CN103488977A - Medical image management system based on SVM - Google Patents
Medical image management system based on SVM Download PDFInfo
- Publication number
- CN103488977A CN103488977A CN201310431609.XA CN201310431609A CN103488977A CN 103488977 A CN103488977 A CN 103488977A CN 201310431609 A CN201310431609 A CN 201310431609A CN 103488977 A CN103488977 A CN 103488977A
- Authority
- CN
- China
- Prior art keywords
- medical image
- medical
- image
- management system
- master pattern
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention discloses a medical image management system based on an SVM. The medical image management system based on the SVM is characterized by comprising an original image database used for storing medical images, a classification image database used for storing the medical images in the original image database in a classification mode according to disease varieties, a feature recognizer used for recognizing the medical features of the medical images in the original image database and a classifier. The classifier is used for recognizing the medical images in the image database through the feature recognizer to obtain the medical features of the medical images, then recognizing diseases and storing the medical images in the classification image database according to the disease varieties. According to the medical image management system based on the SVM, the Adaboost cascade classifier is used for carrying out training in advance, the good recognition rate and the good false accept rate are achieved for different samples, calculation speed is high, a large number of medical images can be recognized and classified fast, and the medical image management system can be used for large-scale medical image data database systems.
Description
Technical field
The invention belongs to medical domain, relate in particular to medical image management wherein.
Background technology
The application of medical image in clinical diagnosis and treatment is increasingly extensive, how to utilize a large amount of medical images, and diagnosis and treatment that auxiliary doctor carries out disease are the problems that current industry is all being studied.An outstanding medical image management system must, according to the careful classification of improving of kinds of Diseases and donor, in order to carry out at any time efficient retrieval, reach information analysis and excavation.Traditional medical image is all the method that adopts artificial cognition, script classify, but, increasing along with medical image, the differences such as the ethnic group especially wherein related to, sex, age, cause its difficulty to artificial cognition increasing, and workload increase day by day.How to address this problem, introduce increasingly mature computer image recognition technology, it is following development trend that replacement manually completes above-mentioned work.The present invention is in this research work, the Research Thinking of oneself is proposed, to a kind of new method is provided, especially utilize the current SVM(Support Vector Machine of maturation, support vector machine, as a kind of trainable machine learning method, the model parameter mature technology after the dependence small-sample learning) technology, realize the Classification Management work of robotization.
Summary of the invention
The present invention is the medical image management system based on SVM according to a kind of mechanized classification of above-mentioned thinking design.
Technical scheme of the present invention is to provide a kind of medical image management system based on SVM, and it is characterized in that: it comprises:
One raw video database, for storing medical image;
One classification image database, for the medical image according to the described raw video database of kinds of Diseases classification storage;
One feature recognizer, for the medical features of the medical image of identifying described raw video database;
One sorter, identified the medical image of described image database for utilizing described feature recognizer, obtains its medical features, then disease identified, and be stored in described classification image database according to kinds of Diseases.
Preferably, described feature recognizer utilizes based on Adaboost cascade classifier training in advance.
Preferably, also comprise:
One master pattern storehouse, for storing the medical image of human organ master pattern;
One standard disease model storehouse, for the medical features of storage standards disease model.
Preferably, the medical image of described master pattern library storage human organ master pattern is classified according to sex, age, ethnic group, height, the body weight of human body.
Preferably, the step that described feature recognizer obtains the medical features of described medical image comprises:
1) image pre-service, to described medical image to be identified become a full member, convergent-divergent, filtering, resolution adjustment process;
2) the organic image standardization is adjusted, to described medical image to be identified be out of shape, local convergent-divergent processes, make organic image outline line wherein be tending towards overlapping with the outline line of medical image in described master pattern storehouse, the medical image in described master pattern storehouse screens and obtains in described master pattern storehouse;
3), in previous step, the outline line of the medical image in organic image outline line and described master pattern storehouse not intersection carries out the genius morbi extraction.
The condition of the medical image preferably, described step 2) in screening described master pattern storehouse comprises: the sex of human body, age, ethnic group, height, body weight.
Medical image management system based on SVM of the present invention utilizes the Adaboost cascade classifier to carry out training in advance, there is discrimination and misclassification rate preferably for different samples, and computing velocity is fast, can carry out fast a large amount of medical image identification and classification, can be used for large-scale Medical imaging system.
The accompanying drawing explanation
Fig. 1 is system architecture diagram of the present invention.
Embodiment
Below the specific embodiment of the present invention is described in further detail.
As shown in Figure 1, a kind of medical image management system based on SVM of the present invention comprises:
Raw video database, classification image database, master pattern storehouse, standard disease model storehouse, feature recognizer and sorter.
Wherein:
The raw video database, for storing medical image and record the information of donor;
The classification image database, for the medical image according to kinds of Diseases classification storage raw video database;
The master pattern storehouse, for the medical image of the classification human organ master pattern of storing different sexes, age, ethnic group, height, body weight;
Standard disease model storehouse, for the medical features of storage standards disease model, comprise the stages of disease.
The feature recognizer be utilize good based on Adaboost cascade classifier training in advance, for the medical features of the medical image of identifying the raw video database.Adaboost is a kind of iterative algorithm, its core concept is to train different sorter (Weak Classifier) for same training set, then these Weak Classifiers are gathered, form a stronger final sorter (strong classifier), it is industry proven technique means already.
And sorter, utilize the feature recognizer to be identified the medical image in image database, obtain its medical features, then standard disease model storehouse major disease model is compared, disease is identified, and be stored in the classification image database according to kinds of Diseases.
Wherein, the step of the medical features of feature recognizer identification medical image comprises:
1) image pre-service, to medical image to be identified become a full member, convergent-divergent, filtering, resolution adjustment process;
2) the organic image standardization is adjusted, to medical image to be identified be out of shape, local convergent-divergent processes, make organic image outline line wherein be tending towards overlapping with the outline line of medical image in the master pattern storehouse, medical image in the master pattern storehouse screens and obtains in the master pattern storehouse, and the condition of screening comprises: the sex of human body, age, ethnic group, height, body weight; As the common practise of industry, can there be larger difference in the sex of different human bodies, age, ethnic group, height, the organic image that body weight is identical, and the image of identical disease same phase is also different;
3), in previous step, the outline line of the medical image in organic image outline line and master pattern storehouse not intersection carries out the genius morbi extraction;
4) medical features of the standard disease model in erworbene Krankenheit feature in previous step and standard disease model storehouse is compared, obtain disease type and developing stage thereof.
Above embodiment is only the present invention's a kind of embodiment wherein, and it describes comparatively concrete and detailed, but can not therefore be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.
Claims (6)
1. the medical image management system based on SVM, it is characterized in that: it comprises:
One raw video database, for storing medical image;
One classification image database, for the medical image according to the described raw video database of kinds of Diseases classification storage;
One feature recognizer, for the medical features of the medical image of identifying described raw video database;
One sorter, identified the medical image of described image database for utilizing described feature recognizer, obtains its medical features, then disease identified, and be stored in described classification image database according to kinds of Diseases.
2. the medical image management system based on SVM according to claim 1 is characterized in that: described feature recognizer utilizes based on Adaboost cascade classifier training in advance.
3. the medical image management system based on SVM according to claim 2 is characterized in that:
One master pattern storehouse, for storing the medical image of human organ master pattern;
One standard disease model storehouse, for the medical features of storage standards disease model.
4. the medical image management system based on SVM according to claim 3, it is characterized in that: the medical image of described master pattern library storage human organ master pattern is classified according to sex, age, ethnic group, height, the body weight of human body.
5. the medical image management system based on SVM according to claim 4, it is characterized in that: the step that described feature recognizer obtains the medical features of described medical image comprises:
1) image pre-service, to described medical image to be identified become a full member, convergent-divergent, filtering, resolution adjustment process;
2) the organic image standardization is adjusted, to described medical image to be identified be out of shape, local convergent-divergent processes, make organic image outline line wherein be tending towards overlapping with the outline line of medical image in described master pattern storehouse, the medical image in described master pattern storehouse screens and obtains in described master pattern storehouse;
3), in previous step, the outline line of the medical image in organic image outline line and described master pattern storehouse not intersection carries out the genius morbi extraction.
6. the medical image management system based on SVM according to claim 5 is characterized in that: the condition of the medical image described step 2) in screening described master pattern storehouse comprises: the sex of human body, age, ethnic group, height, body weight.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310431609.XA CN103488977A (en) | 2013-09-22 | 2013-09-22 | Medical image management system based on SVM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310431609.XA CN103488977A (en) | 2013-09-22 | 2013-09-22 | Medical image management system based on SVM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103488977A true CN103488977A (en) | 2014-01-01 |
Family
ID=49829186
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310431609.XA Pending CN103488977A (en) | 2013-09-22 | 2013-09-22 | Medical image management system based on SVM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103488977A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104392223A (en) * | 2014-12-05 | 2015-03-04 | 青岛科技大学 | Method for recognizing human postures in two-dimensional video images |
| CN106682435A (en) * | 2016-12-31 | 2017-05-17 | 西安百利信息科技有限公司 | System and method for automatically detecting lesions in medical image through multi-model fusion |
| CN108109680A (en) * | 2017-12-20 | 2018-06-01 | 南通艾思达智能科技有限公司 | A kind of method of settlement of insurance claim image bag sorting |
| CN110543894A (en) * | 2019-07-28 | 2019-12-06 | 聊城市光明医院 | Medical image processing method |
| CN111724893A (en) * | 2019-03-20 | 2020-09-29 | 宏碁股份有限公司 | Medical image recognition device and medical image recognition method |
| TWI711050B (en) * | 2019-03-12 | 2020-11-21 | 宏碁股份有限公司 | Medical image recognizition device and medical image recognizition method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110317892A1 (en) * | 2010-06-28 | 2011-12-29 | Ramot At Tel-Aviv University Ltd. | Method and system of classifying medical images |
| CN102799858A (en) * | 2012-06-21 | 2012-11-28 | 华中科技大学 | Automatic medical ultrasonic image recognition method based on redundant characteristic reduction |
-
2013
- 2013-09-22 CN CN201310431609.XA patent/CN103488977A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110317892A1 (en) * | 2010-06-28 | 2011-12-29 | Ramot At Tel-Aviv University Ltd. | Method and system of classifying medical images |
| CN102799858A (en) * | 2012-06-21 | 2012-11-28 | 华中科技大学 | Automatic medical ultrasonic image recognition method based on redundant characteristic reduction |
Non-Patent Citations (1)
| Title |
|---|
| 邢晓芬 等: "基于内容的医学图像分类研究", 《科学技术与工程》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104392223A (en) * | 2014-12-05 | 2015-03-04 | 青岛科技大学 | Method for recognizing human postures in two-dimensional video images |
| CN104392223B (en) * | 2014-12-05 | 2017-07-11 | 青岛科技大学 | Human posture recognition method in two-dimensional video image |
| CN106682435A (en) * | 2016-12-31 | 2017-05-17 | 西安百利信息科技有限公司 | System and method for automatically detecting lesions in medical image through multi-model fusion |
| WO2018120942A1 (en) * | 2016-12-31 | 2018-07-05 | 西安百利信息科技有限公司 | System and method for automatically detecting lesions in medical image by means of multi-model fusion |
| CN106682435B (en) * | 2016-12-31 | 2021-01-29 | 西安百利信息科技有限公司 | System and method for automatically detecting lesion in medical image through multi-model fusion |
| CN108109680A (en) * | 2017-12-20 | 2018-06-01 | 南通艾思达智能科技有限公司 | A kind of method of settlement of insurance claim image bag sorting |
| TWI711050B (en) * | 2019-03-12 | 2020-11-21 | 宏碁股份有限公司 | Medical image recognizition device and medical image recognizition method |
| CN111724893A (en) * | 2019-03-20 | 2020-09-29 | 宏碁股份有限公司 | Medical image recognition device and medical image recognition method |
| CN111724893B (en) * | 2019-03-20 | 2024-04-09 | 宏碁股份有限公司 | Medical image identification device and medical image identification method |
| CN110543894A (en) * | 2019-07-28 | 2019-12-06 | 聊城市光明医院 | Medical image processing method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xu et al. | Deep learning of feature representation with multiple instance learning for medical image analysis | |
| CN103488977A (en) | Medical image management system based on SVM | |
| Sun et al. | Automatic multi-fault recognition in TFDS based on convolutional neural network | |
| Wojek et al. | A dynamic conditional random field model for joint labeling of object and scene classes | |
| CN103473569A (en) | Medical image classification method based on SVM | |
| JP2017520859A (en) | Image object region recognition method and apparatus | |
| CN104361313B (en) | A kind of gesture identification method merged based on Multiple Kernel Learning heterogeneous characteristic | |
| CN103824052A (en) | Multilevel semantic feature-based face feature extraction method and recognition method | |
| CN103745233B (en) | The hyperspectral image classification method migrated based on spatial information | |
| WO2024001123A1 (en) | Image recognition method and apparatus based on neural network model, and terminal device | |
| Guo et al. | Real-time hand detection based on multi-stage HOG-SVM classifier | |
| CN105913034B (en) | Vehicle identification method and device and vehicle | |
| CN108509861B (en) | Target tracking method and device based on combination of sample learning and target detection | |
| CN103489000A (en) | Achieving method of human movement recognition training system | |
| Hussain et al. | A survey of traffic sign recognition systems based on convolutional neural networks | |
| CN115497124A (en) | Identity recognition method and device and storage medium | |
| Tan et al. | A survey of zero shot detection: methods and applications | |
| Manivannan et al. | Hep-2 specimen classification using multi-resolution local patterns and SVM | |
| Najibi et al. | Towards the success rate of one: Real-time unconstrained salient object detection | |
| Xie et al. | Vehicle detection in open parks using a convolutional neural network | |
| Ren et al. | Object detection using boosted local binaries | |
| Hasan et al. | Gesture recognition using modified HSV segmentation | |
| Yan et al. | Making full use of spatial-temporal interest points: an adaboost approach for action recognition | |
| CN105335755A (en) | Media segment-based speaking detection method and system | |
| Xu et al. | Car detection using deformable part models with composite features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140101 |
|
| RJ01 | Rejection of invention patent application after publication |