CN109410267A - A kind of coronary artery segmentation appraisal procedure and system - Google Patents
A kind of coronary artery segmentation appraisal procedure and system Download PDFInfo
- Publication number
- CN109410267A CN109410267A CN201811162031.1A CN201811162031A CN109410267A CN 109410267 A CN109410267 A CN 109410267A CN 201811162031 A CN201811162031 A CN 201811162031A CN 109410267 A CN109410267 A CN 109410267A
- Authority
- CN
- China
- Prior art keywords
- center line
- coronary artery
- noise
- prediction
- line
- 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
- 210000004351 coronary vessel Anatomy 0.000 title claims abstract description 148
- 230000011218 segmentation Effects 0.000 title claims abstract description 74
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000011156 evaluation Methods 0.000 claims abstract description 71
- 230000002159 abnormal effect Effects 0.000 claims abstract description 57
- 238000003776 cleavage reaction Methods 0.000 claims 1
- 230000007017 scission Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 16
- 238000004364 calculation method Methods 0.000 description 12
- 210000001367 artery Anatomy 0.000 description 4
- 208000029078 coronary artery disease Diseases 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 210000000709 aorta Anatomy 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000002059 diagnostic imaging Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000011800 void material Substances 0.000 description 2
- 238000002583 angiography Methods 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 210000004204 blood vessel Anatomy 0.000 description 1
- 238000002591 computed tomography Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000002651 drug therapy Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000005311 nuclear magnetism Effects 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses a kind of coronary artery segmentation appraisal procedure and systems.The appraisal procedure includes: to carry out central line pick-up to coronary artery prediction dividing body, obtains prediction center line；Mark center line is generated using label dividing body；The prediction center line and the mark center line are compared, with abnormal position present in the determination prediction center line；The evaluation index of the coronary artery segmentation is determined according to identified abnormal position.By the appraisal procedure and system, it can reflect the case where coronary artery predicts the noise of dividing body, fracture, false tendency, the exception occurred when dividing to coronary artery makes effective assessment in automation coronary artery reconstruction process.
Description
Technical field
The present invention relates to medical imaging technology field more particularly to a kind of coronary artery segmentation appraisal procedures and system.
Background technique
In the past 50 years, the disease incidence of global coronary heart disease persistently rises, Coronary Artery Disease Intervention Treatment, surgical operation and drug therapy
Means are developed rapidly.There is a kind of dependence and expectation to the operative treatment of coronary heart disease in doctor and patient, this is preced with from the whole world
The patient numbers that shape arteries is rebuild are continuously increased and can find out, this technology has been obtained for widely receiving and recognizing
It can.Therefore, coronary reconstruction has become the most frequently used and most important means for the treatment of coronary artery disease at present, has important
Clinical value and practical significance.
However, using DicLoss as the loss function of segmentation neural network, passing through in automation coronary artery reconstruction process
The machine learning of sample image, neural metwork training obtain coronary artery parted pattern, carry out coronary artery segmentation to data (such as CT images).
It unanimously cannot effectively assess, not only be increased when coronary artery is divided in abnormal conditions such as noise, the fractures that this process occurs
Cost, and the accuracy of coronary artery segmentation cannot also ensure.How effectively model to be assessed, reflection coronary artery can when dividing
The exception that can occur is a urgent problem to be solved.
Summary of the invention
The present invention provides a kind of coronary artery segmentation appraisal procedure and system, can be in automation coronary artery reconstruction process, reflection
Coronary artery predicts the abnormal conditions of dividing body, and the abnormal conditions occurred when dividing to coronary artery make effective assessment, increases coronary artery segmentation
Accuracy.
One aspect of the present invention provides a kind of embodiment of coronary artery segmentation appraisal procedure, comprising: to coronary artery predict dividing body into
Row central line pick-up obtains prediction center line；Mark center line is generated using label dividing body；Compare the prediction center line with
The mark center line, with abnormal position present in the determination prediction center line；According to identified abnormal position come really
The evaluation index of the fixed coronary artery segmentation.
It is described that central line pick-up is carried out to coronary artery prediction dividing body in an embodiment of the invention, it is predicted
Center line, comprising: largest connected body is chosen according to the coronary artery dividing body；The center line for extracting the largest connected body, by institute
The center line of the largest connected body extracted is as the prediction center line.
In an embodiment of the invention, the abnormal position is fracture position, according to identified abnormal position
To determine that the evaluation index of the coronary artery segmentation is to determine that the fracture of the coronary artery segmentation is commented according to identified fracture position
Estimate index.
In an embodiment of the invention, the fracture position determined by determines coronary artery segmentation
Fracture assessment index, comprising: determine each described fracture position to the length and the fracture for predicting center line starting point
The length of starting point of the endpoint of position corresponding label branch hub line in the mark center line to the mark center line
Fracture ratio；Calculate determined by it is all fracture ratios and value；Determine all fracture ratios and in value and the label
The ratio of the quantity for the label branch hub line for including in heart line, to obtain the fracture assessment index.
In an embodiment of the invention, the abnormal position is noise position, according to identified abnormal position
To determine that the evaluation index of the coronary artery segmentation is to determine that the noise of the coronary artery segmentation is commented according to identified noise position
Estimate index.
In an embodiment of the invention, the noise position determined by determines coronary artery segmentation
Noise evaluation index, comprising: determine each described noise position corresponding label branch hub in the mark center line
The endpoint of line to the mark center line starting point length and the noise position to it is described predict center line starting point length
Noise difference；Determine the noise product of each the noise difference and average noise rate；Wherein, the average noise rate is
The ratio of noise centerline length where the average length of noise and the noise position；Determine each described noise product with
The noise value of the prediction centerline length；Calculate determined by all noise values and value；Determine all noise values
And the label branch hub line that includes in value and the mark center line quantity ratio, referred to obtaining the noise evaluation
Mark.
In an embodiment of the invention, the abnormal position is false tendency position, according to identified exception
Position is to determine that the coronary artery is divided according to identified false tendency position come the evaluation index for determining the coronary artery segmentation
False tendency evaluation index.
In an embodiment of the invention, the falseness tendency position determined by determines the coronary artery point
The false tendency evaluation index cut, comprising: determine that each described false tendency position is corresponding in the mark center line
Mark length and the false tendency position to the prediction of the endpoint of branch hub line to the starting point of the mark center line
The false tendency difference of the length of center line starting point；Determine each false tendency difference and the prediction centerline length
False tendency ratio；All false tendency ratios determined by calculating and value；Determine all false tendency ratios and value
With the ratio of the quantity for the label branch hub line for including in the mark center line, referred to obtaining the false tendency assessment
Mark.
Another aspect of the present invention provides a kind of embodiment of coronary artery segmentation assessment system, comprising: prediction centerline cell is used
In carrying out central line pick-up to coronary artery prediction dividing body, prediction center line is obtained；Mark center line unit, for utilizing label point
It cuts body and generates mark center line；Center line comparing unit, for comparing the prediction center line and the mark center line, with true
Abnormal position present in the fixed prediction center line；Assessment unit, it is described for being determined according to identified abnormal position
The evaluation index of coronary artery segmentation.
In an embodiment of the invention, the prediction centerline cell includes: that connected component chooses subelement, is used for
Largest connected body is chosen according to the coronary artery dividing body；Central line pick-up subelement, for extracting in the largest connected body
Heart line, using the center line of the extracted largest connected body as prediction center line.
In an embodiment of the invention, the abnormal position is fracture position, and the assessment unit is used for basis
Identified fracture position determines the fracture assessment index of coronary artery segmentation.
In an embodiment of the invention, the assessment unit includes: fracture ratio calculation subelement, for determining
Each described fracture position is to the length of the prediction center line starting point and the fracture position in the mark center line
The fracture ratio of the length of starting point of the endpoint of corresponding label branch hub line to the mark center line；It is broken ratio summation
Subelement, for all fractures ratios determined by calculating and value；It is broken ratio and assesses subelement, for determining all fractures
The ratio of the quantity of label branch hub line that is ratio and including in value and the mark center line is commented with obtaining the fracture
Estimate index.
In an embodiment of the invention, the abnormal position is noise position, and the assessment unit is also used to root
The noise evaluation index of the coronary artery segmentation is determined according to identified noise position.
In an embodiment of the invention, the assessment unit further include: noise difference computation subunit, for true
The endpoint of each fixed described noise position corresponding label branch hub line in the mark center line is into the label
The noise difference of the length of the starting point of heart line and the length of the noise position to the prediction center line starting point；Noise product meter
Operator unit, for determining the noise product of each the noise difference and average noise rate；Wherein, the average noise rate
For the ratio of noise centerline length where the average length of noise and the noise position；Noise value computation subunit is used
In the noise value for determining each described noise product and the prediction centerline length；Noise value summation subelement, is used
In calculate determined by all noise values and value；Noise value assesses subelement, for determining the sum of all noise values
The ratio of the quantity for the label branch hub line for including in value and the mark center line, to obtain the noise evaluation index.
In an embodiment of the invention, the abnormal position is false tendency position, and the assessment unit is also used
In the false tendency evaluation index for determining the coronary artery segmentation according to identified false tendency position.
In an embodiment of the invention, the assessment unit further include: false tendency difference computation subunit is used
In determining the endpoint of each described false tendency position corresponding label branch hub line in the mark center line to institute
The falseness for stating the length and the length of the false tendency position to the prediction center line starting point of the starting point of mark center line is walked
Potential difference value；False tendency ratio calculation subelement, for determining each false tendency difference and the prediction center line
The false tendency ratio of length；False tendency ratio summation subelement, for calculating identified all false tendency ratios
And value；False tendency ratio assesses subelement, for determine all false tendency ratios and value in the mark center line
Including label branch hub line quantity ratio, to obtain the false tendency evaluation index.
The coronary artery segmentation appraisal procedure provided through the embodiment of the present invention, carries out center line to coronary artery prediction dividing body and mentions
It takes, obtains prediction center line；Mark center line is generated using label dividing body；To the prediction center line and the mark center
Line is compared, with abnormal position present in the determination prediction center line；According to identified abnormal position to determine
State the evaluation index of coronary artery segmentation.Divide appraisal procedure and system using coronary artery provided in an embodiment of the present invention, it can be automatic
Change in coronary artery reconstruction process, the abnormal conditions of reflection coronary artery prediction dividing body, the abnormal conditions occurred when dividing to coronary artery are made
Effectively assessment increases the accuracy of coronary artery segmentation.
It is to be appreciated that the teachings of the present invention does not need to realize whole beneficial effects recited above, but it is specific
Technical solution may be implemented specific technical effect, and other embodiments of the invention can also be realized and not mentioned above
Beneficial effect.
Detailed description of the invention
The following detailed description is read with reference to the accompanying drawings, above-mentioned and other mesh of this specification illustrative embodiments
, feature and will a little become prone to understand.In the accompanying drawings, if showing by way of example rather than limitation of the invention
Dry embodiment, in which:
In the accompanying drawings, identical or corresponding label indicates identical or corresponding part.
Fig. 1 shows a kind of coronary artery segmentation appraisal procedure flow diagram of the embodiment of the present invention；
Fig. 2 shows central line pick-up is carried out to coronary artery prediction dividing body in the embodiment of the present invention, prediction center line is obtained
Flow diagram；
Fig. 3 shows the calculation method flow diagram that evaluation index is broken in the embodiment of the present invention；
Fig. 4 shows the calculation method flow diagram of noise evaluation index in this clearly demarcated embodiment；
Fig. 5 shows the calculation method flow diagram of false tendency evaluation index in present invention implementation；
Fig. 6 shows a kind of coronary artery segmentation assessment system structural schematic diagram of the embodiment of the present invention；
Fig. 7 shows the structural schematic diagram that centerline cell is predicted in the embodiment of the present invention；
Fig. 8 shows the structural schematic diagram of assessment unit in the embodiment of the present invention.
Specific embodiment
To keep the purpose of the present invention, feature, advantage more obvious and understandable, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
It is only a part of the embodiment of the present invention, and not all embodiments.Based on the embodiments of the present invention, those skilled in the art are not having
Every other embodiment obtained under the premise of creative work is made, shall fall within the protection scope of the present invention.
The principle and spirit of the invention are described below with reference to several illustrative embodiments.It should be appreciated that providing this
A little embodiments are used for the purpose of making those skilled in the art can better understand that realizing the present invention in turn, and be not with any
Mode limits the scope of the invention.On the contrary, thesing embodiments are provided so that the present invention is more thorough and complete, and energy
It enough will fully convey the scope of the invention to those skilled in the art.
The technical solution of the present invention is further elaborated in the following with reference to the drawings and specific embodiments.
Fig. 1 shows a kind of coronary artery segmentation appraisal procedure schematic diagram of the embodiment of the present invention.
As shown in Figure 1, the coronary artery of the embodiment of the present invention divides appraisal procedure, comprising: step S101 predicts to divide to coronary artery
Body carries out central line pick-up, obtains prediction center line；Step S102 generates mark center line using label dividing body；Step
S103 compares the prediction center line and the mark center line, with abnormal position present in the determination prediction center line；
Step S104 determines the evaluation index of the coronary artery segmentation according to identified abnormal position.
Here, step S101 and step S102 and in no particular order order, can first carry out step S101, can also first carry out
Step S102 can be also completed at the same time, and obtained prediction center line and generated mark center line, they do not influence each other.
In a specific embodiment, coronary artery prediction dividing body is carried out in the embodiment of the present invention as shown in Figure 2
Central line pick-up obtains the flow diagram of prediction center line.In this embodiment, center is carried out to coronary artery prediction dividing body
Line drawing obtains prediction center line, comprising: step S1011 predicts that dividing body chooses largest connected body according to the coronary artery；Step
Rapid S1012 extracts the center line of the largest connected body, using the center line of the extracted largest connected body as in prediction
Heart line.
In the embodiment, by neural network method, sample image is trained to obtain a prediction model, then
A series of CT (CT Scan) image input prediction models coronarius are obtained into coronary artery prediction dividing body.Obtained hat
Arteries and veins predicts that dividing body is not one, from a maximum connected component is wherein chosen as largest connected body, extracts largest connected body
Center line, using the center line of extracted largest connected body as prediction center line.
The acquisition of prediction model can also be by being labeled to obtain to sample image according to the rule manually set.Input
The image of prediction model is also not limited to CT images, the image that can also be obtained using other Medical Imaging Technologies, such as: blood
The image that pipe photography/artery photography/angiography etc. is obtained using X-ray illuminating method, there are also angiocardiographies, mammogram
The obtained images of technologies such as art, positron emission tomoscan, the imaging of nuclear-magnetism engineering, the detection of medicine ultrasonic can be in this as
Original image.
The blood vessel framework of maximum connection body can be obtained, accordingly by general by carrying out skeletal extraction to largest connected body
Sharp nurse algorithm or kruskal algorithm generate maximum spanning tree or minimum spanning tree, by maximum spanning tree or minimum spanning tree point
Solution just obtains the branch hub line of largest connected body.
In step s 102, mark center line is generated using label dividing body.Here label dividing body is complete and accurate
Coronary artery dividing body, be trained in advance by neural network method.And obtain the mark center line of label dividing body
Method can be identical as the method for center line for extracting largest connected body.
Under normal conditions, when being split to coronary artery, it is possible to will appear the abnormal feelings such as fracture, noise or false tendency
Condition.Here, by that will predict that center line is compared with mark center line, it can be found that fracture present in prediction center line,
The abnormal positions such as noise or false tendency, and then determine that the assessment of the coronary artery segmentation refers to according to identified abnormal position
Mark, the abnormal conditions of reflection coronary artery prediction dividing body, the exception occurred when dividing to coronary artery make effective assessment.
In a specific embodiment, the abnormal position be fracture position, according to identified abnormal position come
The evaluation index for determining the coronary artery segmentation is the evaluation index that the coronary artery segmentation is determined according to identified fracture position.
Fig. 3 shows the calculation method flow diagram that evaluation index is broken in the embodiment of the present invention.In the implementable side
In formula, the fracture position determined by determines the fracture assessment index of coronary artery segmentation, comprising: step
S1041a determines each described fracture position to the length of the prediction center line starting point and the fracture position in the mark
Remember the fracture ratio of the length of starting point of the endpoint of corresponding label branch hub line in center line to the mark center line.Step
Rapid S1042a, calculate determined by it is all fracture ratios and value.Step S1043a, determine all fracture ratios and value and institute
The ratio of the quantity for the label branch hub line for including in mark center line is stated, to obtain the fracture assessment index.
Herein, it should be pointed out that either prediction center line or mark center line, center line starting point refer to lead
Artery issues the point of coronary artery, and endpoint refers to the tip of coronary branches.Fracture assessment index can be counted by following formula
It calculates:
Wherein, EMbrakeIndicate fracture assessment index；
Length_curr_midline is indicated in prediction center line in the presence of the fracture position of fracture to prediction center line starting point
Length；
Length_label_midline indicates abnormal position corresponding label branch hub line in heart line in the markers
Endpoint to mark center line starting point length；In the calculating of fracture assessment index, length_label_midline is indicated
Fracture position in the markers in the heart line starting point of the endpoint of corresponding label branch hub line to mark center line length；
Midline_num indicates the quantity for the label branch hub line for including in mark center line；
I is the number for the label branch hub line for including in mark center line；N is natural number, and the value of n is equal to midline_
The value of num.
The fracture assessment index EM of coronary artery segmentation is calculated by above-mentioned formulabrake, in fracture assessment index EMbrakeMore
Big explanation is less susceptible to breakage problem occur when being split using prediction model to coronary artery.In other words, as by above-mentioned
Formula calculates the fracture assessment index EM of coronary artery segmentationbrakeIt is the bigger the better.
In a specific embodiment, the abnormal position be noise position, according to identified abnormal position come
The evaluation index for determining the coronary artery segmentation is the noise evaluation that the coronary artery segmentation is determined according to identified noise position
Index.
Fig. 4 shows the calculation method flow diagram of noise evaluation index in this clearly demarcated embodiment.In the implementable side
In formula, the noise evaluation index of the coronary artery segmentation is determined according to identified noise position, comprising: step S1041b, really
The endpoint of each fixed described noise position corresponding label branch hub line in the mark center line is into the label
The noise difference of the length of the starting point of heart line and the length of the noise position to the prediction center line starting point.Step
S1042b determines the noise product of each the noise difference and average noise rate, wherein the average noise rate is noise
Average length and the noise position where noise centerline length ratio.Step S1043b is determined and is made an uproar described in each
The noise value of sound product and the prediction centerline length.Step S1044b calculates the sum of identified all noise values
Value.Step S1045b determines label branch hub line that is all noise values and including in value and the mark center line
The ratio of quantity, to obtain the noise evaluation index.
It should also be noted that in the calculating process of noise evaluation index, during either prediction center line still marks
Heart line, center line starting point refer to the point that aorta issues coronary artery, and endpoint refers to the tip of coronary branches.Noise evaluation refers to
Mark can be calculated by following formula:
Wherein, EMnoiseIndicate noise evaluation index；
Length_label_midline indicates abnormal position corresponding label branch hub line in heart line in the markers
Endpoint to mark center line starting point length；In the calculating of noise evaluation index, length_label_midline is indicated
Noise position in the markers in the heart line starting point of the endpoint of corresponding label branch hub line to mark center line length；
In the calculating of noise evaluation index, length_curr_pos indicates that there are the noises of noise in prediction center line
Length of the position to prediction center line starting point；
Midline_num indicates the quantity for the label branch hub line for including in mark center line；
I is the number for the label branch hub line for including in mark center line；N is natural number, and the value of n is equal to midline_
The value of num；
Ratio is average noise rate, indicates the average length and noise position place noise centerline length of noise
Ratio, formula can be passed through
It is calculated；
Herein, the length of noise center line where noise_length indicates the noise position；
C is constant, indicates the average length of noise or the acceptable length of noise, not a changeless numerical value,
Specifically the either empirical value that obtains according to statistical data of medical department can be determined according to practical situation by medical department.
The noise evaluation index EM of coronary artery segmentation is calculated by above-mentioned formulanoise, in noise evaluation index EMnoiseMore
Small explanation is less susceptible to noise problem occur when being split using prediction model to coronary artery.In other words, as by above-mentioned
Formula calculates the fracture assessment index EM of coronary artery segmentationnoiseIt is the smaller the better.
In a specific embodiment, the abnormal position is false tendency position, according to identified exception bits
Setting the evaluation index to determine the coronary artery segmentation is that the coronary artery segmentation is determined according to identified false tendency position
False tendency evaluation index.
Fig. 5 shows the calculation method flow diagram of false tendency evaluation index in present invention implementation.It is implementable at this
In mode, the false tendency evaluation index of the coronary artery segmentation is determined according to identified false tendency position, comprising: step
S1041c determines the endpoint of each the false tendency position corresponding label branch hub line in the mark center line
To the void of the length and the length of the false tendency position to the prediction center line starting point of the starting point of the mark center line
False tendency difference；Step S1042c determines that the falseness of each the false tendency difference and the prediction centerline length is walked
Gesture ratio.Step S1043c, calculate determined by all false tendency ratios and value.Step S1044c determines all falsenesses
The ratio of the quantity of label branch hub line that is tendency ratio and including in value and the mark center line, to obtain the void
False tendency evaluation index.
It should also be noted that either prediction center line is still marked in the calculating process of false tendency evaluation index
Remember center line, center line starting point refers to the point that aorta issues coronary artery, and endpoint refers to the tip of coronary branches.Falseness is walked
Gesture evaluation index can be calculated by following formula:
Wherein, EMmidlineIndicate noise evaluation index；
Length_label_midline indicates abnormal position corresponding label branch hub line in heart line in the markers
Endpoint to mark center line starting point length；In the calculating of false tendency evaluation index, length_label_midline
Indicate false tendency position in the markers in heart line corresponding label branch hub line endpoint to mark center line starting point
Length；
In the calculating of false tendency evaluation index, length_curr_pos indicates that there are falsenesses to walk in prediction center line
Length of the false tendency position of gesture to the starting point for predicting center line；
Midline_num indicates the quantity for the label branch hub line for including in mark center line；
I is the number for the label branch hub line for including in mark center line；N is natural number, and the value of n is equal to midline_
The value of num.
The false tendency evaluation index EM of coronary artery segmentation is calculated by above-mentioned formulamidline, in noise evaluation index
EMmidlineBigger explanation is less susceptible to noise problem occur when being split using prediction model to coronary artery.In other words, as
The fracture assessment index EM of coronary artery segmentation is calculated by above-mentioned formulamidlineIt is the bigger the better.
The fracture assessment index that obtains by the above process, noise evaluation index and false tendency evaluation index, can be with
It can individually be measured in coronary artery cutting procedure, can also be weighted and averaged or even be multiplied, with may be simultaneously when coronary artery is divided
When there are various problems, reflect the prediction situation of model, the exception occurred when dividing to coronary artery makes effective assessment.
Fig. 6 shows a kind of coronary artery segmentation assessment system structural schematic diagram of the embodiment of the present invention.
As shown in fig. 6, the coronary artery of the embodiment of the present invention divides assessment system, comprising: prediction centerline cell 101, label
Centerline cell 102, center line comparing unit 103 and assessment unit 104.
Wherein, it predicts centerline cell 101, for carrying out central line pick-up to coronary artery prediction dividing body, obtains in prediction
Heart line.Mark center line unit 102, for generating mark center line using label dividing body.Center line comparing unit 103 is used
In comparing the prediction center line and the mark center line, with abnormal position present in the determination prediction center line.It comments
Unit 104 is estimated, for determining the evaluation index of the coronary artery segmentation according to identified abnormal position.
Fig. 7 shows the structural schematic diagram that centerline cell is predicted in the embodiment of the present invention.In a specific implementable side
In formula, prediction centerline cell 101 includes: that connected component chooses subelement 1011 and central line pick-up subelement 1012.Wherein, even
Entire body chooses subelement 1011, for choosing largest connected body according to the coronary artery dividing body；Central line pick-up subelement 1012,
For extracting the center line of the largest connected body, using the center line of the extracted largest connected body as prediction center line.
When being split to coronary artery, it is possible to will appear the abnormal conditions such as fracture, noise or false tendency.Here, lead to
It crosses and prediction center line is compared with mark center line, it can be found that fracture, noise or falseness present in prediction center line
The abnormal positions such as tendency, and then determine according to identified abnormal position the evaluation index of the coronary artery segmentation, reflect coronary artery
The abnormal conditions for predicting dividing body, the exception occurred when dividing to coronary artery make effective assessment.
In a specific embodiment, when the abnormal position is fracture position, the assessment unit is used for basis
Identified fracture position determines the fracture assessment index of coronary artery segmentation.When the abnormal position is noise position, institute
State the noise evaluation index that assessment unit is also used to determine the coronary artery segmentation according to identified noise position.The exception bits
When being set to false tendency position, the assessment unit is also used to determine the coronary artery point according to identified false tendency position
The false tendency evaluation index cut.
Fig. 8 shows the structural schematic diagram of assessment unit in the embodiment of the present invention.
In a specific embodiment, the abnormal position is fracture position, and the assessment unit 104 is used for basis
Identified fracture position determines the fracture assessment index of coronary artery segmentation.By breaking present in discovery prediction center line
Position is split, the fracture assessment index of the coronary artery segmentation, reflection coronary artery prediction segmentation are determined according to identified fracture position
The case where crack conditions of body, the fracture occurred when dividing to coronary artery, makes effective assessment.
In the embodiment, assessment unit 104 includes: fracture ratio calculation subelement 1041a, fracture ratio summation
Subelement 1042a and fracture ratio assess subelement 1043a.
Wherein, it is broken ratio calculation subelement 1041a, for determining each described fracture position to the pre- measured center
The length of line starting point is with the endpoint of the fracture position corresponding label branch hub line in the mark center line described in
The fracture ratio of the length of the starting point of mark center line.It is broken ratio summation subelement 1042a, it is identified all for calculating
Be broken ratio and value.Be broken ratio and assess subelement 1043a, for determine all fracture ratios and in value and the label
The ratio of the quantity for the label branch hub line for including in heart line, to obtain the fracture assessment index.
In a specific embodiment, the abnormal position is noise position, and the assessment unit 104 is also used to root
The noise evaluation index of the coronary artery segmentation is determined according to identified noise position.By making an uproar present in discovery prediction center line
Sound position determines the noise evaluation index of the coronary artery segmentation, reflection coronary artery prediction segmentation according to identified noise position
The case where noise situations of body, the noise occurred when dividing to coronary artery, makes effective assessment.
In the embodiment, assessment unit 104 includes: noise difference computation subunit 1041b, the calculating of noise product
Subelement 1042b, noise value computation subunit 1043b, noise value summation subelement 1044b and noise value assessment are single
First 1045b.
Wherein, noise difference computation subunit 1041b, for determining each described noise position in the mark center
The length of starting point of the endpoint of corresponding label branch hub line to the mark center line and the noise position to institute in line
State the noise difference of the length of prediction center line starting point；Noise product computation subunit 1042b makes an uproar described in each for determining
The noise product of sound difference and average noise rate；Wherein, the average noise rate is the average length and the noise bits of noise
The ratio of noise centerline length where setting；Noise value computation subunit 1043b, for determining each described noise product
With the noise value of the prediction centerline length；Noise value is summed subelement 1044b, is made an uproar for all determined by calculating
Acoustic ratio value and value；Noise value assesses subelement 1045b, for determine all noise values and value and the mark center
The ratio of the quantity for the label branch hub line for including in line, to obtain the noise evaluation index.
In a specific embodiment, the abnormal position is false tendency position, and the assessment unit 104 is also used
In the false tendency evaluation index for determining the coronary artery segmentation according to identified false tendency position.By in discovery prediction
Falseness tendency position present in heart line determines the false tendency of the coronary artery segmentation according to identified false tendency position
The case where evaluation index, reflection coronary artery predict the false tendency situation of dividing body, the false tendency occurred when to coronary artery segmentation work
Effectively assessment out.
In the embodiment, assessment unit 104 includes: false tendency difference computation subunit 1041c, false tendency
Ratio calculation subelement 1042c, false tendency ratio summation subelement 4043c and false tendency ratio assess subelement 1044c.
Wherein, false tendency difference computation subunit 1041c, for determining each described false tendency position described
The length and the falseness of starting point of the endpoint of corresponding label branch hub line to the mark center line in mark center line
False tendency difference of the tendency position to the length for predicting center line starting point；False tendency ratio calculation subelement 1042c,
For determining the false tendency ratio of each the false tendency difference and the prediction centerline length；False tendency ratio
Sum subelement 1043c, for all false tendency ratios determined by calculating and value；False tendency ratio assesses subelement
1044c, for determining label branch hub line that is all false tendency ratios and including in value and the mark center line
The ratio of quantity, to obtain the false tendency evaluation index.
Assessment unit 104 in the present invention is not limited to fracture, noise or the falseness being likely to occur when dividing to coronary artery
Tendency problem is assessed, and can also be assessed other any the problem of may relate to, only be needed to assessment unit at this time
104 carry out corresponding extension, to obtain corresponding evaluation index.
A kind of coronary artery segmentation assessment system of the embodiment of the present invention is run using the principle of above-mentioned coronary artery segmentation appraisal procedure,
This is no longer going to repeat them.The coronary artery segmentation appraisal procedure and system provided through the embodiment of the present invention, predicts to divide to coronary artery
Body carries out central line pick-up, obtains prediction center line；Mark center line is generated using label dividing body；To the prediction center line
It is compared with the mark center line, with abnormal position present in the determination prediction center line；According to identified different
Normal position determines the evaluation index of the coronary artery segmentation, can be in automation coronary artery reconstruction process, reflection coronary artery prediction point
The case where cutting noise, the fracture, false tendency of body, the exception occurred when dividing to coronary artery makes effective assessment.
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.Moreover, particular features, structures, materials, or characteristics described
It may be combined in any suitable manner in any one or more of the embodiments or examples.In addition, without conflicting with each other, this
The technical staff in field can be by the spy of different embodiments or examples described in this specification and different embodiments or examples
Sign is combined.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic." first " is defined as a result, the feature of " second " can be expressed or hidden
It include at least one this feature containing ground.In the description of the present invention, the meaning of " plurality " is two or more, unless otherwise
Clear specific restriction.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, protection scope of the present invention should be based on the protection scope of the described claims.
Claims (10)
1. a kind of coronary artery divides appraisal procedure characterized by comprising
Central line pick-up is carried out to coronary artery prediction dividing body, obtains prediction center line；
Mark center line is generated using label dividing body；
The prediction center line and the mark center line are compared, with abnormal position present in the determination prediction center line；
The evaluation index of the coronary artery segmentation is determined according to identified abnormal position.
2. coronary artery according to claim 1 divides appraisal procedure, which is characterized in that described to be carried out to coronary artery prediction dividing body
Central line pick-up obtains prediction center line, comprising:
Predict that dividing body chooses largest connected body according to the coronary artery；
The center line for extracting the largest connected body, using the center line of the extracted largest connected body as pre- measured center
Line.
3. coronary artery according to claim 1 divides appraisal procedure, which is characterized in that the abnormal position is fracture position；
The abnormal position determined by is according to identified cleavage come the evaluation index for determining the coronary artery segmentation
Set the fracture assessment index to determine the coronary artery segmentation.
4. coronary artery according to claim 3 divides appraisal procedure, which is characterized in that the fracture position according to determined by
To determine the fracture assessment index of the coronary artery segmentation, comprising:
Determine each described fracture position to the length of the prediction center line starting point and the fracture position in the label
The fracture ratio of the length of starting point of the endpoint of corresponding label branch hub line to the mark center line in center line；
Calculate determined by it is all fracture ratios and value；
Determine the ratio of the quantity of label branch hub line that is all fracture ratios and including in value and the mark center line,
To obtain the fracture assessment index.
5. coronary artery according to claim 1 divides appraisal procedure, which is characterized in that the abnormal position is noise position；
The abnormal position determined by is according to identified noise bits come the evaluation index for determining the coronary artery segmentation
Set the noise evaluation index to determine the coronary artery segmentation.
6. coronary artery according to claim 5 divides appraisal procedure, which is characterized in that the noise position according to determined by
To determine the noise evaluation index of the coronary artery segmentation, comprising:
Determine the endpoint of each described noise position corresponding label branch hub line in the mark center line described in
The noise difference of the length of the starting point of mark center line and the length of the noise position to the prediction center line starting point；
Determine the noise product of each the noise difference and average noise rate；Wherein, the average noise rate is noise
The ratio of noise centerline length where average length and the noise position；
Determine the noise value of each described noise product and the prediction centerline length；
Calculate determined by all noise values and value；
Determine the ratio of the quantity of label branch hub line that is all noise values and including in value and the mark center line,
To obtain the noise evaluation index.
7. coronary artery according to claim 1 divides appraisal procedure, which is characterized in that the abnormal position is false tendency position
It sets；
The abnormal position determined by is to be walked according to identified falseness come the evaluation index for determining the coronary artery segmentation
Gesture position determines the false tendency evaluation index of coronary artery segmentation.
8. coronary artery according to claim 7 divides appraisal procedure, which is characterized in that the falseness tendency according to determined by
Position determines the false tendency evaluation index of coronary artery segmentation, comprising:
Determine that each described false tendency position endpoint of corresponding label branch hub line in the mark center line arrives
The falseness of the length of the starting point of the mark center line and the length of the false tendency position to the prediction center line starting point
Tendency difference；
Determine the false tendency ratio of each the false tendency difference and the prediction centerline length；
All false tendency ratios determined by calculating and value；
Determine the quantity of label branch hub line that is all false tendency ratios and including in value and the mark center line
Ratio, to obtain the false tendency evaluation index.
9. a kind of coronary artery divides assessment system characterized by comprising
It predicts centerline cell, for carrying out central line pick-up to coronary artery prediction dividing body, obtains prediction center line；
Mark center line unit, for generating mark center line using label dividing body；
Center line comparing unit, for comparing the prediction center line and the mark center line, with the determination pre- measured center
Abnormal position present in line；
Assessment unit, for determining the evaluation index of the coronary artery segmentation according to identified abnormal position.
10. coronary artery according to claim 9 divides assessment system, which is characterized in that the prediction centerline cell includes:
Connected component chooses subelement, for choosing largest connected body according to the coronary artery dividing body；
Central line pick-up subelement, for extracting the center line of the largest connected body, by the extracted largest connected body
Center line is as prediction center line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811162031.1A CN109410267B (en) | 2018-09-30 | 2018-09-30 | Coronary artery segmentation evaluation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811162031.1A CN109410267B (en) | 2018-09-30 | 2018-09-30 | Coronary artery segmentation evaluation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109410267A true CN109410267A (en) | 2019-03-01 |
| CN109410267B CN109410267B (en) | 2020-02-07 |
Family
ID=65466687
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811162031.1A Active CN109410267B (en) | 2018-09-30 | 2018-09-30 | Coronary artery segmentation evaluation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109410267B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110021016A (en) * | 2019-04-01 | 2019-07-16 | 数坤（北京）网络科技有限公司 | A kind of calcification detection method |
| CN111312374A (en) * | 2020-01-21 | 2020-06-19 | 上海联影智能医疗科技有限公司 | Medical image processing method, device, storage medium and computer equipment |
| CN113222964A (en) * | 2021-05-27 | 2021-08-06 | 推想医疗科技股份有限公司 | Method and device for generating coronary artery central line extraction model |
| CN110310256B (en) * | 2019-05-30 | 2021-09-21 | 上海联影智能医疗科技有限公司 | Coronary stenosis detection method, coronary stenosis detection device, computer equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104867147A (en) * | 2015-05-21 | 2015-08-26 | 北京工业大学 | SYNTAX automatic scoring method based on coronary angiogram image segmentation |
| CN106108925A (en) * | 2015-05-04 | 2016-11-16 | 西门子保健有限责任公司 | In medical image, Whole Body Bone Scanning removes the method and system with visualization of blood vessels |
| CN108198213A (en) * | 2018-02-01 | 2018-06-22 | 上海联影医疗科技有限公司 | It is a kind of mark tubular structure method and device, tubular structure processing terminal |
-
2018
- 2018-09-30 CN CN201811162031.1A patent/CN109410267B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106108925A (en) * | 2015-05-04 | 2016-11-16 | 西门子保健有限责任公司 | In medical image, Whole Body Bone Scanning removes the method and system with visualization of blood vessels |
| CN104867147A (en) * | 2015-05-21 | 2015-08-26 | 北京工业大学 | SYNTAX automatic scoring method based on coronary angiogram image segmentation |
| CN108198213A (en) * | 2018-02-01 | 2018-06-22 | 上海联影医疗科技有限公司 | It is a kind of mark tubular structure method and device, tubular structure processing terminal |
Non-Patent Citations (2)
| Title |
|---|
| MICHIEL SCHAAP: "Standardized evaluation methodology and reference database for evaluating coronary artery centerline extraction algorithms", 《MEDICAL IMAGE ANALYSIS》 * |
| 赵洁: "双源CT冠状动脉三维分割算法的研究", 《生物医学工程研究》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110021016A (en) * | 2019-04-01 | 2019-07-16 | 数坤（北京）网络科技有限公司 | A kind of calcification detection method |
| CN110310256B (en) * | 2019-05-30 | 2021-09-21 | 上海联影智能医疗科技有限公司 | Coronary stenosis detection method, coronary stenosis detection device, computer equipment and storage medium |
| CN111312374A (en) * | 2020-01-21 | 2020-06-19 | 上海联影智能医疗科技有限公司 | Medical image processing method, device, storage medium and computer equipment |
| CN111312374B (en) * | 2020-01-21 | 2024-03-22 | 上海联影智能医疗科技有限公司 | Medical image processing method, medical image processing device, storage medium and computer equipment |
| CN113222964A (en) * | 2021-05-27 | 2021-08-06 | 推想医疗科技股份有限公司 | Method and device for generating coronary artery central line extraction model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109410267B (en) | 2020-02-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10706545B2 (en) | Systems and methods for analysis of anatomical images | |
| CN109410267A (en) | A kind of coronary artery segmentation appraisal procedure and system | |
| KR102491988B1 (en) | Methods and systems for using quantitative imaging | |
| CN109069014B (en) | System and method for estimating healthy lumen diameter and stenosis quantification in coronary arteries | |
| US10349910B2 (en) | Method and apparatus for assessing blood vessel stenosis | |
| US10748451B2 (en) | Methods and systems for generating fluid simulation models | |
| EP3567525A1 (en) | Systems and methods for analysis of anatomical images each captured at a unique orientation | |
| CN110866914A (en) | Method, system, device and medium for evaluating cerebral aneurysm hemodynamic index | |
| CN106037710A (en) | Synthetic data-driven hemodynamic determination in medical imaging | |
| US20080071160A1 (en) | Displaying A Tracheobronchial Tree | |
| CN105877767A (en) | Coronary artery image angiography method and device | |
| CN108491770A (en) | A kind of data processing method based on fracture image | |
| EP4002387A1 (en) | Cad device and method for analysing medical images | |
| Affane et al. | Robust deep 3-D architectures based on vascular patterns for liver vessel segmentation | |
| Pepe et al. | Automated cross-sectional view selection in CT angiography of aortic dissections with uncertainty awareness and retrospective clinical annotations | |
| US20230260133A1 (en) | Methods for acquiring aorta based on deep learning and storage media | |
| Klein et al. | Automatic segmentation of the wire frame of stent grafts from CT data | |
| Kiraly et al. | Analysis of arterial subtrees affected by pulmonary emboli | |
| Aswathy et al. | CAD systems for automatic detection and classification of COVID-19 in nano CT lung image by using machine learning technique | |
| Jabbar et al. | Deep learning based classification of wrist cracks from X-ray imaging | |
| KR20240041355A (en) | Coronary artery disease analysis system | |
| Al-Rifaie et al. | Deploying swarm intelligence in medical imaging | |
| EP4113434A1 (en) | Generation of plaque information | |
| Huang et al. | A Dual-Functional System for the Classification and Diameter Measurement of Aortic Dissections Using CTA Volumes via Deep Learning | |
| Peng et al. | Aortic dissection diagnosis based on sequence information and deep learning |
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 | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 100102 No. 501 No. 12, 5th floor, No. 6, Wangjing Dongyuan District 4, Chaoyang District, Beijing Patentee after: Shukun (Beijing) Network Technology Co.,Ltd. Address before: 100102 No. 501 No. 12, 5th floor, No. 6, Wangjing Dongyuan District 4, Chaoyang District, Beijing Patentee before: SHUKUN (BEIJING) NETWORK TECHNOLOGY Co.,Ltd. |
|
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 100102 No. 501 No. 12, 5th floor, No. 6, Wangjing Dongyuan District 4, Chaoyang District, Beijing Patentee after: Shukun Technology Co.,Ltd. Address before: 100102 No. 501 No. 12, 5th floor, No. 6, Wangjing Dongyuan District 4, Chaoyang District, Beijing Patentee before: Shukun (Beijing) Network Technology Co.,Ltd. |