CN109635846A - A kind of multiclass medical image judgment method and system - Google Patents
A kind of multiclass medical image judgment method and system Download PDFInfo
- Publication number
- CN109635846A CN109635846A CN201811368056.7A CN201811368056A CN109635846A CN 109635846 A CN109635846 A CN 109635846A CN 201811368056 A CN201811368056 A CN 201811368056A CN 109635846 A CN109635846 A CN 109635846A
- Authority
- CN
- China
- Prior art keywords
- medical image
- image
- processing
- multiclass
- algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Pathology (AREA)
- Databases & Information Systems (AREA)
- Image Analysis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of multiclass medical image judgment method and system, method includes processing medical image to obtain characteristics of image, analyzes different types of medical image, and the classifier based on random deep woods is to obtain optimal feature subset;Handling optimal feature subset by support vector machines with output category result and assigns medical image corresponding type label;Corresponding disease image Processing Algorithm is selected to handle the medical image again to obtain genius morbi, based on sorter network processing genius morbi to export diagnostic result based on type label.System is for executing corresponding method.The present invention, to obtain characteristics of image, obtains optimal feature subset based on classifier by processing medical image;Pass through support vector machines output category result, select disease image Processing Algorithm processing medical image to obtain genius morbi according to classification results, based on sorter network processing genius morbi to export diagnostic result, it can be improved Medical Image Processing speed, reduce the processing load of doctor.
Description
Technical field
The present invention relates to technical field of image processing, especially a kind of multiclass medical image judgment method and system.
Background technique
In clinic, the most of of medical data are medical image, and with the update of image acquisition technology and equipment at
This reduction, from now in the medical field, the importance of medical image can deepen, and the quantity generated during diagnosis can also increase
Add.Meanwhile the quantity of skilled medical practitioner will not increase significantly, in this case, identification medical image will for doctor
The too many time is occupied, working efficiency is caused to reduce, is unfavorable for medical diagnosis, increases the probability of happening of malpractice.
Meanwhile various medical images is many kinds of, the image of different types has each different feature, it is desirable to ripe
Practice that grasp the identification of image of all or most be very difficult thing, if either increase image recognition personnel or
The training burden that person increases the image recognition of doctor is all non-efficiency and valuableness, it is therefore desirable to which one can reduce conscientiously doctor and sentence
The method of disconnected medical image.
Summary of the invention
The present invention is directed to solve at least some of the technical problems in related technologies.For this purpose, of the invention
One purpose is to provide a kind of multiclass medical image judgment method and system.
The technical scheme adopted by the invention is that: a kind of multiclass medical image judgment method, comprising steps of processing medicine figure
As to obtain characteristics of image, the classifier based on random deep woods screens described image feature to obtain optimal feature subset;Pass through
The support vector machines processing optimal feature subset is with output category result and assigns the medical image corresponding type label;
Corresponding disease image Processing Algorithm is selected to handle the medical image again to obtain genius morbi based on the type label,
Based on the sorter network processing genius morbi to export diagnostic result.
Further, the sorter network is based on nearest neighbor algorithm and/or handles institute in random deep woods and/or BP neural network
Genius morbi is stated to export diagnostic result.
Further, the step of processing medical image is to obtain characteristics of image includes: based on airspace or frequency domain filtering
The denoising for carrying out medical image handles medical image to realize the segmentation of area-of-interest, based on preset based on template matching
Feature identification model handles area-of-interest to obtain characteristics of image.
Further, the step of classifier of the random deep woods screens described image feature includes: according to characteristics of image
Subcharacter setting corresponding number Random Forest model, calculate the different degree of each subcharacter, reject different degree and be less than threshold
The subcharacter of value is to obtain optimal feature subset.
Further, it is suitable for plurality of medical diagnostic imaging, such as: galactophore image, lung, ultrasonic wave, blood cell and sperm etc.,
Comprising steps of carrying out the denoising of medical image based on airspace or frequency domain filtering, medicine figure is handled based on sech template matching algorithm
As to determine area-of-interest, the segmentation of area-of-interest is realized based on watershed algorithm and RSF model.
It is of the present invention another solution is that a kind of multiclass medical image judges system, comprising: primary module,
For handling medical image to obtain characteristics of image, the classifier screening described image feature based on random deep woods is optimal to obtain
Character subset;Secondary module, for handling the optimal feature subset by support vector machines with output category result and assigning
The corresponding type label of the medical image;Final module, for being selected at corresponding disease image based on the type label
Adjustment method handles the medical image again to obtain genius morbi, is examined based on the sorter network processing genius morbi with exporting
Disconnected result.
Further, the sorter network is based on nearest neighbor algorithm and/or handles institute in random deep woods and/or BP neural network
Genius morbi is stated to export diagnostic result.
Further, the step of processing medical image is to obtain characteristics of image includes: based on airspace or frequency domain filtering
The denoising for carrying out medical image handles medical image to realize the segmentation of area-of-interest, based on preset based on template matching
Feature identification model handles area-of-interest to obtain characteristics of image.
Further, the step of classifier of the random deep woods screens described image feature includes: according to characteristics of image
Subcharacter setting corresponding number Random Forest model, calculate the different degree of each subcharacter, reject different degree and be less than threshold
The subcharacter of value is to obtain optimal feature subset.
Further, the primary module is also used to be carried out the denoising of medical image based on airspace or frequency domain filtering, is based on
Sech template matching algorithm handles medical image to determine area-of-interest, based on watershed algorithm and RSF model to realize sense
The segmentation in interest region.
The beneficial effects of the present invention are: the present invention handles medical image to obtain characteristics of image, based on dividing for random deep woods
Class device screens described image feature to obtain optimal feature subset;The optimal feature subset is handled by support vector machines with defeated
Classification results out select corresponding disease image Processing Algorithm to handle the medical image again to obtain disease according to classification results
Sick feature can be improved Medical Image Processing speed, drop based on the sorter network processing genius morbi to export diagnostic result
The processing load of low doctor.
Detailed description of the invention
Fig. 1 is the schematic diagram of medical image judgment method of the invention;
Fig. 2 is a kind of schematic diagram of Medical Image Processing frame of the invention.
Specific embodiment
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.
Embodiment 1
The present embodiment is for the shortcomings that illustrating the prior art and resolving ideas of the invention
It is quite reasonable selection that intracorporal internal organ are detected by way of microwave imaging, because not will cause the damage of internal organ
Wound, will not cause the trace of body surface;Thus derive the method for the various health status by medical image acquisition human body, and
Reality prove these methods be it is effective, with and come be image data amount caused by the broader applications of medical image increasing
Add.
In the case where artificial intelligence is not able to achieve high resolution yet, in fact, the confirmation of the last state of an illness there is still a need for by
Mankind doctor is responsible for, this just brings very big homework burden to doctor, and the fatigue on body can bring the decline of recognition capability,
It is very big unfavorable that this can bring to the diagnosis of the state of an illness, leads to malpractice.
Therefore, the present invention provides a kind of multiclass medical image judgment method as shown in Figure 1, comprising steps of S1, processing doctor
Image is learned to obtain characteristics of image, based on the classifier screening described image feature of random deep woods to obtain optimal feature subset;
S2, the optimal feature subset is handled by support vector machines with output category result and assigns the medical image corresponding class
Type label;S3, corresponding disease image Processing Algorithm is selected to handle the medical image again to obtain based on the type label
Genius morbi is taken, based on the sorter network processing genius morbi to export diagnostic result.
Its principle is that there are many existing algorithm for being related to medical image, this programme be extract these schemes it is common
The step of, the identification of primary is realized according to these steps, after completing primary identification, is being carried out once according to special algorithm
Processing is to obtain better treatment effect.
Wherein, handle medical image (referred to as image) comprising steps of
Denoising, for Image filter arithmetic, is generally divided into two types, one is airspace filters, another is then
In frequency domain filtering, the present embodiment, using image filtering is carried out on airspace, (frequency domain filtering algorithm is on the medical image
Good effect can be obtained), under normal circumstances, being filtered to medical image is for smoothed image, and therefore, we make mostly
The pretreatment of medical image is carried out with median filtering, mean filter, wherein median filtering is the surrounding neighbors collection with pixel
The intermediate value of gray value replaces the gray value at preimage vegetarian refreshments in conjunction, and the removal for salt-pepper noise and patch noise, effect is very
Obviously, and mean filter be then using the gray value in pixel surrounding neighbors set mean value replace preimage vegetarian refreshments gray scale
Value.
Template matching, template matching algorithm are a kind of common methods of basic target positioning in pattern-recognition, mainly
It is by definition template image, the similarity of the sub-image area of calculation template image and target image finds target position;
It, generally can be using circular shuttering, spherical template, Gaussian template and hyperbolic just for the suspicious lesions zone location of medical image
Template is cut, the specific function that handles includes Gaussian function and hyperbolic tangent function (sech function), wherein Gaussian function is relatively more suitable
Detection microcalcifications are closed, and hyperbolic tangent function is suitble to detect lump.Therefore, the present embodiment uses sech template matching to calculate
Method is difficult with a kind of template matching of size to all lesion regions simultaneously because the shape of lesion region, different sizes.
Therefore, the template matching based on single size can not detect all lesion regions well.Therefore, more rulers can be taken
Degree strategy, constructs the template with sizes size, matches to target image, arrive respectively to multiple dimensioned template detection
As a result it is merged, obtains final testing result (determining lesion region).
Feature identification carries out the extraction of feature after the determination for completing above-mentioned lesion region, these features include doctor
The number of 4 connected regions after learning the dimension of image, gray average, variance, binary conversion treatment and number, the face of 8 connected regions
Colour moment, gray level co-occurrence matrixes, histograms of oriented gradients and local binary patterns etc. (i.e. subcharacter);These features are all in image
Commonly used index or parameter in processing, acquisition methods are more, and the present embodiment is without further explanation.
After carrying out feature extraction to medical image, obtained feature vector dimension is very big, it means that calculation amount
It is sizable.Therefore, we will carry out feature selecting, screen by random forest to each feature, and rejecting is not
So important feature obtains optimal feature subset.In addition, the algorithm for Medical Images Classification has very much, such as commonly
Three kinds of classifiers, k nearest neighbor (k-NearestNeighbor, KNN), support vector machines and neural network classifier, for extraction
To feature vector classified (i.e. judgement belongs to the image of which kind of disease), the present embodiment chooses support vector machines as classification
Device is used for output category result.
After completing above-mentioned judgement medical image, then corresponding Processing Algorithm can be selected according to specific disease type,
The present embodiment is by taking the medical image of mammary gland as an example, after completing above-mentioned deterministic process:
It takes the mode of median filtering to realize denoising first, passes through sech template matching algorithm calculation template image and target
Similarity graph picture between image to similarity graph as binary conversion treatment, then passes through a series of preset rules removal false positive
Region, remaining region are the suspicious mass region (i.e. area-of-interest) detected.
Then being handled by preset feature identification model can be with lump region, wherein feature identification model is in the nature
The recognizer of above-mentioned various features, the technical field of the acquisitions of the parameters such as dimension, gray average, variance in image procossing
It is open more sufficient basic technology, the present embodiment is without additional explanation.
Finally by trained classifier, (i.e. sorter network is carried out by disease datas such as MIAS data sets
Training is to obtain the ability of classification) carry out feature classification (this subseries the result is that diagnostic result, technological essence are bases
The degree of similitude exports a corresponding value).
Embodiment 2
For the present embodiment for explaining preset rules on the basis of embodiment 1, the purposes of the preset rules is at image
A part of content/data of exclusion of selectivity, specific rule include: during reason
1) region area:, can by the lump diameter in analysis MIAS data set (i.e. William Kang Xing breast cancer data set)
To judge the possible area of lump in a certain range, pectoral region is likely to more than this range;
2) shape: being more than centainly using the length-width ratio of rectangle as form factor according to the minimum outsourcing rectangle of candidate region
Size then can remove;
3) gray average: the pixel value in lump region is generally relatively high, then can determine that the region not lower than certain threshold value
It is lump;
4) gray variance: the variance in lump region generally all less (is not more than certain threshold value);
5) morphological feature: eccentricity and circularity.
By the formulation of above-mentioned rule, false positive region can be removed and filter out suspicious region.
It is above-mentioned filter out suspicious region on the basis of, in addition the present embodiment proposes a kind of watershed of binding marker control
The method that algorithm and RSF model combine is to realize the segmentation for lump:
Area-of-interest is split using the watershed algorithm of marking of control, obtains the approximate location in lump region;
Extract the boundary in lump region, the coordinate of record delimitation point;
The closed curve that the boundary point coordinate recorded is formed is used as the initial profile line of RSF model
RSF model, which is finely divided, to be cut, and accurate lump profile is obtained.
Embodiment 3
The present embodiment provides a kind of Medical Image Processing frames as shown in Figure 2, comprising:
As the medical image databases in medical image source, from the various features of medicine image zooming-out (including color moment,
Gray level co-occurrence matrixes, histograms of oriented gradients and local binary patterns etc.), various features are passed through into random forest Fusion Features,
Classification processing based on SVM classifier exports the classification (i.e. class label) of image, according to class label, from disease image
Reason algorithm data-base chooses suitable Processing Algorithm to handle medical image and obtain various features (including color moment, gray scale are total
Raw matrix, histograms of oriented gradients and local binary patterns etc., at this point, marking it for genius morbi), then according to KNN, (K is close
It is adjacent), SVM (support vector machines), BPNN (neural network) classifier (any combination) classified (i.e. output judging result).
It is to be illustrated to preferable implementation of the invention, but the invention is not limited to the implementation above
Example, those skilled in the art can also make various equivalent variations on the premise of without prejudice to spirit of the invention or replace
It changes, these equivalent deformations or replacement are all included in the scope defined by the claims of the present application.
Claims (10)
1. a kind of multiclass medical image judgment method, which is characterized in that comprising steps of
Medical image is handled to obtain characteristics of image, the classifier screening described image feature based on random deep woods is optimal to obtain
Character subset;
The optimal feature subset is handled with output category result by support vector machines and to assign the medical image corresponding
Type label;
Corresponding disease image Processing Algorithm is selected to handle the medical image again to obtain disease based on the type label
Feature, based on the sorter network processing genius morbi to export diagnostic result.
2. a kind of multiclass medical image judgment method according to claim 1, which is characterized in that the sorter network is based on
Nearest neighbor algorithm and/or the genius morbi is handled to export diagnostic result in random deep woods and/or BP neural network.
3. a kind of multiclass medical image judgment method according to claim 1 or 2, which is characterized in that the processing medicine
The step of image is to obtain characteristics of image include:
The denoising that medical image is carried out based on airspace or frequency domain filtering, it is interested to realize based on template matching processing medical image
The segmentation in region, based on preset feature identification model processing area-of-interest to obtain characteristics of image.
4. a kind of multiclass medical image judgment method according to claim 3, which is characterized in that point of the random deep woods
Class device screen described image feature the step of include:
The Random Forest model of corresponding number is set according to the subcharacter of characteristics of image, calculates the different degree of each subcharacter,
It rejects different degree and is less than the subcharacter of threshold value to obtain optimal feature subset.
5. a kind of multiclass medical image judgment method according to claim 3, comprising steps of
The denoising that medical image is carried out based on airspace or frequency domain filtering, based on sech template matching algorithm processing medical image with true
Determine area-of-interest, the segmentation of area-of-interest is realized based on watershed algorithm and RSF model.
6. a kind of multiclass medical image judges system characterized by comprising
Primary module, for handling medical image to obtain characteristics of image, the classifier based on random deep woods screens described image
Feature is to obtain optimal feature subset;
Secondary module, for handling the optimal feature subset by support vector machines with output category result and assigning the doctor
Learn the corresponding type label of image;
Final module, for selecting corresponding disease image Processing Algorithm to handle the medicine figure again based on the type label
As handling the genius morbi based on sorter network to export diagnostic result to obtain genius morbi.
7. a kind of multiclass medical image according to claim 6 judges system, which is characterized in that the sorter network is based on
Nearest neighbor algorithm and/or the genius morbi is handled to export diagnostic result in random deep woods and/or BP neural network.
8. a kind of multiclass medical image according to claim 6 or 7 judges system, which is characterized in that the processing medicine
The step of image is to obtain characteristics of image include:
The denoising that medical image is carried out based on airspace or frequency domain filtering, it is interested to realize based on template matching processing medical image
The segmentation in region, based on preset feature identification model processing area-of-interest to obtain characteristics of image.
9. a kind of multiclass medical image according to claim 8 judges system, which is characterized in that point of the random deep woods
Class device screen described image feature the step of include:
The Random Forest model of corresponding number is set according to the subcharacter of characteristics of image, calculates the different degree of each subcharacter,
It rejects different degree and is less than the subcharacter of threshold value to obtain optimal feature subset.
10. a kind of multiclass medical image according to claim 8 judges system, it is suitable for claim 5 the method,
It is characterized in that, the primary module, is also used to carry out the denoising of medical image based on airspace or frequency domain filtering, is based on sech template
Matching algorithm handles medical image to determine area-of-interest, based on watershed algorithm and RSF model to realize area-of-interest
Segmentation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811368056.7A CN109635846B (en) | 2018-11-16 | 2018-11-16 | Multi-type medical image judging method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811368056.7A CN109635846B (en) | 2018-11-16 | 2018-11-16 | Multi-type medical image judging method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109635846A true CN109635846A (en) | 2019-04-16 |
CN109635846B CN109635846B (en) | 2023-06-02 |
Family
ID=66068370
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811368056.7A Active CN109635846B (en) | 2018-11-16 | 2018-11-16 | Multi-type medical image judging method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109635846B (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110070083A (en) * | 2019-04-24 | 2019-07-30 | 深圳市微埃智能科技有限公司 | Image processing method, device, electronic equipment and computer readable storage medium |
CN110517239A (en) * | 2019-08-20 | 2019-11-29 | 东软医疗系统股份有限公司 | A kind of medical image detection method and device |
CN110600107A (en) * | 2019-08-29 | 2019-12-20 | 上海联影智能医疗科技有限公司 | Method for screening medical images, computer device and readable storage medium |
CN110610756A (en) * | 2019-07-26 | 2019-12-24 | 赛诺威盛科技(北京)有限公司 | Method for realizing automatic classified printing of films based on DICOM image information |
CN111062943A (en) * | 2020-03-18 | 2020-04-24 | 南京景三医疗科技有限公司 | Plaque stability determination method and device and medical equipment |
CN111091574A (en) * | 2019-12-21 | 2020-05-01 | 中国人民解放军第四军医大学 | Medical image segmentation method based on single pixel characteristics |
CN111863232A (en) * | 2020-08-06 | 2020-10-30 | 罗春华 | Remote disease intelligent diagnosis system based on block chain and medical image |
CN112150565A (en) * | 2020-09-25 | 2020-12-29 | 呙明 | Medical image management system |
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 |
CN113077891A (en) * | 2021-04-15 | 2021-07-06 | 王小娟 | Big data disease diagnosis system based on algorithm, block chain and medical image |
CN113537099A (en) * | 2021-07-21 | 2021-10-22 | 招商局重庆交通科研设计院有限公司 | Dynamic detection method for fire smoke in highway tunnel |
WO2021217854A1 (en) * | 2020-04-30 | 2021-11-04 | 平安科技(深圳)有限公司 | False positive filtering method, device, equipment, and storage medium |
CN113936191A (en) * | 2021-10-21 | 2022-01-14 | 平安国际智慧城市科技股份有限公司 | Picture classification model training method, device, equipment and storage medium |
CN115100646A (en) * | 2022-06-27 | 2022-09-23 | 武汉兰丁智能医学股份有限公司 | Cell image high-definition rapid splicing identification marking method |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120321174A1 (en) * | 2011-06-15 | 2012-12-20 | Siemens Aktiengesellschaft | Image Processing Using Random Forest Classifiers |
US20150148677A1 (en) * | 2013-11-22 | 2015-05-28 | General Electric Company | Method and system for lesion detection in ultrasound images |
CN104732213A (en) * | 2015-03-23 | 2015-06-24 | 中山大学 | Computer-assisted lump detecting method based on mammary gland magnetic resonance image |
CN105279483A (en) * | 2015-09-28 | 2016-01-27 | 华中科技大学 | Fall-down behavior real-time detection method based on depth image |
US20160260211A1 (en) * | 2013-10-12 | 2016-09-08 | H. Lee Moffitt Cancer Center And Research Institute, Inc. | Systems and methods for diagnosing tumors in a subject by performing a quantitative analysis of texture-based features of a tumor object in a radiological image |
CN107330892A (en) * | 2017-07-24 | 2017-11-07 | 内蒙古工业大学 | A kind of sunflower disease recognition method based on random forest method |
CN107845098A (en) * | 2017-11-14 | 2018-03-27 | 南京理工大学 | Liver cancer image full-automatic partition method based on random forest and fuzzy clustering |
CN108464840A (en) * | 2017-12-26 | 2018-08-31 | 安徽科大讯飞医疗信息技术有限公司 | A kind of breast lump automatic testing method and system |
-
2018
- 2018-11-16 CN CN201811368056.7A patent/CN109635846B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120321174A1 (en) * | 2011-06-15 | 2012-12-20 | Siemens Aktiengesellschaft | Image Processing Using Random Forest Classifiers |
US20160260211A1 (en) * | 2013-10-12 | 2016-09-08 | H. Lee Moffitt Cancer Center And Research Institute, Inc. | Systems and methods for diagnosing tumors in a subject by performing a quantitative analysis of texture-based features of a tumor object in a radiological image |
US20150148677A1 (en) * | 2013-11-22 | 2015-05-28 | General Electric Company | Method and system for lesion detection in ultrasound images |
CN104732213A (en) * | 2015-03-23 | 2015-06-24 | 中山大学 | Computer-assisted lump detecting method based on mammary gland magnetic resonance image |
CN105279483A (en) * | 2015-09-28 | 2016-01-27 | 华中科技大学 | Fall-down behavior real-time detection method based on depth image |
CN107330892A (en) * | 2017-07-24 | 2017-11-07 | 内蒙古工业大学 | A kind of sunflower disease recognition method based on random forest method |
CN107845098A (en) * | 2017-11-14 | 2018-03-27 | 南京理工大学 | Liver cancer image full-automatic partition method based on random forest and fuzzy clustering |
CN108464840A (en) * | 2017-12-26 | 2018-08-31 | 安徽科大讯飞医疗信息技术有限公司 | A kind of breast lump automatic testing method and system |
Non-Patent Citations (4)
Title |
---|
孔倩等: "基于多目标神经网络的前列腺癌诊断方法", 《系统工程理论与实践》 * |
孟江等: "基于CT图像纹理分析的结肠息肉鉴别研究", 《中国医学装备》 * |
汪丽丽等: "基于Spark的肿瘤基因混合特征选择方法", 《计算机工程》 * |
沈健等: "基于边缘检测的多类别医学图像分类方法", 《数据采集与处理》 * |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110070083A (en) * | 2019-04-24 | 2019-07-30 | 深圳市微埃智能科技有限公司 | Image processing method, device, electronic equipment and computer readable storage medium |
CN110610756A (en) * | 2019-07-26 | 2019-12-24 | 赛诺威盛科技(北京)有限公司 | Method for realizing automatic classified printing of films based on DICOM image information |
CN110517239A (en) * | 2019-08-20 | 2019-11-29 | 东软医疗系统股份有限公司 | A kind of medical image detection method and device |
CN110600107A (en) * | 2019-08-29 | 2019-12-20 | 上海联影智能医疗科技有限公司 | Method for screening medical images, computer device and readable storage medium |
CN110600107B (en) * | 2019-08-29 | 2022-07-26 | 上海联影智能医疗科技有限公司 | Method for screening medical images, computer device and readable storage medium |
CN111091574A (en) * | 2019-12-21 | 2020-05-01 | 中国人民解放军第四军医大学 | Medical image segmentation method based on single pixel characteristics |
CN111091574B (en) * | 2019-12-21 | 2024-04-02 | 中国人民解放军第四军医大学 | Medical image segmentation method based on single pixel characteristics |
CN111062943A (en) * | 2020-03-18 | 2020-04-24 | 南京景三医疗科技有限公司 | Plaque stability determination method and device and medical equipment |
CN111062943B (en) * | 2020-03-18 | 2020-06-09 | 南京景三医疗科技有限公司 | Plaque stability determination method and device and medical equipment |
WO2021217854A1 (en) * | 2020-04-30 | 2021-11-04 | 平安科技(深圳)有限公司 | False positive filtering method, device, equipment, and storage medium |
CN111863232A (en) * | 2020-08-06 | 2020-10-30 | 罗春华 | Remote disease intelligent diagnosis system based on block chain and medical image |
CN111863232B (en) * | 2020-08-06 | 2021-02-19 | 深圳市柯尼达巨茂医疗设备有限公司 | Remote disease intelligent diagnosis system based on block chain and medical image |
CN112150565A (en) * | 2020-09-25 | 2020-12-29 | 呙明 | Medical image management system |
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 |
CN113077891A (en) * | 2021-04-15 | 2021-07-06 | 王小娟 | Big data disease diagnosis system based on algorithm, block chain and medical image |
CN113537099A (en) * | 2021-07-21 | 2021-10-22 | 招商局重庆交通科研设计院有限公司 | Dynamic detection method for fire smoke in highway tunnel |
CN113936191A (en) * | 2021-10-21 | 2022-01-14 | 平安国际智慧城市科技股份有限公司 | Picture classification model training method, device, equipment and storage medium |
CN115100646A (en) * | 2022-06-27 | 2022-09-23 | 武汉兰丁智能医学股份有限公司 | Cell image high-definition rapid splicing identification marking method |
CN115100646B (en) * | 2022-06-27 | 2023-01-31 | 武汉兰丁智能医学股份有限公司 | Cell image high-definition rapid splicing identification marking method |
Also Published As
Publication number | Publication date |
---|---|
CN109635846B (en) | 2023-06-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109635846A (en) | A kind of multiclass medical image judgment method and system | |
Beevi et al. | Automatic mitosis detection in breast histopathology images using convolutional neural network based deep transfer learning | |
Akram et al. | Identification and classification of microaneurysms for early detection of diabetic retinopathy | |
dos Santos Ferreira et al. | Convolutional neural network and texture descriptor-based automatic detection and diagnosis of glaucoma | |
George et al. | Remote computer-aided breast cancer detection and diagnosis system based on cytological images | |
Akram et al. | Detection and classification of retinal lesions for grading of diabetic retinopathy | |
Lachure et al. | Diabetic retinopathy using morphological operations and machine learning | |
Dundar et al. | Computerized classification of intraductal breast lesions using histopathological images | |
CN109447998B (en) | Automatic segmentation method based on PCANet deep learning model | |
CN108257135A (en) | The assistant diagnosis system of medical image features is understood based on deep learning method | |
CN104899926A (en) | Medical image segmentation method and device | |
CN109614869A (en) | A kind of pathological image classification method based on multi-scale compress rewards and punishments network | |
CN104299242A (en) | Fluorescence angiography fundus image extraction method based on NGC-ACM | |
Abiyev et al. | Fuzzy neural networks for identification of breast cancer using images' shape and texture features | |
CN115100474B (en) | Thyroid gland puncture image classification method based on topological feature analysis | |
Hassan et al. | A dilated residual hierarchically fashioned segmentation framework for extracting Gleason tissues and grading prostate cancer from whole slide images | |
Al Jannat et al. | Detection of multiple sclerosis using deep learning | |
Alnabelsi | Cervical cancer diagnostic system using adaptive fuzzy moving k-means algorithm and fuzzy min-max neural network | |
Syed et al. | Detection of tumor in MRI images using artificial neural networks | |
Balakrishnan | NDC-IVM: An automatic segmentation of optic disc and cup region from medical images for glaucoma detection | |
Ding et al. | Classification of chromosome karyotype based on faster-rcnn with the segmatation and enhancement preprocessing model | |
Prem et al. | Classification of exudates for diabetic retinopathy prediction using machine learning | |
CN110428405A (en) | Method, relevant device and the medium of lump in a kind of detection biological tissue images | |
Siddiqui et al. | Attention based covid-19 detection using generative adversarial network | |
Li et al. | Automatic detecting and recognition of casts in urine sediment images |
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 |