CN104715484B - Automatic tumor imaging region segmentation method based on improved level set - Google Patents
Automatic tumor imaging region segmentation method based on improved level set Download PDFInfo
- Publication number
- CN104715484B CN104715484B CN201510124586.7A CN201510124586A CN104715484B CN 104715484 B CN104715484 B CN 104715484B CN 201510124586 A CN201510124586 A CN 201510124586A CN 104715484 B CN104715484 B CN 104715484B
- Authority
- CN
- China
- Prior art keywords
- region
- pet image
- split
- tumor
- initial zero
- 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.)
- Active
Links
- 206010028980 Neoplasm Diseases 0.000 title claims abstract description 79
- 238000000034 method Methods 0.000 title claims abstract description 36
- 230000011218 segmentation Effects 0.000 title claims abstract description 27
- 238000003384 imaging method Methods 0.000 title claims abstract description 19
- 230000003902 lesion Effects 0.000 claims abstract description 46
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000009499 grossing Methods 0.000 claims abstract description 6
- 238000010276 construction Methods 0.000 claims abstract description 3
- 238000001914 filtration Methods 0.000 claims description 7
- 230000003628 erosive effect Effects 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 5
- 230000037396 body weight Effects 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 239000003795 chemical substances by application Substances 0.000 claims description 2
- 230000010339 dilation Effects 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 239000012141 concentrate Substances 0.000 claims 1
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 3
- 238000011156 evaluation Methods 0.000 abstract description 3
- 238000001356 surgical procedure Methods 0.000 abstract description 3
- 201000011510 cancer Diseases 0.000 description 7
- 208000019065 cervical carcinoma Diseases 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 206010008342 Cervix carcinoma Diseases 0.000 description 3
- 208000006105 Uterine Cervical Neoplasms Diseases 0.000 description 3
- 201000010881 cervical cancer Diseases 0.000 description 3
- 238000004445 quantitative analysis Methods 0.000 description 3
- 230000002285 radioactive effect Effects 0.000 description 3
- 239000000243 solution Substances 0.000 description 3
- 208000024719 uterine cervix neoplasm Diseases 0.000 description 3
- 206010006187 Breast cancer Diseases 0.000 description 2
- 208000026310 Breast neoplasm Diseases 0.000 description 2
- 206010058467 Lung neoplasm malignant Diseases 0.000 description 2
- 238000002591 computed tomography Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000002347 injection Methods 0.000 description 2
- 239000007924 injection Substances 0.000 description 2
- 201000005202 lung cancer Diseases 0.000 description 2
- 208000020816 lung neoplasm Diseases 0.000 description 2
- 238000002600 positron emission tomography Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 201000001275 rectum cancer Diseases 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 210000004999 sex organ Anatomy 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- ZJNLYGOUHDJHMG-UHFFFAOYSA-N 1-n,4-n-bis(5-methylhexan-2-yl)benzene-1,4-diamine Chemical compound CC(C)CCC(C)NC1=CC=C(NC(C)CCC(C)C)C=C1 ZJNLYGOUHDJHMG-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000010429 evolutionary process Effects 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- 208000014829 head and neck neoplasm Diseases 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 231100000225 lethality Toxicity 0.000 description 1
- 230000003211 malignant effect Effects 0.000 description 1
- 208000026037 malignant tumor of neck Diseases 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 208000025440 neoplasm of neck Diseases 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000008506 pathogenesis Effects 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 210000004872 soft tissue Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001225 therapeutic effect Effects 0.000 description 1
- 210000001519 tissue Anatomy 0.000 description 1
- 210000002700 urine Anatomy 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Abstract
The present invention provides a kind of automatic tumor imaging region segmentation method based on improved level set, including：Obtain the original PET image to be split comprising lesion region and pre-processed and positioned so that it is determined that pretreated lesion region PET image to be split；According to the CT images of lesion region and the pretreated lesion region PET image construction hypergraph to be split, so as to primarily determine that the rough tumor region in PET image is initial zero level collection；Improved Level Set Method is performed so that it is determined that tumor region to the initial zero level collection；Edge-smoothing processing is performed to the tumor region according to morphology operations.The method of the invention, which can be realized, fast and accurately splits tumor region, so that assisted surgery doctor carries out diagnoses and treatment and curative effect evaluation.
Description
Technical field
The present invention relates to image processing techniques, more particularly to a kind of automatic tumor imaging region based on improved level set
Dividing method.
Background technology
Cervical carcinoma is one of most common three big malignant tumour of female sex organ, is crisis women life and influences life
One of major malignant tumor of bioplasm amount, occupy female sex organ malignant tumour first.According to World Health Organization subordinate
International cancer research institution (The International Agency for Research on Cancer) 3 days 2 months it is in place
2014 worlds cancer report that general headquarters in Lyons, France deliver, 012 year new cases in the whole world of cervical cancer 2 reach more than 50 ten thousand
Example, breast cancer, the carcinoma of the rectum, lung cancer are only second in female malignant, ranked fourth position, death caused by same period cervical carcinoma
More than more than 260,000 people, lethality is only second to breast cancer, lung cancer, the carcinoma of the rectum, occupies the female cancer death rate the 4th.In less-developed state
In family women, cervical carcinoma is most common cancer.In recent years, the incidence of the cervical carcinoma of young woman, which has, increases trend, becomes
One of three big major cancers that young woman is susceptible to suffer from.China is cervical cancer pathogenesis and dead big country, and morbidity and mortality are equal
Account for 1/3rd of the world.Therefore it is particularly significant for the Accurate Diagnosis of cervical cancer patient.Positron emission tomography
Scan (positron emission tomography, PET) and CT scan (Computed
Tomography, CT) molecular image means are used as, it is the current common detection means in clinical tumor field, utilizes PET/CT pairs
Tumour, which carries out quantitative analysis, can provide accurate diagnostic message for clinic and aid in formulating therapeutic scheme.
The index of current clinically most common quantitative analysis be standard uptake value (standard uptake value,
SUV), SUV be equal to lesion radioactive concentration (kBq/ml) divided by injection dosage (MBq) again divided by weight (kg), secondly tumour
Volume, i.e. MTV values, are also proved to the recurrence and assessment prognosis that can predict tumour.But these quantitative targets all rely on it is swollen
The accurate of knurl region is delineated.In addition in the Radiation treatment plans for cervical carcinoma, also rely on the accurate of target area and delineate.Consider
The poorly efficient and higher subjectivity that work point in one's hands is cut, the segmentation of automatic accurate Cervical Tumor is very necessary.But with its
He compares tumour, and Cervical Tumor region is delineated then in face of more challenges：On the one hand, due to the decay of tumour and uterine neck essence
Coefficient is identical, therefore is difficult to accurately differentiate on CT images；On the other hand, due to the position of bladder and uterine neck very close to, and
The radioactive activity of urine in bladder is more than or is approximately equal to the radioactive activity of tumour, therefore also is difficult in PET image
Automatically extracted.
The content of the invention
The present invention provides a kind of automatic tumor imaging region segmentation method based on improved level set, for solving uterine neck
The problem of being difficult to automatic distinguishing tumor region and bladder area in tumor segmentation, so that user can fast and accurately split palace
Neck cancer tumour carries out diagnoses and treatment and curative effect evaluation so as to assisted surgery doctor.
Automatic tumor imaging region segmentation method of the invention based on improved level set includes：
Obtain the original PET image to be split comprising lesion region and pre-processed and positioned so that it is determined that pre-processing
Lesion region PET image to be split afterwards；
Hypergraph is constructed according to the CT images of lesion region and the pretreated lesion region PET image to be split, from
And it is initial zero level collection to primarily determine that the rough tumor region in the PET image；
Improved Level Set Method is performed so that it is determined that tumor region to the initial zero level collection；
Edge-smoothing processing is performed to the tumor region according to morphology operations.
Beneficial effects of the present invention are：
The present invention proposes a kind of automatic tumor imaging region segmentation method based on improved level set, solves uterine neck
The problem of being difficult to automatic distinguishing tumor region and bladder area in tumor segmentation, user is set fast and accurately to split uterine neck
Tumor carries out diagnoses and treatment and curative effect evaluation so as to assisted surgery doctor, and the method for the invention has speed fast, and precision is high,
The advantages of strong robustness, test result indicates that, this technology accurately can automatically delineate Cervical Tumor, realize tumour and bladder
Automatic distinguishing, has great practical value in clinical diagnosis and treatment.
Brief description of the drawings
Fig. 1 is the flow chart of the automatic tumor imaging region segmentation method of the invention based on improved level set；
Fig. 2 is that positioning is to be split described in the automatic tumor imaging region segmentation method of the invention based on improved level set
The schematic diagram of lesion region；
Fig. 3 is improved level described in the automatic tumor imaging region segmentation method of the invention based on improved level set
Schematic diagram during diversity method iteration；
Fig. 4 is 3 of the application automatic tumor imaging region segmentation method of the invention based on improved level set typical
The schematic diagram of the segmentation result of tumor region and goldstandard in cervical carcinoma data；
Fig. 5 is that the segmentation result of the automatic tumor imaging region segmentation method of the invention based on improved level set and gold are marked
The accurate uniformity schematic diagram when carrying out quantitative analysis.
Embodiment
Fig. 1 is the flow chart of the automatic tumor imaging region segmentation method of the invention based on improved level set, such as Fig. 1 institutes
Show, the automatic tumor imaging region segmentation method of the invention based on improved level set includes：
S1, obtain the original PET image to be split comprising lesion region and pre-processed and positioned so that it is determined that pre-
Lesion region PET image to be split after processing；
Preferably, it is described obtain comprising lesion region original PET image to be split and pre-processed and positioned from
And determining pretreated lesion region PET image to be split includes：
To the voxel gray values in the original PET image to be split comprising lesion region divided by the development injected
Agent dose and patient body weight are to be converted to SUV values, then carry out gaussian filtering and up-sampling, so that point of PET image to be split
Resolution is identical with CT images, is finally positioned according to the SUV values and determines pretreated lesion region PET image to be split.
Preferably, it is described to be positioned according to the SUV values and determine pretreated lesion region PET image bag to be split
Include：
Preprocessing process, including：
By the gray value of each voxel of PET image by divided by injection 18F-FDG dosage and patient body weight be converted into
SUV values, then gaussian filtering and up-sampling are carried out, make its resolution ratio identical with CT images.Meanwhile CT images also carry out identical height
This filtering；
And position fixing process, including：
Calculate SUV peak values (SUVpeak, the expression each cut into slices in the original PET image to be split comprising lesion region
The average value of the SUV values of each voxel in 26 neighborhoods of corresponding SUVmax namely the voxel of maximum SUV values), choose more than foot most
Corresponding section is cut into slices as lesion region place so that it is determined that pre- place between the two neighboring minimum SUV values of big SUV peak values
Lesion region PET image to be split after reason.
S2, the CT images according to lesion region and the pretreated lesion region PET image construction hypergraph to be split,
So as to primarily determine that the rough tumor region in the PET image is initial zero level collection；
Preferably, the CT images according to lesion region and the pretreated lesion region PET image to be split
Hypergraph is constructed, so as to primarily determine that the rough tumor region in the PET image includes for initial zero level collection：
CT image normalizations to the pretreated lesion region PET image to be split with lesion region and after being multiplied
Result structure hypergraph, recycle fuzzy C-means clustering, morphological erosion and generic threshold value method to primarily determine that the PET figures
Rough tumor region as in is initial zero level collection, including：
Hypergraph is first built, including：
Each voxel of hypergraph is made of three features, is respectively：PET image corresponds to the normalized SUV values of voxel (i.e.
SUV/SUVmax), CT images correspond to the normalized HU values (HU/HUmax, HU represent the CT values of each voxel) of voxel, and
Their product.According to root tissue specificity, hypergraph can be divided into four parts：(a) representative tumour all bigger SUV and HU；(b)
SUV high but low HU representative bladder；(c) SUV is low but the representative of HU high other soft tissues；(d) SUV and HU is than relatively low generation
Table background.
Fuzzy C-means clustering is recycled to be divided into 4 classes to hypergraph, wherein the representative tumour that three features are all bigger, profit
The bladder wall of tumor region may be divided into by mistake by being eroded with the method for morphological erosion, and then this region is commonly used using clinical
40% SUVmax be threshold value, obtain rough tumor region.
S3, perform improved Level Set Method so that it is determined that tumor region to the initial zero level collection；
Preferably, it is described that improved Level Set Method is performed to initial zero level collection so that it is determined that tumor region includes：
The gradient fields information of pretreated lesion region PET image to be split is added to the initial zero level collection
In, new EVOLUTION EQUATION is built, to the rough tumor region in the PET image with the resolution ratio with the original PET image
Identical resolution ratio performs down-sampling and obtains initial zero level collection；
Successive ignition is carried out in the EVOLUTION EQUATION according to finite difference calculus to the initial zero level collection to determine finally
Split the tumor region completed；
Preferably, the improved Level Set Method includes：
The dark characteristic of middle bright limb is presented in view of the tumour after gaussian filtering and bladder, so tumour and bladder
The gradient field direction at edge is on the contrary, the new level set movements equation EVOLUTION EQUATION as described below that therefore can be built is：
Wherein, IσIt is pretreated lesion region PET image to be split, (variance of Gaussian kernel is σ), φ is initial zero
Level set or initial zero level collection carry out the function of the level set after iteration several times,<*>Represent the IσWith the gradient of φ to
Angle between amount, | * | represent the vectorial amplitude, c1And c2Initial zero level collection (or initial zero level is represented respectively
Collection) inside and outside voxel average gray value, i.e.,:
Wherein, Ω represents image-region,
δ functions are obtained by following smooth function δ ε approximations：
Φ is to adjust initial zero level collection or initial zero level collection to carry out the function of the level set after iteration several times and make
Serialization, need to carry out following gaussian filtering to level set function after each iteration：
φ=Gσ* φ,
Wherein, GσIt is the Gaussian kernel that variance is σ, * represents convolution algorithm, and initial function φ 0 is defined as follows：
Wherein, c0 is normal number, and R0 is to be down-sampled to the rough tumor region that original PET image resolution ratio obtains；
In view of discrete can be turned to for 3-D view, level set function φ (x, y, z, t)Wherein (i, j, k)
For space coordinate, n is time coordinate, then EVOLUTION EQUATION discrete can turn to：
Wherein L is that then evolutionary process can be iterated as the following formula on the right of the equal sign of EVOLUTION EQUATION：
Fig. 3 is the schematic diagram of iterative process.
S4, according to morphology operations to the tumor region perform edge-smoothing processing.
Preferably, it is described that tumor region execution edge-smoothing processing is included according to morphology operations：
Morphology opening operation and closed operation are performed to the tumor region with chondritic element to eliminate edge protuberance
And filling cavity；
Preferably, wherein specific morphology operations refer to the tumor region obtained using chondritic element to S103
Edge protuberance is eliminated using morphology opening operation and closed operation, radius is that the chondritic element s (x, y, z) of r is as follows：
It is described that morphology opening operation and closed operation are performed to the tumor region to eliminate edge with chondritic element
Raised and filling cavity includes：Carry out opening operation and closed operation processing respectively using formula (6) and (7)：
Opening operation processing：
Closed operation is handled：
Wherein,Represent dilation operation symbol,Represent erosion operation symbol, M represents the binary map of the tumor region
Picture.
It should be noted that for verification effectiveness of the invention and practicality, we carry out on clinical PET/CT images
Experiment, goldstandard by two expert's manual segmentation results average value.
By largely testing, (i.e. Dice similarity coefficient, weigh segmentation to Dice likeness coefficients
As a result the Duplication between goldstandard) for 91.80 ± 2.46%, Hausdorff distances (weigh segmentation result and goldstandard it
Between maximum mismatch degree) be 77.79 ± 2.18mm, Fig. 4 is the goldstandard and segmentation result of three groups of typical images；Fig. 5 (a)
It is the quantitative target SUVmean, Bland- MTV and goldstandard between of the tumour using this technology split (b)
Altman schemes, and illustrates the high consistency of this technology and goldstandard.Experiment shows that our method is well positioned to meet clinic and examines
Disconnected and auxiliary formulates the demand for the treatment of plan, has huge practical value.
Finally it should be noted that：The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Pipe is described in detail the present invention with reference to foregoing embodiments, it will be understood by those of ordinary skill in the art that：Its according to
Can so modify to the technical solution described in foregoing embodiments, either to which part or all technical characteristic into
Row equivalent substitution；And these modifications or replacement, the essence of appropriate technical solution is departed from various embodiments of the present invention technology
The scope of scheme.
Claims (5)
- A kind of 1. automatic tumor imaging region segmentation method based on improved level set, it is characterised in that including：Obtain the original PET image to be split comprising lesion region and pre-processed and positioned so that it is determined that pretreated Lesion region PET image to be split；According to the CT images of lesion region and the pretreated lesion region PET image construction hypergraph to be split, so that just Step determines that the rough tumor region in the PET image is initial zero level collection；Improved Level Set Method is performed so that it is determined that tumor region to the initial zero level collection；Edge-smoothing processing is performed to the tumor region according to morphology operations.
- 2. the automatic tumor imaging region segmentation method according to claim 1 based on improved level set, its feature exist In described to obtain the original PET image to be split comprising lesion region and pre-processed and positioned so that it is determined that pre-processing Lesion region PET image to be split afterwards includes：To the voxel gray values in the original PET image to be split comprising lesion region divided by the developer agent injected Measure with patient body weight to be converted to SUV values, then carry out gaussian filtering and up-sampling, finally position and determine according to the SUV values Pretreated lesion region PET image to be split；The CT images and the pretreated lesion region PET image to be split according to lesion region constructs hypergraph, from And primarily determine that the rough tumor region in the PET image includes for initial zero level collection：CT image normalizations to the pretreated lesion region PET image to be split with lesion region and the knot after being multiplied Fruit builds hypergraph, recycles fuzzy C-means clustering, morphological erosion and generic threshold value method to primarily determine that in the PET image Rough tumor region be initial zero level collection；It is described that improved Level Set Method is performed to initial zero level collection so that it is determined that tumor region includes：The gradient fields information of pretreated lesion region PET image to be split is added to the initial zero level to concentrate, structure New EVOLUTION EQUATION is built, to the rough tumor region in the PET image with identical with the resolution ratio of the original PET image Resolution ratio performs down-sampling and obtains initial zero level collection；Successive ignition is carried out to the initial zero level collection in the EVOLUTION EQUATION according to finite difference calculus and determines tumor region.
- 3. the automatic tumor imaging region segmentation method according to claim 2 based on improved level set, its feature exist In described to be positioned according to SUV values and determine that pretreated lesion region PET image to be split includes：The SUV peak values each cut into slices in the original PET image to be split comprising lesion region are calculated, are chosen more than foot maximum SUV peak values two neighboring minimum SUV values between corresponding section as section where lesion region so that it is determined that pretreatment Lesion region PET image to be split afterwards；Correspondingly, the EVOLUTION EQUATION is：Wherein, IσIt is pretreated lesion region PET image to be split, (variance of Gaussian kernel is σ), φ is initial zero level Collection or initial zero level collection carry out the function of the level set after iteration several times,<*>Represent the IσWith the gradient vector of φ it Between angle, | * | represent the vectorial amplitude, c1And c2Represent respectively in initial zero level collection (or initial zero level collection) The average gray value of outer voxel, i.e.,:Wherein, Ω represents image-region,δ functions are by following smooth function δεApproximation obtains：Φ is to adjust function and the company of being allowed to that initial zero level collection or initial zero level collection carry out the level set after iteration several times Continuousization, needs to carry out following gaussian filtering to level set function after each iteration：φ=Gσ* φ,Wherein, GσIt is the Gaussian kernel that variance is σ, * represents convolution algorithm, initial function φ0It is defined as follows：Wherein, c0It is normal number, R0To be down-sampled to the rough tumor region that original PET image resolution ratio obtains.
- 4. the automatic tumor imaging region segmentation method according to claim 1 based on improved level set, its feature exist In described that tumor region execution edge-smoothing processing is included according to morphology operations：Morphology opening operation and closed operation are performed to the tumor region with chondritic element to eliminate edge protuberance and fill out Fill cavity.
- 5. the automatic tumor imaging region segmentation method according to claim 4 based on improved level set, its feature exist In, it is described with chondritic element morphology opening operation and closed operation are performed to the tumor region eliminate edge protuberance and Filling cavity includes：Carry out opening operation and closed operation processing respectively using formula (6) and (7)：Opening operation processing：Closed operation is handled：Wherein,Represent dilation operation symbol,Represent erosion operation symbol, M represents the bianry image of the tumor region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510124586.7A CN104715484B (en) | 2015-03-20 | 2015-03-20 | Automatic tumor imaging region segmentation method based on improved level set |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510124586.7A CN104715484B (en) | 2015-03-20 | 2015-03-20 | Automatic tumor imaging region segmentation method based on improved level set |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104715484A CN104715484A (en) | 2015-06-17 |
| CN104715484B true CN104715484B (en) | 2018-04-13 |
Family
ID=53414778
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510124586.7A Active CN104715484B (en) | 2015-03-20 | 2015-03-20 | Automatic tumor imaging region segmentation method based on improved level set |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104715484B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056611B (en) * | 2016-06-03 | 2019-01-11 | 上海交通大学 | Level set image segmentation method and its system based on area information and marginal information |
| CN106447682A (en) * | 2016-08-29 | 2017-02-22 | 天津大学 | Automatic segmentation method for breast MRI focus based on Inter-frame correlation |
| TWI622959B (en) * | 2017-06-30 | 2018-05-01 | 中國醫藥大學附設醫院 | A medical image processing system and its method |
| CN108257134B (en) * | 2017-12-21 | 2022-08-23 | 深圳大学 | Nasopharyngeal carcinoma focus automatic segmentation method and system based on deep learning |
| CN108961222A (en) * | 2018-06-19 | 2018-12-07 | 江西大福医疗科技股份有限公司 | A kind of cervical carcinoma early screening recognition methods based on gynecatoptron image |
| CN109035208A (en) * | 2018-06-29 | 2018-12-18 | 上海联影医疗科技有限公司 | Recognition methods, device and the PET system in hypermetabolism region |
| CN108986114B (en) * | 2018-07-11 | 2022-03-29 | 中南大学 | Automatic abdominal CT sequence image liver segmentation method based on level set and shape descriptor |
| CN109671054A (en) * | 2018-11-26 | 2019-04-23 | 西北工业大学 | The non-formaldehyde finishing method of multi-modal brain tumor MRI |
| CN112837324A (en) * | 2021-01-21 | 2021-05-25 | 山东中医药大学附属医院 | Automatic tumor image region segmentation system and method based on improved level set |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1814323A (en) * | 2005-01-31 | 2006-08-09 | 重庆海扶(Hifu)技术有限公司 | Focusing ultrasonic therapeutical system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080030497A1 (en) * | 2005-12-08 | 2008-02-07 | Yangqiu Hu | Three dimensional modeling of objects |
-
2015
- 2015-03-20 CN CN201510124586.7A patent/CN104715484B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1814323A (en) * | 2005-01-31 | 2006-08-09 | 重庆海扶(Hifu)技术有限公司 | Focusing ultrasonic therapeutical system |
Non-Patent Citations (2)
| Title |
|---|
| 《Segmentation of Mouse Dynamic PET Images Using a Multiphases Level Set Method》;Jinxiu ChengLiao 等;《Phys Med Biol》;20101107;第55卷(第21期);全文 * |
| 《基于改进水平集方法的Micro-CT 鼠脑三维自动化分割》;陈诗烨 等;《中国医疗器械杂志》;20121231;第36卷(第3期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104715484A (en) | 2015-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104715484B (en) | Automatic tumor imaging region segmentation method based on improved level set | |
| Charron et al. | Automatic detection and segmentation of brain metastases on multimodal MR images with a deep convolutional neural network | |
| Li et al. | Deep learning for variational multimodality tumor segmentation in PET/CT | |
| US10269122B2 (en) | Knowledge-based automatic image segmentation | |
| US11443433B2 (en) | Quantification and staging of body-wide tissue composition and of abnormal states on medical images via automatic anatomy recognition | |
| Peng et al. | Segmentation of lung in chest radiographs using hull and closed polygonal line method | |
| Linguraru et al. | Liver and tumor segmentation and analysis from CT of diseased patients via a generic affine invariant shape parameterization and graph cuts | |
| Pandey et al. | A systematic review of the automatic kidney segmentation methods in abdominal images | |
| Di et al. | Automatic liver tumor segmentation from CT images using hierarchical iterative superpixels and local statistical features | |
| Abdollahi et al. | Radiomics-guided radiation therapy: opportunities and challenges | |
| Basu et al. | RadFormer: Transformers with global–local attention for interpretable and accurate Gallbladder Cancer detection | |
| Zhao et al. | Clinical applications of deep learning in breast MRI | |
| Deng et al. | Fusion of FDG-PET image and clinical features for prediction of lung metastasis in soft tissue sarcomas | |
| Wang et al. | Bowelnet: Joint semantic-geometric ensemble learning for bowel segmentation from both partially and fully labeled ct images | |
| Mohammed et al. | Liver segmentation: A survey of the state-of-the-art | |
| Vo et al. | Effects of multiple filters on liver tumor segmentation from CT images | |
| Shim et al. | Fully automated breast segmentation on spiral breast computed tomography images | |
| Ibragimov et al. | Automated hepatobiliary toxicity prediction after liver stereotactic body radiation therapy with deep learning-based portal vein segmentation | |
| Rachmawati et al. | Automatic whole-body bone scan image segmentation based on constrained local model | |
| Ying et al. | Weakly supervised segmentation of uterus by scribble labeling on endometrial cancer MR images | |
| Sun et al. | Application of artificial intelligence nuclear medicine automated images based on deep learning in tumor diagnosis | |
| Feng et al. | Multi-stage fully convolutional network for precise prostate segmentation in ultrasound images | |
| Amritha et al. | Liver tumor segmentation and classification using deep learning | |
| Saeku et al. | Liver and tumor segmentation in selective internal radiation therapy 99m Tc-MAA SPECT/CT images using MANet and histogram adjustment | |
| Baydoun et al. | Auto-contouring FDG-PET/MR images for cervical cancer radiation therapy: An intelligent sequential approach using focally trained, shallow U-Nets |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |