrast, roughening and directive property as Tamura textural characteristics, accessed is surpassed
Acoustic image is handled to obtain corresponding contrast, roughening and directive property, and characteristics of image includes contrast, roughening and sensing
Property, they form a three dimensions point.The computational methods of contrast, roughening and directive property refer to first embodiment of the invention
Description.
Matching unit 403, for characteristics of image to be matched with the pre-set image feature in template database, acquisition
With degree, when matching degree is more than predetermined threshold value, pre-set image feature corresponding to maximum matching degree is obtained, wherein, template database
It is stored with a variety of abnormal image information and corresponding pre-set image feature.
In the present embodiment, ultrasonoscopy processing system also establishes unit including template database, for utilizing multiple data
The ultrasound image data at center establishes the template database by machine learning method.The ultrasound image data can be point
The ultrasound image data that cloth gathers in data centers such as different zones, Different hospital or clinics.It is pre-stored in template database
There are a variety of different abnormal image information and its corresponding pre-set image feature, abnormal image information and pre-set image feature are one by one
It is corresponding；Wherein, abnormal image information claims including exception name, the edge contour of the form of abnormal image, abnormal image, exception are divided
Type, abnormal classification.
Further, matching bag is carried out with reference to figure 2, the pre-set image feature by characteristics of image and template database
Include：
201st, characteristics of image and the distance of the pre-set image feature in template database are calculated；
The method for calculating characteristics of image and the distance of the pre-set image feature in template database includes but is not limited to calculate
Euclidean distance, histogram intersection distance or included angle cosine distance between the two；In the present embodiment, from calculate characteristics of image with
The Euclidean distance of pre-set image feature in template database come be made whether comprising abnormal image judge, the image of ultrasonoscopy
Feature includes contrast, roughening and directive property；And pre-set image feature also includes contrast, roughening and directive property.
The distance for calculating characteristics of image and the pre-set image feature in template database is to calculate Euclidean therebetween
Distance.
202nd, according to distance compared with preset value, the matching degree is obtained.
, can be using the ratio of the distance being calculated and preset value as matching degree, the number of matching degree in the present embodiment
For one or more.Certainly other also can be selected apart from calculation, such as histogram intersection distance or included angle cosine distance,
Here repeat no more.
When one or more matching degree of above-mentioned acquisition has the matching degree more than predetermined threshold value, default threshold will be greater than
The matching degree sequence of value, and obtain pre-set image feature corresponding to maximum matching degree.
In actual use, feature extraction unit 402 and matching unit 403 can be realized using computer, pop one's head in and count
Calculation machine is connected and the ultrasonoscopy collected is sent into computer, and ultrasonoscopy is handled using computer, is judged super
Whether acoustic image includes abnormal image.
Output unit 404, for exporting abnormal image information corresponding with the pre-set image feature of matching degree maximum.
In the present embodiment, output unit 404 is searched corresponding to the pre-set image feature that matching degree is maximum in template database
Abnormal image information, and output display, to help user to analyze and process ultrasonoscopy, so as to improve the speed of analyzing and processing
Degree and the degree of accuracy.In addition, output unit 404 can also be exported abnormal image information by way of voice broadcast.
Example IV
A kind of ultrasonoscopy processing system will be described in detail for fourth embodiment of the invention, the ultrasonoscopy processing
The concrete structure of system refers to Fig. 5, and Fig. 5 is the structural representation of ultrasonoscopy processing system described in fourth embodiment of the invention
Figure, including image acquisition unit 501, feature extraction unit 502, matching unit 503, output unit 504, image measurement unit
505 and display unit 506.
Image acquisition unit 501, for obtaining the ultrasonoscopy of destination object；
Feature extraction unit 502, for carrying out feature extraction to ultrasonoscopy to obtain characteristics of image；
Matching unit 503, for characteristics of image to be matched with the pre-set image feature in template database, acquisition
With degree, when matching degree is more than predetermined threshold value, pre-set image feature corresponding to maximum matching degree is obtained, wherein, template database
It is stored with a variety of abnormal image information and corresponding pre-set image feature；
Output unit 504, for exporting abnormal image information corresponding with the pre-set image feature of matching degree maximum；
Image measurement unit 505, for when matching degree is more than predetermined threshold value, measuring ultrasonoscopy to obtain abnormal image
Feature；
The present embodiment is upgraded to obtain on the basis of third embodiment of the invention, the ultrasound figure in the present embodiment
Picture processing system also includes image measurement unit 505, for when matching degree is more than predetermined threshold value, measuring ultrasonoscopy to obtain
Abnormal image feature, abnormal image feature include number, the size of abnormal image, the actual edge of abnormal image of abnormal image
Profile.
Display unit 506, the abnormal image feature and the abnormal image information are shown for contrasting.
In the present embodiment, display unit 506, the display content of display unit includes the abnormal image information of template database
Abnormal image feature is obtained with measurement, abnormal image information claims including exception name, the edge of the form of abnormal image, abnormal image
Profile, Anomalous typing, abnormal classification, abnormal image feature include the number, the size of abnormal image, abnormal image of abnormal image
Actual edge profile.The information that the actual measured value (i.e. abnormal image feature) of ultrasonoscopy and matching are obtained is (i.e. abnormal
Image information) contrast, improves the degree of accuracy that ultrasonoscopy is handled.
In actual use, feature extraction unit 502, matching unit 503, image measurement unit 504 can use computer
To realize, probe is connected with computer and the ultrasonoscopy collected is sent into computer, using computer to ultrasonoscopy
Handled, judge whether ultrasonoscopy measures acquisition abnormal image feature comprising abnormal image and to ultrasonoscopy.
A kind of ultrasonoscopy processing system in the present invention, characteristics of image is obtained by carrying out feature extraction to ultrasonoscopy,
Template database Auto-matching, abnormal image corresponding to the maximum pre-set image feature of output matching degree are realized using characteristics of image
Information, auxiliary user quickly and accurately carry out Treatment Analysis to ultrasonoscopy, improve the efficiency and standard of ultrasonoscopy Treatment Analysis
Exactness.
Above is the preferable implementation to the present invention is illustrated, but the invention is not limited to the implementation
Example, those skilled in the art can also make a variety of equivalent variations on the premise of without prejudice to spirit of the invention or replace
Change, these equivalent deformations or replacement are all contained in the application claim limited range.
Claims (10)
1. a kind of ultrasonoscopy processing method, it is characterised in that comprise the following steps：
Obtain the ultrasonoscopy of destination object；
Feature extraction is carried out to the ultrasonoscopy to obtain characteristics of image；
Described image feature is matched with the pre-set image feature in template database, obtains matching degree；The template number
A variety of abnormal image information and corresponding pre-set image feature are contained according to stock；
When the matching degree is more than predetermined threshold value, pre-set image feature corresponding to maximum matching degree is obtained；
Export abnormal image information corresponding with the pre-set image feature that the matching degree is maximum.
2. ultrasonoscopy processing method according to claim 1, it is characterised in that utilize the ultrasound figure of multiple data centers
As data by machine learning method establish the template database.
3. ultrasonoscopy processing method according to claim 1, it is characterised in that described by described image feature and template
Pre-set image feature in database, which carries out matching, to be included：
Calculate the distance of described image feature and the pre-set image feature in template database；
By the distance compared with preset value, the matching degree is obtained.
4. ultrasonoscopy processing method according to claim 1, it is characterised in that
The feature extraction includes color characteristic and/or textural characteristics and/or shape facility and/or spatial relation characteristics extraction；
The abnormal image information claim including exception name and/or the edge contour of the form of abnormal image and/or abnormal image and/
Or Anomalous typing and/or abnormal classification.
5. ultrasonoscopy processing method according to claim 1, it is characterised in that the ultrasonoscopy processing method is also wrapped
Include step：
When the matching degree is more than predetermined threshold value, the ultrasonoscopy is measured to obtain abnormal image feature, wherein, it is described different
Normal characteristics of image includes the number and/or the size of abnormal image and/or the actual edge profile of abnormal image of abnormal image；
Export the abnormal image feature.
6. ultrasonoscopy processing method according to claim 5, it is characterised in that contrast shows the abnormal image feature
With the abnormal image information.
A kind of 7. ultrasonoscopy processing system, it is characterised in that including
Image acquisition unit, for obtaining the ultrasonoscopy of destination object；
Feature extraction unit, for carrying out feature extraction to the ultrasonoscopy to obtain characteristics of image；
Matching unit, for described image feature to be matched with the pre-set image feature in template database, obtain matching
Degree, when the matching degree is more than predetermined threshold value, obtains pre-set image feature corresponding to maximum matching degree, wherein, the template
Database purchase has a variety of abnormal image information and corresponding pre-set image feature；
Output unit, for exporting abnormal image information corresponding with the pre-set image feature of matching degree maximum.
8. ultrasonoscopy processing system according to claim 7, it is characterised in that the ultrasonoscopy processing system is also wrapped
Include：
Template database establishes unit, for being established using the ultrasound image data of multiple data centers by machine learning method
The template database.
9. ultrasonoscopy processing system according to claim 7, it is characterised in that the ultrasonoscopy processing system is also wrapped
Include：
Image measurement unit, for when the matching degree is more than predetermined threshold value, measuring the ultrasonoscopy to obtain Abnormal Map
As feature.
10. ultrasonoscopy processing system according to claim 9, it is characterised in that the ultrasonoscopy processing system is also
Including：
Display unit, the abnormal image feature and the abnormal image information are shown for contrasting.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710917423.3A CN107679574A (en) | 2017-09-29 | 2017-09-29 | Ultrasonoscopy processing method and system |
| PCT/CN2018/108296 WO2019062843A1 (en) | 2017-09-29 | 2018-09-28 | Ultrasound image processing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710917423.3A CN107679574A (en) | 2017-09-29 | 2017-09-29 | Ultrasonoscopy processing method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107679574A true CN107679574A (en) | 2018-02-09 |
Family
ID=61139094
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710917423.3A Pending CN107679574A (en) | 2017-09-29 | 2017-09-29 | Ultrasonoscopy processing method and system |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN107679574A (en) |
| WO (1) | WO2019062843A1 (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108804547A (en) * | 2018-05-18 | 2018-11-13 | 深圳华声医疗技术股份有限公司 | Ultrasonoscopy teaching method, device and computer readable storage medium |
| CN109101525A (en) * | 2018-06-19 | 2018-12-28 | 黑龙江拓盟科技有限公司 | A kind of medical image comparison method based on image comparison identification |
| CN109498064A (en) * | 2018-12-29 | 2019-03-22 | 深圳开立生物医疗科技股份有限公司 | Ultrasonic scanning control method and ultrasonic diagnostic equipment |
| WO2019062843A1 (en) * | 2017-09-29 | 2019-04-04 | 深圳开立生物医疗科技股份有限公司 | Ultrasound image processing method and system |
| CN109602445A (en) * | 2018-12-06 | 2019-04-12 | 余姚市华耀工具科技有限公司 | Spleen defect detection platform |
| CN109925054A (en) * | 2019-03-04 | 2019-06-25 | 杭州三坛医疗科技有限公司 | Determine the householder method, device and system, readable storage medium storing program for executing in target spot path |
| CN109925053A (en) * | 2019-03-04 | 2019-06-25 | 杭州三坛医疗科技有限公司 | Determination methods, devices and systems, the readable storage medium storing program for executing of operation pathway |
| CN110269641A (en) * | 2019-06-21 | 2019-09-24 | 深圳开立生物医疗科技股份有限公司 | A kind of ultrasonic imaging auxiliary bootstrap technique, system, equipment and storage medium |
| WO2020118534A1 (en) * | 2018-12-11 | 2020-06-18 | 广东医动科技有限公司 | Automatic detection method and system using ultrasonic robot |
| CN111528910A (en) * | 2020-05-12 | 2020-08-14 | 上海深至信息科技有限公司 | Ultrasonic detection system for cataract of eye |
| WO2020177725A1 (en) * | 2019-03-04 | 2020-09-10 | 杭州三坛医疗科技有限公司 | Target path determining method |
| CN112839590A (en) * | 2018-10-08 | 2021-05-25 | 皇家飞利浦有限公司 | Method and system for determining a supplemental ultrasound view |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104680481A (en) * | 2013-11-28 | 2015-06-03 | 深圳迈瑞生物医疗电子股份有限公司 | Ultrasonic auxiliary scanning method and ultrasonic auxiliary scanning system |
| CN106250707A (en) * | 2016-08-12 | 2016-12-21 | 王双坤 | A kind of based on degree of depth learning algorithm process head construction as the method for data |
| CN106510759A (en) * | 2016-11-26 | 2017-03-22 | 汕头市超声仪器研究所有限公司 | Semiautomatic ultrasonic diagnosis method |
| CN106580368A (en) * | 2016-11-26 | 2017-04-26 | 汕头市超声仪器研究所有限公司 | Full-automatic ultrasonic diagnosis method |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070160275A1 (en) * | 2006-01-11 | 2007-07-12 | Shashidhar Sathyanarayana | Medical image retrieval |
| CN105232081A (en) * | 2014-07-09 | 2016-01-13 | 无锡祥生医学影像有限责任公司 | Medical ultrasound assisted automatic diagnosis device and medical ultrasound assisted automatic diagnosis method |
| CN107679574A (en) * | 2017-09-29 | 2018-02-09 | 深圳开立生物医疗科技股份有限公司 | Ultrasonoscopy processing method and system |
-
2017
- 2017-09-29 CN CN201710917423.3A patent/CN107679574A/en active Pending
-
2018
- 2018-09-28 WO PCT/CN2018/108296 patent/WO2019062843A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104680481A (en) * | 2013-11-28 | 2015-06-03 | 深圳迈瑞生物医疗电子股份有限公司 | Ultrasonic auxiliary scanning method and ultrasonic auxiliary scanning system |
| CN106250707A (en) * | 2016-08-12 | 2016-12-21 | 王双坤 | A kind of based on degree of depth learning algorithm process head construction as the method for data |
| CN106510759A (en) * | 2016-11-26 | 2017-03-22 | 汕头市超声仪器研究所有限公司 | Semiautomatic ultrasonic diagnosis method |
| CN106580368A (en) * | 2016-11-26 | 2017-04-26 | 汕头市超声仪器研究所有限公司 | Full-automatic ultrasonic diagnosis method |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019062843A1 (en) * | 2017-09-29 | 2019-04-04 | 深圳开立生物医疗科技股份有限公司 | Ultrasound image processing method and system |
| CN108804547A (en) * | 2018-05-18 | 2018-11-13 | 深圳华声医疗技术股份有限公司 | Ultrasonoscopy teaching method, device and computer readable storage medium |
| CN109101525A (en) * | 2018-06-19 | 2018-12-28 | 黑龙江拓盟科技有限公司 | A kind of medical image comparison method based on image comparison identification |
| CN112839590A (en) * | 2018-10-08 | 2021-05-25 | 皇家飞利浦有限公司 | Method and system for determining a supplemental ultrasound view |
| CN109602445A (en) * | 2018-12-06 | 2019-04-12 | 余姚市华耀工具科技有限公司 | Spleen defect detection platform |
| WO2020118534A1 (en) * | 2018-12-11 | 2020-06-18 | 广东医动科技有限公司 | Automatic detection method and system using ultrasonic robot |
| CN109498064A (en) * | 2018-12-29 | 2019-03-22 | 深圳开立生物医疗科技股份有限公司 | Ultrasonic scanning control method and ultrasonic diagnostic equipment |
| CN109925054A (en) * | 2019-03-04 | 2019-06-25 | 杭州三坛医疗科技有限公司 | Determine the householder method, device and system, readable storage medium storing program for executing in target spot path |
| CN109925053A (en) * | 2019-03-04 | 2019-06-25 | 杭州三坛医疗科技有限公司 | Determination methods, devices and systems, the readable storage medium storing program for executing of operation pathway |
| WO2020177725A1 (en) * | 2019-03-04 | 2020-09-10 | 杭州三坛医疗科技有限公司 | Target path determining method |
| CN110269641A (en) * | 2019-06-21 | 2019-09-24 | 深圳开立生物医疗科技股份有限公司 | A kind of ultrasonic imaging auxiliary bootstrap technique, system, equipment and storage medium |
| CN111528910A (en) * | 2020-05-12 | 2020-08-14 | 上海深至信息科技有限公司 | Ultrasonic detection system for cataract of eye |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019062843A1 (en) | 2019-04-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107679574A (en) | Ultrasonoscopy processing method and system | |
| Sobhaninia et al. | Fetal ultrasound image segmentation for measuring biometric parameters using multi-task deep learning | |
| Wang et al. | Automatic analysis of lateral cephalograms based on multiresolution decision tree regression voting | |
| US9256941B2 (en) | Microcalcification detection and classification in radiographic images | |
| Durrleman et al. | Registration, atlas estimation and variability analysis of white matter fiber bundles modeled as currents | |
| Durrleman et al. | Inferring brain variability from diffeomorphic deformations of currents: an integrative approach | |
| US20080292194A1 (en) | Method and System for Automatic Detection and Segmentation of Tumors and Associated Edema (Swelling) in Magnetic Resonance (Mri) Images | |
| Ma et al. | Hierarchical segmentation and identification of thoracic vertebra using learning-based edge detection and coarse-to-fine deformable model | |
| Hasan et al. | Automated screening of MRI brain scanning using grey level statistics | |
| CN105931225B (en) | A method of based on realtime graphic detection technique analyzing crystal growth form and Size Distribution | |
| CN116664559B (en) | Machine vision-based memory bank damage rapid detection method | |
| CN104881680A (en) | Alzheimer's disease and mild cognitive impairment identification method based on two-dimension features and three-dimension features | |
| WO2015106374A1 (en) | Multidimensional texture extraction method based on brain nuclear magnetic resonance images | |
| WO2019238104A1 (en) | Computer apparatus and method for implementing classification detection of pulmonary nodule images | |
| Toews et al. | A feature-based approach to big data analysis of medical images | |
| Xie et al. | Fabric defect detection method combing image pyramid and direction template | |
| Demirci et al. | Cortical thickness systematically varies with curvature and depth in healthy human brains | |
| Mustapha et al. | Design and development of a content-based medical image retrieval system for spine vertebrae irregularity | |
| CN110660044B (en) | Method for rapidly detecting bone tissue structural morphological abnormality and electronic device | |
| Daenzer et al. | VolHOG: a volumetric object recognition approach based on bivariate histograms of oriented gradients for vertebra detection in cervical spine MRI | |
| Abeysinghe et al. | Segmentation-free skeletonization of grayscale volumes for shape understanding | |
| CN105488798B (en) | SAR image method for measuring similarity based on point set contrast | |
| CN107590806B (en) | Detection method and system based on brain medical imaging | |
| Matuk et al. | Biomedical applications of geometric functional data analysis | |
| Qi et al. | Age estimation from MR images via 3D convolutional neural network and densely connect |
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 |
Application publication date: 20180209 |
|
| RJ01 | Rejection of invention patent application after publication |