CN108491770A - A kind of data processing method based on fracture image - Google Patents
A kind of data processing method based on fracture image Download PDFInfo
- Publication number
- CN108491770A CN108491770A CN201810191832.4A CN201810191832A CN108491770A CN 108491770 A CN108491770 A CN 108491770A CN 201810191832 A CN201810191832 A CN 201810191832A CN 108491770 A CN108491770 A CN 108491770A
- Authority
- CN
- China
- Prior art keywords
- image
- fracture
- data processing
- line
- ray
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- 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
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The present invention relates to a kind of data processing methods based on fracture image, include the following steps：a）Fracture image is imported into data processing centre；b）Image recognition is carried out to the image of importing；c）Predefined parameter value measurement is carried out to the image after identification, obtains one group of data value；d）By step c）In obtained data value handled after export.Orthopaedics doctor can be helped to interpret blueprints and judge to skeletal image the state of an illness by this method, greatly improve the working efficiency and quality of doctor.
Description
Technical field
The invention belongs to field of medical technology more particularly to a kind of data processing methods based on fracture image.
Background technology
As living-pattern preservation and the exacerbation of aging, China's incidence of fracture increase year by year.In the diagnosis of the disease
And in treatment, X-ray is of great significance, the indispensable inspection in almost all patient's diagnosis and treatment processes.However it is limited to China
Medical environment and medical level, fracture X-ray can be accomplished accurately to read, make full use of and be in actual clinical work
One thing for being difficult.Therefore, the mistaken diagnosis of China's fracture patient, rate of missed diagnosis are higher, not only bring massive losses to patient health,
Also add the risk of medical tangle.And the therapy of fracture is various, it is desirable that doctor makes a concrete analysis of according to conditions of patients, this
The diagnosis and treatment doctor not only sturdy professional skill of to master is needed, the experiences in diagnosis and treatment for having abundant is also needed, once diagnosis and treatment method is wrong,
Serious consequence equally can be caused to doctor and patient.Problem above is all badly in need of being resolved by effective measures.
Fracture image can show quantity, form, position and the relationship with surrounding structure of fracture line, and can be by X-ray
The measurement of all multi-parameters of piece further quantifies the degree of displacement fracture or angulation.However to X-ray of fracturing in actual clinical work
It is a thing for being difficult that can accomplish accurately to read, make full use of.The application process of fracture X-ray is artificial reading at present
Figure, this mode have the following problems：
1. strongly professional
For shifting the case apparent, wound is larger, most doctors can accurately discern whether there is fracture.
But due to the uniqueness of instrument of skeleton anatomical structure, irregular, the misaligned fracture site for certain forms, or displacement is not
Apparent fracture, the insufficient doctor of part professional experiences are just difficult to provide correct diagnosis.
2. wrong diagnosis and escape
The fracture patient state of an illness occurs suddenly, especially the patient of multiple injury, wound weight and situation complexity, gold consultation time
It is of short duration.And emergency treatment heavy workload, every patient see and treat patients limited time, can not measure certain important images ginsengs in detail and accurately
Number, this can influence diagnosis and treatment treatment, cause wrong diagnosis and escape, miss patient's best occasion for the treatment.
3. parting is complicated
AO partings of fracturing have great importance for disease treatment, but parting rule is complicated, and hardly possible note is easily forgotten, even if
Be veteran orthopaedics doctor also often can caused by the reasons such as forgeing, neglecting parting mistake, influence successive treatment.
Invention content
In order to solve the above technical problems, the present invention provides a kind of data processing method based on fracture image.Pass through the party
Method can help orthopaedics doctor to interpret blueprints and judge to skeletal image the state of an illness, greatly improve the working efficiency and quality of doctor.
The present invention is achieved through the following technical solutions：A kind of data processing method based on fracture image, including
Following steps：
A) fracture image is imported into data processing centre；
B) image recognition is carried out to the image of importing；
C) predefined parameter value measurement is carried out to the image after identification, obtains one group of data value；
D) it is exported after being handled the data value obtained in step c).
Further, the fracture image is X-ray.
Further, in step b), the image of described pair of importing carries out image recognition, further includes step：
B1 bone contours line) is indicated on the X-ray；
B2) fracture line is indicated on the X-ray；
B3 tag line and/or identification point) are indicated on the X-ray；
B4) each parameter value and the tag line or identification point are partly or entirely shown on the X-ray.
Further, step b) further includes：B5 the quantity and/or fracture line and tag line or identification point of fracture line) are identified
Position relationship.
Further, step c) further includes, according to different fracture parting demands, measuring the ginseng needed for corresponding parting
Number, and be stored in preset database.
Further, the Data processing is intracardiac is preset with fracture typing data library, and in step d), after processing
The data value and the typing data library carry out Data Matching, genotyping result is exported according to matching result.
Further, the typing data library includes AO typing datas library.
Further, the Data processing it is intracardiac be also preset with it is right for each type of fracture institute under different bone partings
The therapeutic scheme answered exports corresponding therapeutic scheme while exporting AO genotyping results together.
Further, step b) carries out image recognition, including following pre-treatment step to the image of importing：
B01) first described image is filtered, denoising, the transformation of re-segmenting linear gradation；
B02 image segmentation, extraction edge) are carried out after the above step, then carry out morphology after carrying out Connected area disposal$
Filtering；
B03 data) are moved into contour shape from image block；
B04 the Combined expression layer for) learning image block texture mode and contour shape mode by CNN, forms from image block
To the access of texture；
B05 contours extract) is carried out along the access, obtains the contour shape of skeletal structure.
Further, step b03) further include：
Image-zooming is carried out to the fracture image and obtains interested region.
Further, step b) further includes：B06 instrument of skeleton anatomical structure in image) is obtained using multiple linear regression grader
Possibility combination.
Further, by the way of human-computer interaction, human assistance indicates bone contours line on the X-ray；With/
Or,
By the way of human-computer interaction, human assistance indicates fracture line on the X-ray；And/or
By the way of human-computer interaction, human assistance indicates tag line and/or identification point on the X-ray.
Further, further include that the data value storage after measuring is gone forward side by side in data processing centre's database in step c)
Row Classification Management.
Further, further include usage time recurrent neural network in step c), to add label per an example image, and will
Image and the corresponding data value result under the label are deposited in data processing centre's database.
Further, the fracture image is wrist fracture image or intertrochanteric fracture image.
Therefore, a kind of utilization " computer/artificial intelligence identification-survey calculation-analysis " system of the present invention, to patients with fractures
X-ray progress is quick, accurately identifies and measures, and obtains related important parameter, comprehensive analysis derives fracture parting, and can be based on
The method that parting proposes therapeutic scheme.The diagosis speed and precision of fracture X-ray image can be improved in the invention, solves artificial read tablet and misses
The high problem of rate of missed diagnosis is examined, improves emergency treatment and orthopedist to the medical efficiency and quality of medical care in fracture diagnosis and treatment process.
Description of the drawings
Fig. 1 is the flow chart of the method for the present invention.
Specific implementation mode
The present invention will be described in detail below with reference to the accompanying drawings and embodiments.Below in conjunction with attached drawing to the principle of the present invention
It is described with feature, it should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application
It can be combined with each other.The given examples are served only to explain the present invention, is not intended to limit the scope of the present invention.
Classification of fracture is varied, and fracture of distal radius is most common fracture in clinical upper limb fracture, accounts about complete
The AO partings of the 1/6 of body fracture, fracture of distal radius have important directive significance for the treatment of fracture.Therefore, in order to more
Good explanation technical solution disclosed in this invention below has the present invention as specific embodiment to choose fracture of distal radius
Body explanation.
Explanation of nouns：
Fracture：Fracture is that the continuity of bone structure is broken completely or partially.It can be caused by violence or the strain of accumulation property, often
For-a position fracture, minority is multiple fracture.It can restore original function through appropriate processing, most patients in time, it is a small number of
Patient can leave different degrees of sequelae.
Fracture of distal radius：Fracture within span distal aspect of the radius 3cm.
Link together referred to as joint between bone and bone, and articular surface is the contact surface for each related bone for participating in composition joint.
Center reference point (CRP)：As upper, lunar incisura slaps the midpoint of line between cornu dorsale for wrist joint positive side position.It can be more
It is accurate to measure oar inclination angle and ulnar variance, center reference point generally within the coronal-plane at lunar incisura center, do not inclined by the palm or
The back of the body inclines to determine.
Oar inclination angle (ruler drift angle) refers to the line and radius long axis from processus styloideus radii to CRP as an anatomic parameter
Formed angle between vertical line.Normal wrist oar inclination angle is about 24 °.In general, fracture of distal radius patient's oar inclination angle<15 ° of tools
There is opposite operative indication.
Radius height：The cripetura situation of evaluation radius is also used for as another parameter.It is by measuring perpendicular to reference line
The distance between two straight lines of Radial long axis can obtain concrete numerical value, this two lines is respectively to pass through processus styloideus radii vertex
Vertical line and by CRP vertical line.The normal value of radius height is 11.6mm.
Slap inclination angle：Refer to that the so-called angle of dorsal border line is slapped in the vertical line of Radial axis and side position as in.Normal carpal is closed
Section palm inclination angle is about 10 °.As in, palm inclination angle is used to measure the angulation situation of articular surface for side position.The fracture of palmar displacement is usual
It is shown as the increase at palm inclination angle, these fracture are extremely unstable, need a degree of fixation.
Fracture line：Fracture line is exactly the line not healed that the position fractured is presented in X on pieces, it is that clinically judgement is
The important symbol of no fracture, if fracture line, which does not disappear, illustrates that fracture does not heal also.It is generally divided into completeness fracture according to fracture line
And incomplete fracture.
Transverse fracture：Fracture line almost with backbone axis oriented normal.
Oblique fracture：Fracture line and backbone longitudinal axis out of plumb.
Spiral fracture：Fracture line is twist.
Comminuted fracture：Fragment of fracturing is more than 2 pieces, and as when fracture line is T-shaped or Y shape, also known as T shapes or Y shape are fractured.
Intercalation is fractured：After fracture, compact bone substance is inserted into cancellous bone.
Artificial intelligence：Artificial intelligence (Artificial Intelligence), english abbreviation AI.It is research, exploitation
Theory, a new technological sciences of method, technology and application system for the intelligence for simulating, extending and extend people.Manually
Intelligence is a branch of computer science, it attempts to understand essence of intelligence, and produce it is a kind of it is new can be with human intelligence
The research of the intelligence machine that similar mode is made a response, the field includes robot, language identification, image recognition, natural language
Speech processing and expert system etc..Artificial intelligence can be to the simulation of the information process of the consciousness of people, thinking.Artificial intelligence is not people
Intelligence, but can think deeply as people, may also be more than people intelligence.
Convolutional neural networks (Convolutional Neural Network, CNN) are a kind of feedforward neural networks, it
Artificial neuron can respond the surrounding cells in a part of coverage area, have outstanding performance for large-scale image procossing.It is wrapped
Include convolutional layer (convolutional layer) and pond layer (pooling layer).
RCNN (Region-based Convolutional Neural Network), using convolutional neural networks, plays a game
Portion's area information is extracted to carry out a kind of method of target detection.The process main points of RCNN can be simply expressed as：
1. carrying out region of interest ROI to input picture using selective search methods to extract
2. the region of extraction is put into the CNN of pre-training and carries out feature extraction
3. SVM classifier is carried out fine-tuning, output label is changed to N+1
4. train this SVM, with trained SVM for CNN come out feature vector obtain detection target classification and
The recurrence of bounding box.
Fast-RCNN, Faster-RCNN are further improved image recognition algorithms on the basis of RCNN, improve survey
Speed when examination and training, and reduce trained space.
Attached drawing 1 is please referred to, it is specific for a kind of one kind of the data processing method based on fracture image provided by the present invention
The flow chart of embodiment.
One, image obtains
The normotopia X-ray of wrist joint image, side position X-ray are imported into data processing centre, if it is necessary, can also be it
Its X-ray importing is entered.
Lead-in mode can be introduced directly into after being obtained from hospital imaging department PACS system using DICOM systems, can also
It is to be imported after being shot by way of shooting picture.The data processing centre is used to carry out data processing to the image of importing, can
Can also be the processing system of the mobile devices such as mobile phone, tablet computer to be computer software.The data processing centre can
Can also be that image is sent to Data processing by way of telecommunication so that treated that display end is integrated with data
After the heart, required result is back to the transmitting terminal after data processing centre is handled.
For by the bone of type, required X-ray is not different yet, some bones may need normotopia and side position
Two, some bones may only need a normotopia, also some light to have normotopia and side position not enough, may also need
Want loxosis or other images.
Two, image recognition
Image recognition is carried out to the image of importing, needs to pre-process image, include the following steps：
B01) first described image is filtered, the transformation of re-segmenting linear gradation；
B02 image segmentation, extraction edge) are carried out after the above step, then carry out morphology after carrying out Connected area disposal$
Filtering；
B03) data are moved from image block using multi-modal depth Boltzmann machine (Deep Boltzmann Machine)
Move on to contour shape；
B04 the Combined expression layer for) learning image block texture mode and contour shape mode by CNN, forms from image block
To the access of texture；
B05) gibbs sampler (Gibbs Sampling) method is used to carry out contours extract along the access, obtains bone
The contour shape of structure.
B06 multiple linear regression grader) is used to obtain the possibility combination of instrument of skeleton anatomical structure in image.
After image preprocessing, the operations such as crossed, identified to image are needed, are specifically included：
B1 bone contours line) is indicated on the X-ray；
B2) fracture line is indicated on the X-ray；
B3 tag line and/or identification point) are indicated on the X-ray；
B4) each parameter value and the tag line or identification point are partly or entirely shown on the X-ray；
B5 the quantity of fracture line and the position relationship of fracture line and tag line or identification point) are identified.
Above-mentioned tag line and identification point, according to different bone types, tag line and identification point are also different, and wrist is closed
For section, when carrying out the data processing of fracture image, need to identify center reference point (CRP), which is identification point, should
Reference point is located at wrist joint positive side position as upper, and lunar incisura slaps the midpoint of line between cornu dorsale.It can more acurrate measurement oar inclination angle
And ulnar variance, center reference point are not inclined or are carried on the back by the palm and inclined to determine generally within the coronal-plane at lunar incisura center.
In addition, the angle line of oar inclination angle (ruler drift angle), refers to the line from processus styloideus radii to CRP as an anatomic parameter
The formed angle between the vertical line of radius long axis, you can be marked as a tag line.Normal wrist oar inclination angle
About 24 °.In general, fracture of distal radius patient's oar inclination angle<15 ° have opposite operative indication.
Another parameter is radius height, and the cripetura situation of evaluation radius is also used for as another parameter.Pass through measurement
The distance between two straight lines perpendicular to reference line, that is, Radial long axis can obtain concrete numerical value, this two lines is respectively to pass through
Cross the vertical line on processus styloideus radii vertex and the vertical line by CRP.The normal value of radius height is 11.6mm.
During carrying out image recognition, if there is the growth and development situation of a few bodies is more special, it is therefore possible to use often
Rule algorithm can not automatic identification or doctor have preference, can also be by the way of human-computer interaction, by doctor in the X-ray
Upper human assistance indicates bone contours line, and either the human assistance on the X-ray indicates fracture line or in the X-ray
On piece human assistance indicates tag line and/or identification point.
During being identified, if ontogeny situation is very normal, image data is also more clear, is easy to know
Not, can operation be identified by computer completely.
In another specific implementation mode, a part can also be needed by doctor special by the way of human-computer interaction
The line or point of mark carry out auxiliary drafting, remaining is by computer digital animation center processing.
Above-mentioned several different modes can be met more demands, be improved this method by doctor's voluntarily selection operation
The scope of application.
The method being identified using the above-mentioned image to importing, accuracy is high, and adaptive ability is strong, and processing speed
Efficient quick.
Three, image analysis, parameter measurement
Image recognition work after the completion of, so that it may with radius structure is analyzed according to a series of constraintss, is measured with
And displacement fracture or angulation degree are quantified.Image analysis includes excluding background to mix, and identifies key area, ulna, oar
Bone, ulnoradial joint and the identification of surrounding carpal bone, abnormal fracture line, sclerite identification, measure relevant parameter (quantity of fracture line, position
Set and inclination angle slapped in joint relation of plane, radius height, ruler drift angle), as parting input parameter.The concrete meaning of each parameter is asked
Referring to above-mentioned explanation of nouns.
Since doctor for patient when performing an operation, only also it is difficult to be implemented according to simple operation plan, in order to
Easily check that image, data processing centre can add tag line or mark on image at any time during surgery convenient for doctor
Point, as needed, what doctor can be selective, each parameter value and the tag line or identification point are partly or entirely shown
In on the image.In this way, can directly print with tagged image, more convenient doctor looks at any time when performing the operation
It sees, and eliminates various drawbacks of artificial crossed.
Four, data analysis
According to parameter value measured directly, partial data is needed to obtain new data after carrying out data analysis, to meet
The demand of doctor facilitates doctor to judge the state of an illness according to analysis result.
Five, result exports
In terms of exporting result, in order to maximumlly reduce the workload of doctor, medical efficiency, data processing centre are improved
It is inside preset with the database of various bone analysis, such as AO analytical databases, the data value and the parting to treated
Database carries out Data Matching and storage, is exported genotyping result according to matching result.
Therapeutic scheme is fast and accurately directly obtained in order to facilitate doctor, the Data processing is intracardiac to be also preset with needle
To the therapeutic scheme corresponding to each type under various bone partings, while exporting genotyping result, by corresponding therapeutic scheme
It exports together.
In addition, in order to improve accuracy and the efficiency of processing system, after the image picture of one case of every processing, use
Time recurrent neural network, to add label per an example image, and by under the label image and the corresponding data value
As a result it deposits in data processing centre's database.In this manner it is possible to data are constantly accumulated, the continuous artificial intelligence for enhancing system
Energy deep learning effect, and then image processing speed can be made to be getting faster, and also it is more and more accurate.
Method provided by the present invention passes through CNN image recognition series techniques (RCNN, Fast-CNN, Faster- first
CNN the primary image information for) obtaining wrist joint positive side position X-ray obtains the specific of fracture line by being further processed to image
Positioning by computer automatic analysis, is calculated the above-mentioned key parameter of acquisition, is further derived using parameter and can get specific AO
Parting or other partings finally propose optimal treatment suggestion according to parting.
The explanation that above-mentioned specific embodiment is only carried out using wrist joint bone as example, this method can be used for femur
The processing of the image data of intertrochanteric fracture image or other bones, details are not described herein for specific method.
Data processing method disclosed in this invention based on fracture image, is suitable for auxiliary diagnosis, the parting in fracture
And auxiliary treatment, patients with fractures's X-ray can be carried out quickly, accurately identify and measure, obtain related important parameter, comprehensive point
Fracture parting is derived in analysis, and a variety of parting schemes can be provided according to clinical demand to same fracture X-ray, and can
Corresponding therapeutic scheme is proposed based on each parting, for assisting operation conceptual design.Fracture X-ray image can be improved in the invention
Diagosis speed and precision can fast, accurately obtain the critical data of fracture x-ray picture, solve artificial read tablet Misdiagnosis
The high problem of rate improves emergency treatment and orthopedist to the medical efficiency and quality of medical care in fracture diagnosis and treatment process.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiments or example.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, changes, replacing and modification.
Claims (10)
1. a kind of data processing method based on fracture image, which is characterized in that include the following steps：
A) fracture image is imported into data processing centre；
B) image recognition is carried out to the image of importing；
C) predefined parameter value measurement is carried out to the image after identification, obtains one group of data value；
D) it is exported after being handled the data value obtained in step c).
2. the data processing method according to claim 1 based on fracture image, which is characterized in that the fracture image is
X-ray.
3. the data processing method according to claim 2 based on fracture image, which is characterized in that described in step b)
Image recognition is carried out to the image of importing, further includes step：
B1 bone contours line) is indicated on the X-ray；
B2) fracture line is indicated on the X-ray；
B3 tag line and/or identification point) are indicated on the X-ray；
B4) each parameter value and the tag line or identification point are partly or entirely shown on the X-ray.
4. the data processing method according to claim 3 based on fracture image, which is characterized in that step b) further includes：
B5 the quantity of fracture line and/or the position relationship of fracture line and the tag line and/or identification point) are identified.
5. the data processing method according to claim 4 based on fracture image, which is characterized in that step c) further includes,
According to different fracture parting demands, the parameter needed for corresponding parting is measured, and be stored in preset database.
6. the data processing method according to claim 1 based on fracture image, which is characterized in that the Data processing
It is intracardiac to be preset with fracture typing data library, and in step d), the data value and the typing data library to treated
Data Matching is carried out, is exported genotyping result according to matching result；The typing data library includes AO typing datas library；The number
According to being also preset in processing center for the therapeutic scheme corresponding to each type of fracture under different bone partings, output AO divides
While type result, corresponding therapeutic scheme is exported together.
7. the data processing method according to claim 1 based on fracture image, which is characterized in that step b) is to importing
Image carries out image recognition, including following pre-treatment step：
B01) first described image is filtered, denoising, the transformation of re-segmenting linear gradation；
B02 image segmentation, extraction edge) are carried out after the above step, then carry out morphologic filter after carrying out Connected area disposal$；
B03 data) are moved into contour shape from image block；
B04 the Combined expression layer for) learning image block texture mode and contour shape mode by CNN, forms from image block to line
The access of reason；
B05 contours extract) is carried out along the access, obtains the contour shape of skeletal structure；
Step b03) further include：Image-zooming is carried out to the fracture image and obtains interested region；
B06 multiple linear regression grader) is used to obtain the possibility combination of instrument of skeleton anatomical structure in image.
8. a kind of data processing method based on fracture image according to claim 3, which is characterized in that
By the way of human-computer interaction, human assistance indicates bone contours line on the X-ray；And/or
By the way of human-computer interaction, human assistance indicates fracture line on the X-ray；And/or
By the way of human-computer interaction, human assistance indicates tag line and/or identification point on the X-ray.
9. the data processing method based on fracture image according to claim 1-8, which is characterized in that in step c), also
Including
Data value storage after measurement in data processing centre's database and is subjected to Classification Management；
Usage time recurrent neural network, to add label per an example image, and by under the label image and corresponding institute
Data value result is stated to deposit in data processing centre's database.
10. the data processing method based on fracture image according to claim 1-9, which is characterized in that the fracture shadow
As being wrist fracture image or intertrochanteric fracture image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810191832.4A CN108491770B (en) | 2018-03-08 | 2018-03-08 | Data processing method based on fracture image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810191832.4A CN108491770B (en) | 2018-03-08 | 2018-03-08 | Data processing method based on fracture image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108491770A true CN108491770A (en) | 2018-09-04 |
| CN108491770B CN108491770B (en) | 2023-05-30 |
Family
ID=63338047
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810191832.4A Active CN108491770B (en) | 2018-03-08 | 2018-03-08 | Data processing method based on fracture image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108491770B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109106481A (en) * | 2018-09-18 | 2019-01-01 | 北京爱康宜诚医疗器材有限公司 | The determination method and device of acetabular bone defect degree |
| CN109124836A (en) * | 2018-09-18 | 2019-01-04 | 北京爱康宜诚医疗器材有限公司 | The determination method and device of acetabular bone defect processing mode |
| CN111354057A (en) * | 2020-03-10 | 2020-06-30 | 中南大学 | Bone fracture line map drawing method based on image deformation technology |
| TWI701680B (en) * | 2018-08-19 | 2020-08-11 | 長庚醫療財團法人林口長庚紀念醫院 | Method and system of analyzing medical images |
| CN111899880A (en) * | 2020-08-03 | 2020-11-06 | 暨南大学附属第一医院（广州华侨医院） | Lumbar vertebra trabecular load stress change and hidden fracture artificial risk assessment method |
| CN111933281A (en) * | 2020-09-30 | 2020-11-13 | 平安科技（深圳）有限公司 | Disease typing determination system, method, device and storage medium |
| CN112435269A (en) * | 2020-12-02 | 2021-03-02 | 山东中医药大学 | Distal radius fracture image processing method based on fast-RCNN |
| CN113194864A (en) * | 2019-01-18 | 2021-07-30 | 加图立大学校产学协力团 | Virtual internal fixture generation method and device based on image restoration |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002096280A2 (en) * | 2001-05-29 | 2002-12-05 | Innovationsagentur Gesellschaft M.B.H. | Method and device for obtaining information relating to a bone fracture |
| EP1293925A1 (en) * | 2001-09-18 | 2003-03-19 | Agfa-Gevaert | Radiographic scoring method |
| CN1682236A (en) * | 2002-08-20 | 2005-10-12 | 成像治疗仪股份有限公司 | Methods and devices for analysis of X-ray images |
| US20070274584A1 (en) * | 2004-02-27 | 2007-11-29 | Leow Wee K | Method and System for Detection of Bone Fractures |
| CN101452577A (en) * | 2008-11-26 | 2009-06-10 | 沈阳东软医疗系统有限公司 | Rib auto-demarcating method and device |
| US20140233820A1 (en) * | 2012-11-01 | 2014-08-21 | Virginia Commonweath University | Segmentation and Fracture Detection in CT Images |
| CN106023094A (en) * | 2016-05-10 | 2016-10-12 | 中南大学 | Image-based bone tissue microstructure restoration system and restoration method thereof |
| CN107154038A (en) * | 2016-04-22 | 2017-09-12 | 孔德兴 | A kind of visual fracture of rib aided diagnosis method of rib |
-
2018
- 2018-03-08 CN CN201810191832.4A patent/CN108491770B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002096280A2 (en) * | 2001-05-29 | 2002-12-05 | Innovationsagentur Gesellschaft M.B.H. | Method and device for obtaining information relating to a bone fracture |
| EP1293925A1 (en) * | 2001-09-18 | 2003-03-19 | Agfa-Gevaert | Radiographic scoring method |
| CN1682236A (en) * | 2002-08-20 | 2005-10-12 | 成像治疗仪股份有限公司 | Methods and devices for analysis of X-ray images |
| US20070274584A1 (en) * | 2004-02-27 | 2007-11-29 | Leow Wee K | Method and System for Detection of Bone Fractures |
| CN101452577A (en) * | 2008-11-26 | 2009-06-10 | 沈阳东软医疗系统有限公司 | Rib auto-demarcating method and device |
| US20140233820A1 (en) * | 2012-11-01 | 2014-08-21 | Virginia Commonweath University | Segmentation and Fracture Detection in CT Images |
| CN107154038A (en) * | 2016-04-22 | 2017-09-12 | 孔德兴 | A kind of visual fracture of rib aided diagnosis method of rib |
| CN106023094A (en) * | 2016-05-10 | 2016-10-12 | 中南大学 | Image-based bone tissue microstructure restoration system and restoration method thereof |
Non-Patent Citations (2)
| Title |
|---|
| N. UMADEVI等: "Multiple classification system for fracture detection in human bone x-ray images", 《2012 THIRD INTERNATIONAL CONFERENCE ON COMPUTING, COMMUNICATION AND NETWORKING TECHNOLOGIES (ICCCNT\"12)》 * |
| 卫娇: "基于BP神经网络的CT图像骨皮质分割", 《医用生物力学》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI701680B (en) * | 2018-08-19 | 2020-08-11 | 長庚醫療財團法人林口長庚紀念醫院 | Method and system of analyzing medical images |
| CN109106481A (en) * | 2018-09-18 | 2019-01-01 | 北京爱康宜诚医疗器材有限公司 | The determination method and device of acetabular bone defect degree |
| CN109124836A (en) * | 2018-09-18 | 2019-01-04 | 北京爱康宜诚医疗器材有限公司 | The determination method and device of acetabular bone defect processing mode |
| CN113194864A (en) * | 2019-01-18 | 2021-07-30 | 加图立大学校产学协力团 | Virtual internal fixture generation method and device based on image restoration |
| CN111354057A (en) * | 2020-03-10 | 2020-06-30 | 中南大学 | Bone fracture line map drawing method based on image deformation technology |
| CN111899880A (en) * | 2020-08-03 | 2020-11-06 | 暨南大学附属第一医院（广州华侨医院） | Lumbar vertebra trabecular load stress change and hidden fracture artificial risk assessment method |
| CN111933281A (en) * | 2020-09-30 | 2020-11-13 | 平安科技（深圳）有限公司 | Disease typing determination system, method, device and storage medium |
| CN112435269A (en) * | 2020-12-02 | 2021-03-02 | 山东中医药大学 | Distal radius fracture image processing method based on fast-RCNN |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108491770B (en) | 2023-05-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108491770A (en) | A kind of data processing method based on fracture image | |
| Tanzi et al. | Hierarchical fracture classification of proximal femur X-Ray images using a multistage Deep Learning approach | |
| Ogiela et al. | Image languages in intelligent radiological palm diagnostics | |
| Liu et al. | Artificial intelligence to detect the femoral intertrochanteric fracture: The arrival of the intelligent-medicine era | |
| CN110189324B (en) | Medical image processing method and processing device | |
| CN108309334A (en) | A kind of data processing method of spine X-ray image | |
| Mittal et al. | Bone Fracture Segmentation Using Cascaded Convolutional Neural Networks | |
| Abbas et al. | Lower leg bone fracture detection and classification using faster RCNN for X-rays images | |
| Sathish Kumar et al. | A comparative experimental analysis and deep evaluation practices on human bone fracture detection using x‐ray images | |
| Zeng et al. | TUSPM-NET: A multi-task model for thyroid ultrasound standard plane recognition and detection of key anatomical structures of the thyroid | |
| CN110610766A (en) | Apparatus and storage medium for deriving probability of disease based on symptom feature weight | |
| Wang et al. | Accuracy and reliability analysis of a machine learning based segmentation tool for intertrochanteric femoral fracture CT | |
| Pham et al. | Chest x-rays abnormalities localization and classification using an ensemble framework of deep convolutional neural networks | |
| Poojary et al. | Optimization Technique Based Approach for Image Segmentation | |
| Duryea et al. | Neural network based automated algorithm to identify joint locations on hand/wrist radiographs for arthritis assessment | |
| CN111127636B (en) | Intelligent complex intra-articular fracture desktop-level three-dimensional diagnosis system | |
| Hohlmann et al. | Segmentation of the Scaphoid Bone in Ultrasound Images: A comparison of two machine learning architectures for in-vivo segmentation | |
| Chang et al. | CT manifestations of gallbladder carcinoma based on neural network | |
| Wang et al. | Deep Learning‐Based Postoperative Recovery and Nursing of Total Hip Arthroplasty | |
| Wang | Anomaly detection of arm X-Ray based on deep learning | |
| CN111209945A (en) | AI-based medical image auxiliary identification method and system for department of imaging | |
| Shukla et al. | COVID-19 detection using raw chest x-ray images | |
| Lydia et al. | Analysis of Advanced Deep Learning Approaches for the Multiple Bone Fracture detection | |
| CN117542528B (en) | Ankylosing spondylitis hip joint affected risk marking system based on image histology | |
| Nanthini et al. | Application of CNN and Recurrent Neural Network Method for Osteosarcoma Bone Cancer Detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |