CN108829815B - Medical image screening method - Google Patents
Medical image screening method Download PDFInfo
- Publication number
- CN108829815B CN108829815B CN201810597945.4A CN201810597945A CN108829815B CN 108829815 B CN108829815 B CN 108829815B CN 201810597945 A CN201810597945 A CN 201810597945A CN 108829815 B CN108829815 B CN 108829815B
- Authority
- CN
- China
- Prior art keywords
- image
- images
- screening
- label
- labels
- 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
Images
Classifications
-
- 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
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
Abstract
The invention provides a medical image screening method, which comprises the steps of obtaining medical scanning images from an image scanning terminal and grouping the images; the image packets comprise phase packets; calling label items corresponding to the image contents in the corresponding groups, and marking all the images in the groups according to the labels; the labels and the images are in one-to-one correspondence, and each image corresponds to one label representing the characteristic identifier of the image; and generating an image index catalog according to the identified label. Compared with the prior art, the staff only need to relate whether the image sample meets the business condition, and do not need to put a great deal of effort in complicated work such as image copying, file directory building and the like, the average time of marking the image by the technician is 2-3 seconds, and the speed is multiplied.
Description
Technical Field
The invention relates to a medical image screening method, and relates to the field of medical image screening.
Background
The image samples used for the training of the artificial intelligent model of the medical image are screened, summarized and sorted in a manual sorting mode, namely, medical images are opened one by one in a computer folder by a technician and then the images meeting the conditions are copied into other files, so that the purposes of screening, classifying and sorting are achieved. However, this method has a very low efficiency, a high error rate, non-uniform classification of images (high probability of manual input errors), and even correction after errors are found is difficult, and even frustrated for technicians. The images screened in the way are used for model training, the result is difficult to achieve the expectation, the screened images need to be cleaned again for retraining, the repeated image screening prolongs the model training period, and a technical team is tired.
The invention is a technical scheme, a technician only needs to complete the marking of a model training image sample quickly, accurately and easily, and the system can automatically complete the classification, sorting and storage of marked images. Firstly, the image screening speed is increased by times, then the subsequent image filing is completely finished automatically by the system, thus completely avoiding errors caused by manual operation, the image can be directly read and trained by manual intelligence after filing, and finally, the technician responsible for image screening is easy and pleasant.
Disclosure of Invention
The invention aims to provide a medical image screening method which has the characteristics of rapidly screening hospital influence images and generating an image index catalog.
The technical scheme adopted by the invention is as follows:
a medical image screening method comprises screening a medical image,
acquiring medical scanning images from an image scanning terminal, and grouping the images; the image packets comprise phase packets;
calling a label item corresponding to the image content in the corresponding group, and marking all the images in the group according to the label; the labels and the images are in one-to-one correspondence, and each image corresponds to one label representing the characteristic identifier of the image;
and generating an image index catalog according to the identified label.
The image grouping also comprises scanning part grouping, and the scanning part grouping is carried out first, and then phase grouping is carried out.
The image grouping also comprises a scanning personnel name grouping, wherein after the personnel names are grouped, the scanning parts are grouped, and then the phase grouping is carried out.
The method also includes grouping the images according to image naming.
The method further comprises the steps of checking whether any of the label unused identification, the reused identification and the image unidentified condition occurs before generating the image index catalog, and prompting an error if the label unused identification, the reused identification and the image unidentified condition occur.
The method further comprises the steps of checking whether any of the label unused identification, the label reused identification and the image unidentified identification occurs before generating the image index catalog, and prompting an error and giving an error reason if the label unused identification, the label reused identification and the image unidentified identification occur.
The error reason comprises an error item, wherein the error item comprises any of a label unused identifier, a label reused identifier and an image unidentified identifier.
The error reasons further comprise any of extracting and presenting labels which are not identified by using, labels which are identified by reusing, images corresponding to the labels which are identified by reusing and images which are not identified.
The method further comprises the step of reading image data information with successful label identification as a sample to carry out model training.
The method also includes recording the whole process of image screening, reading image screening records in statistical analysis, and counting the image screening amount and screening efficiency.
Compared with the prior art, the staff only need to relate whether the image sample meets the business condition, and do not need to put a great deal of effort in complicated work such as image copying, file directory building and the like, the average time of marking the image by the technician is 2-3 seconds, and the speed is multiplied.
Drawings
Fig. 1 is a schematic structural diagram of a medical image screening system according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of an original data reading process according to an embodiment of the present invention.
FIG. 3 is a flowchart illustrating an embodiment of the present invention for screening and automatically archiving images to an artificial intelligence center.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Any feature disclosed in this specification (including any accompanying drawings) may be replaced by alternative features serving equivalent or similar purposes, unless expressly stated otherwise. That is, unless expressly stated otherwise, each feature is only an example of a generic series of equivalent or similar features.
A medical image screening method comprises screening a medical image,
acquiring medical scanning images from an image scanning terminal, and grouping the images; the image groupings include phase (phase of a scan, such as a scan phase of a CT scan) groupings;
calling label items corresponding to the image contents in the corresponding groups, and marking all the images in the groups according to the labels; the labels and the images are in one-to-one correspondence, and each image corresponds to one label representing the characteristic identifier of the image;
and generating an image index catalog according to the identified label.
As one specific example, in a medical image scan (e.g., a CT enhanced scan), a technician first selects a sequence of scans in the device based on the site under examination of the patient, and then begins the scan. The entire scan may go through 4 phases (4 phases, not all scans go through 4 phases, here four phases are just an example), scout, plateau, arterial and venous phases respectively. The simultaneously generated image records the patient's name, ID number, scan sequence name, location, session name, and scan layer thickness. The system displays the images in groups according to the storage path of the images and the naming rule of each image.
As an image grouping rule, the image grouping rule is not invariable, in the scheme of the invention, the most basic grouping rule is grouped according to the scanning period, the image grouping is completed according to the file name and is stored in the database, and the database is selected as the basis of screening.
If the scanning part grouping is included, such as abdomen, chest, head and the like, the scanning parts are firstly grouped and then are periodically grouped.
If the method comprises the scanning of the grouping of names of the personnel, such as Zhang III, Li IV, Wang V and the like, the grouping of the names of the personnel is firstly carried out, then the grouping of the scanning parts is carried out, and finally the grouping of the phase is carried out.
Before the image is labeled, a label item needs to be selected, and the label item can be selected manually by a technician or automatically according to obvious grouping characteristics. The label item is called from a label library which is generated in advance and used for calling and selecting.
For example: patient Xiaoming (XIAOMING) requires an abdominal CT enhancement scan with an ID of 00001.
Then the path of the small CT image save:
the first level folder name is: CT _ XIAO _ MING _ 00001;
the name of the secondary folder is as follows: ABDOMEN _ HX _ ABDOMEN _ C _ WJ _ (ADULT);
there are 4 folders in the tertiary folder and are respectively: topogram (for scout), Abdomen-7-0 (for abdominal scout phase, scan layer thickness 7 mm), Abdomen-A-7-0 (for abdominal arterial phase, scan layer thickness 7 mm), Abdomen-V-7-0 (for abdominal venous phase, scan layer thickness 7 mm) wherein Abdomen indicates "Abdomen", A indicates "artery", V indicates "vein", and 7-0 indicates "scan layer thickness of 7.0 mm".
The number of CT images is:
there are 2 images under the Topogram folder;
200 images are arranged under an Abdomen-7-0 folder;
200 images are arranged under an Abdomen-A-7-0 folder;
200 images are placed under the Abdomen-V-7-0 folder.
Dividing the Xiaoming CT images into 4 groups of images according to the naming rule and displaying the images respectively, simultaneously marking the scanning part of the image as the abdomen and the layer thickness as 7.0 mm, and after the technician finishes screening the Xiaoming 4 groups of CT images, the system loads 4 groups of images of the next patient.
By the imported image folder name, a site to be examined by the patient, such as the abdomen, is obtained. The system displays corresponding labels beside each image of different phases according to the inspection part.
For example: the abdominal label is in different phases:
topogram (scout): foreign matter in the positioning image, no lifting of hands in the positioning image, no drinking of water in the positioning image, and positioning of the patient in the center of the scanning field are avoided;
abdomen-7-0 (plateau): upper liver marginal layer, upper spleen marginal layer, lower stomach marginal layer, lower spleen marginal layer, lower gallbladder marginal layer, lower left kidney marginal critical layer, lower right kidney marginal critical layer, motion artifact, radial artifact, primary liver cancer, liver metastatic cancer, other pathological changes of liver, gallbladder lesion, pancreatic lesion, stomach lesion, spleen lesion, abdominal lesion, kidney lesion, and other pathological changes;
Abdomen-A-7-0 (arterial phase): upper liver margin critical layer, upper spleen margin critical layer, lower stomach margin critical layer, lower spleen margin critical layer, lower gall bladder margin critical layer, lower left kidney margin critical layer, lower right kidney margin critical layer, early arterial stage, late arterial stage, motion artifact, radial artifact, primary liver cancer, liver metastatic cancer, other liver lesions, gall bladder lesion, pancreatic lesion, stomach lesion, spleen lesion, abdominal lesion, kidney lesion, other lesions;
Abdomen-V-7-0 (venous phase): upper liver margin critical layer, upper spleen margin critical layer, lower stomach margin critical layer, lower spleen margin critical layer, lower gall bladder margin critical layer, lower left kidney margin critical layer, lower right kidney margin critical layer, early venous stage, late venous stage, motion artifact, radial artifact, primary liver cancer, liver metastatic cancer, other liver lesions, gall bladder lesion, pancreatic lesion, stomach lesion, spleen lesion, abdominal lesion, kidney lesion, other lesions;
the technician selects the corresponding label based on the characteristics of each image. Each image corresponds to one label, and the situation that one image corresponds to a plurality of labels cannot occur. As an embodiment of the present invention, the technician selects one label and the label turns blue, and the remaining labels automatically turn gray. If the label is to be changed and a new label is clicked directly, the previously selected label will change to grey and the new label to blue.
For example: the technician now selects an image in Abdomen-V-7-0 (venous phase) as "suprahepatic margin", the corresponding other label for that image will become gray, and the label "suprahepatic margin" will become blue indicating that the image is labeled as "suprahepatic margin".
After the image screening is completed, the database will finish archiving and forming an index. The artificial intelligence training can be directly used, and can also be applied to other systems.
In the scheme of the invention, a worker only needs to relate whether the image sample meets the service condition, and does not need to put a great deal of energy in complicated work such as image copying, file directory building and the like, the average time of marking the image by a technician is 2-3 seconds, and the speed is multiplied. The system can complete the classification, arrangement and storage of the marked images, completely avoid errors caused by manual operation, and finally make technicians responsible for image screening feel comfortable. In addition, the screened images can be directly used for artificial intelligence training and can help hospitals to store standard images.
The method further comprises the steps of checking whether any of the label unused identification, the reused identification and the image unidentified condition occurs before generating the image index catalog, and prompting an error if the label unused identification, the reused identification and the image unidentified condition occur.
As an embodiment of the invention, after the technician selects a group of images to be screened, the system will determine whether the screened images have errors according to the phase in which the group of images is located. If not, the submission is successful and the system automatically loads the next set of images. If there is an error, the technician can correct it in time. Thus, the technician only needs to pay attention to whether the image per se meets the selection standard, and does not need to waste a great deal of energy in the tedious work of image copying, file directory establishment and the like. A skilled technician can complete the labeling of an image every 2-3 seconds.
For example: in the abdominal CT enhancement scan, in each of three groups of images in the horizontal scan stage, the arterial stage and the venous stage, each of 7 labels of the suprahepatic margin critical layer, the superior splenic margin critical layer, the inferior gastric margin critical layer, the inferior splenic margin critical layer, the inferior gall bladder critical layer, the inferior left kidney margin critical layer and the inferior right kidney margin critical layer can be labeled only once (the 7 labels in each group of images can only correspond to 7 different images), if the number of the images is not enough for 7, the system can prompt a technician that the label is not used (for example, only 6 images are labeled, but the 'superior hepatic margin critical layer' is not used, the system prompts the technician that the 'superior hepatic margin critical layer' has no corresponding image), and the technician can supplement in time, so that the omission is avoided. If a label is marked on multiple images, the system will prompt the technician that the label is reused, leaving only one (e.g., if 3 images are marked as "suprahepatic margin", then the system will prompt the technician that 3 images are marked as "suprahepatic margin", please leave one, while displaying the 3 images together, the technician simply cancels two images on the 3 images).
The method further comprises the steps of checking whether any of the label unused identification, the label reused identification and the image unidentified identification occurs before generating the image index catalog, and prompting an error and giving an error reason if the label unused identification, the label reused identification and the image unidentified identification occur.
As an embodiment of the present invention, the image data information obtained by the marking is saved as a result of the current screening, and the result of the current screening becomes one of the screened image databases.
As an embodiment of the invention, the error cause comprises an error item, and the error item comprises any of a label unused identifier, a label reused identifier and an image unidentified identifier.
The error reasons further comprise any of extracting and presenting labels which are not identified by using, labels which are identified by reusing, images corresponding to the labels which are identified by reusing and images which are not identified.
As an implementation mode of the invention, when the image data is submitted to the server side, the program compares the marked image data with each group of labels, and if the label is not used, the program compares the marked image data with each group of labels. The program screens the image data to form an unlabeled image database; if there are multiple image data to which the same tag belongs at the same time, the program will filter the partial image data to form a repeated marked image database. As one embodiment, image information in the unlabeled image database and the unlabeled database is extracted and presented to the user. And the user completes the characteristic identification step by step, newly generated image data is updated to a graphic database, the screened data item file is submitted to a server side in an HTTP mode, and the server side evaluates the screened image data and feeds the result back to the user.
The method further comprises the step of reading image data information with successful label identification as a sample to carry out model training.
The method also includes recording the whole process of image screening, reading image screening records in statistical analysis, and counting the image screening amount and screening efficiency. In one embodiment of the invention, the system reads the image screening record of each technician in the statistical analysis, and calculates the image screening amount and screening efficiency of the technicians.
The medical image screening system adopted by the invention, as shown in fig. 1, comprises a display layer, a business layer and a data processing layer.
The display layer comprises a user interaction interface and is used for receiving the feature identification input by the user and the appointed label mark and displaying the image data screened in each step and the corresponding label information at the user side.
The business layer comprises a screening data generation module, a label generation module, an associated information combination module, a data verification module and an image screening flow control module.
A screening data generation module: grouping the data to be selected as a basic selection database for the screening;
as an example: the system now reads CT examination images of 10 patients simultaneously, and the system will divide the 10 patient images into 10 large groups, each under a scout, sweep, arterial, venous, 4 groups of images, for a total of 40 images. The image screening by the technician for labeling the images will be performed on the basis of the 40 sets of images, and then the 40 sets of images are the base selected database for the screening.
A tag generation module: feature identification for user screening of image data. The system obtains the body part currently examined by the patient through image naming, and then different labels are generated according to different phases.
As an example: the system was obtained by naming the image for a small CT examination, the examined region being the abdomen. Then a label corresponding to the abdomen will be displayed in each set of images that are small and clear, such as: upper hepatic margin critical layer, upper spleen margin critical layer, lower gastric margin critical layer, lower spleen margin critical layer, lower gallbladder margin critical layer, lower left kidney margin critical layer, lower right kidney margin critical layer, etc.
The associated information combination module: and establishing a corresponding relation between the image and the label, and meanwhile, grouping the image according to the label to form an image group based on the label.
A data checking module: and the method is used for verifying the corresponding relation between the screened image data and the label. And checking whether the corresponding relation between the image and the label is correct and the use condition of the label according to different parts of the patient to be checked.
As an example: the system obtains the images through the naming of the images of the small clear CT examination, the examined part is the abdomen, and then the system checks whether 3 groups of images in the CT images have images of each organ edge (such as the upper liver edge critical layer, the upper spleen edge critical layer, the lower stomach edge critical layer and the like) and whether the label of each organ edge is used for multiple times.
The image screening flow control module: the system is used for coordinating and controlling the work among all modules of the service layer and is also responsible for the initialization when the system is started and the cache cleaning work when the system is quitted;
as an example: the business process of the controlling technician in the system is controlled by the module. When the technician enters the system, the system displays the data according to the technician authority. The technician must navigate through the images when screening the images, the system completes the image recognition and grouping, the system presents different labels based on the image information, the technician selects a label for each image, the system verifies if correct after completing the image tagging, if incorrect, the system prompts the technician to modify until correct, the next set of images is loaded, and the technician clears the cache when exiting. The above steps are handled by the process control module, and the technician only completes the image screening strictly according to the process.
Artificial intelligence butt joint module: the screened data has the conditions for participating in artificial intelligence training, and the artificial intelligence program reads and finishes the screened image data and the labels through the interface of the module.
The data processing layer comprises data access and storage and is used for accessing the selected image database and storing the image data information obtained by screening as the database of the screening result.
As shown in fig. 2 and 3, the system analyzes and stores various tags in the EXCEL table into the system by setting a tag read path. And analyzing and storing all the images to be screened into the system by setting an image reading path. The technician enters the image tagging page and completes tagging of each image. After the marking is finished, the system automatically stores the images and the corresponding labels in a grouping mode, establishes indexes, and then directly calls the screened images to perform model training.
Claims (9)
1. A medical image screening method comprises screening a medical image,
acquiring medical scanning images from an image scanning terminal, and grouping the images; the image packets comprise phase packets; obtaining the body part of the current patient to be examined through image naming, and then generating different labels according to different phases;
calling label items corresponding to the image contents in the corresponding groups, and marking all the images in the groups according to the labels; the labels and the images are in one-to-one correspondence, and each image corresponds to one label representing the characteristic identifier of the image;
generating an image index catalog according to the identified label;
before generating the image index catalog, checking whether any of the label unused identification, the reused identification and the image unidentified condition occurs, and if so, prompting an error;
comparing the marked image data with each group of labels, and screening the image data to form an unmarked image database if the labels are not used; if the same label belongs to a plurality of image data at the same time, screening the partial image data to form a repeatedly marked image database; extracting image information in the unmarked image database and the repeated marking database, and presenting the image information to a user; and the user completes the characteristic identification step by step, newly generated image data is updated to an image database, the screened data item file is submitted to a server side in an HTTP mode, and the server side evaluates the screened image data and feeds the result back to the user.
2. The method of claim 1, wherein the image grouping further comprises a scan portion grouping, and the scan portion grouping is performed first and then a phase grouping is performed.
3. The method of claim 2, wherein the image grouping further comprises a group of names of scan personnel, wherein the group of names of scan personnel is first performed, then the group of scan parts is performed, and then the group of phases is performed.
4. The method for screening medical image according to one of claims 1 to 3, further comprising grouping images according to image names.
5. The method for screening medical image according to claim 1, further comprising, before generating the image index directory, checking whether any of the label unused identifier, the label reused identifier and the image unidentified identifier occurs, and if so, prompting an error and giving a reason for the error.
6. The method for screening medical image according to claim 5, wherein the error cause includes error items, and the error items include any of the labeled unused identifier, the labeled reused identifier, and the labeled image unidentified.
7. The method for screening medical image according to claim 6, wherein the error cause further comprises extracting and presenting any of labels of unused marks, labels of reused marks, images corresponding to labels of reused marks, and images of unidentified marks.
8. The method for screening medical image according to claim 1, further comprising reading the image data information with successfully labeled tag as a sample for model training.
9. The method for screening medical image according to claim 1, further comprising recording the whole process of screening the image, reading the image screening record in the statistical analysis, and counting the image screening amount and the screening efficiency.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810597945.4A CN108829815B (en) | 2018-06-12 | 2018-06-12 | Medical image screening method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810597945.4A CN108829815B (en) | 2018-06-12 | 2018-06-12 | Medical image screening method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108829815A CN108829815A (en) | 2018-11-16 |
| CN108829815B true CN108829815B (en) | 2022-06-07 |
Family
ID=64143537
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810597945.4A Active CN108829815B (en) | 2018-06-12 | 2018-06-12 | Medical image screening method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108829815B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110349654B (en) * | 2019-05-31 | 2023-10-20 | 平安科技（深圳）有限公司 | Image processing method, device, storage medium and computer equipment |
| CN110176294A (en) * | 2019-05-31 | 2019-08-27 | 数坤（北京）网络科技有限公司 | A kind of dispatching method, device and the readable storage medium storing program for executing of blood vessel CTA image data |
| CN110689958A (en) * | 2019-09-09 | 2020-01-14 | 杭州憶盛医疗科技有限公司 | Cancer pathology auxiliary diagnosis method based on artificial intelligence technology |
| CN111430010A (en) * | 2020-03-30 | 2020-07-17 | 王博 | System and method for deducing scanning sequence phase based on DICOM image information |
| CN111598883B (en) * | 2020-05-20 | 2023-05-26 | 重庆工程职业技术学院 | Calibration label equipment for acquiring cloud data medical images and working method |
| CN111931762B (en) * | 2020-09-25 | 2021-07-30 | 广州佰锐网络科技有限公司 | AI-based image recognition solution method, device and readable storage medium |
| CN112102945B (en) * | 2020-11-09 | 2021-02-05 | 电子科技大学 | Device for predicting severe condition of COVID-19 patient |
| CN113130067B (en) * | 2021-04-01 | 2022-11-08 | 上海市第一人民医院 | Intelligent reminding method for ultrasonic examination based on artificial intelligence |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0947937A2 (en) * | 1998-04-02 | 1999-10-06 | Canon Kabushiki Kaisha | Image search apparatus and method |
| CN101421746A (en) * | 2006-04-11 | 2009-04-29 | 索尼株式会社 | Image classification based on a mixture of elliptical color models |
| CN105045886A (en) * | 2015-07-23 | 2015-11-11 | 青岛海信医疗设备股份有限公司 | Importing method of DICOM (Digital Imaging and Communications in Medicine) image |
| CN105518679A (en) * | 2015-03-26 | 2016-04-20 | 北京旷视科技有限公司 | Image management method and image synchronization method |
| CN107229826A (en) * | 2017-05-23 | 2017-10-03 | 深圳市菲森科技有限公司 | A kind of correction Image Management apparatus and method for orthodontic |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135952A (en) * | 2011-03-16 | 2011-07-27 | 华为终端有限公司 | Electronic file labeling method and device |
| CN102546755A (en) * | 2011-12-12 | 2012-07-04 | 华中科技大学 | Data storage method of cloud storage system |
| CN103207879B (en) * | 2012-01-17 | 2016-03-30 | 阿里巴巴集团控股有限公司 | The generation method and apparatus of image index |
| CN103150379A (en) * | 2013-03-13 | 2013-06-12 | 北京东田教育科技有限公司 | Indexed management method for message subdirectory |
| JP2015087903A (en) * | 2013-10-30 | 2015-05-07 | ソニー株式会社 | Apparatus and method for information processing |
| CN103995889B (en) * | 2014-06-03 | 2017-11-03 | 广东欧珀移动通信有限公司 | Picture classification method and device |
| CN104572905B (en) * | 2014-12-26 | 2018-09-04 | 小米科技有限责任公司 | Print reference creation method, photo searching method and device |
| CN106649610A (en) * | 2016-11-29 | 2017-05-10 | 北京智能管家科技有限公司 | Image labeling method and apparatus |
| CN107292117B (en) * | 2017-07-14 | 2021-03-16 | 中国科学院上海技术物理研究所 | Processing method and device for quality guarantee of mass shared medical images |
| CN107622104B (en) * | 2017-09-11 | 2020-03-06 | 中央民族大学 | Character image identification and marking method and system |
| CN107871011B (en) * | 2017-11-21 | 2020-04-24 | Oppo广东移动通信有限公司 | Image processing method, image processing device, mobile terminal and computer readable storage medium |
-
2018
- 2018-06-12 CN CN201810597945.4A patent/CN108829815B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0947937A2 (en) * | 1998-04-02 | 1999-10-06 | Canon Kabushiki Kaisha | Image search apparatus and method |
| CN101421746A (en) * | 2006-04-11 | 2009-04-29 | 索尼株式会社 | Image classification based on a mixture of elliptical color models |
| CN105518679A (en) * | 2015-03-26 | 2016-04-20 | 北京旷视科技有限公司 | Image management method and image synchronization method |
| CN105045886A (en) * | 2015-07-23 | 2015-11-11 | 青岛海信医疗设备股份有限公司 | Importing method of DICOM (Digital Imaging and Communications in Medicine) image |
| CN107229826A (en) * | 2017-05-23 | 2017-10-03 | 深圳市菲森科技有限公司 | A kind of correction Image Management apparatus and method for orthodontic |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108829815A (en) | 2018-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108829815B (en) | Medical image screening method | |
| CN1615489B (en) | Image reporting method and system | |
| JP5328146B2 (en) | Medical image processing apparatus, medical image processing method and program | |
| CN111027294B (en) | Method, device and system for summarizing table | |
| Long et al. | Content-based image retrieval in medicine: retrospective assessment, state of the art, and future directions | |
| US20130251233A1 (en) | Method for creating a report from radiological images using electronic report templates | |
| CN101896939A (en) | Labeling a segmented object | |
| US20090106047A1 (en) | Integrated solution for diagnostic reading and reporting | |
| CN102667786A (en) | Protocol guided imaging procedure | |
| CN112331348B (en) | Analysis method and system for set marking, data, project management and non-programming modeling | |
| CN111476776B (en) | Chest lesion position determination method, system, readable storage medium and apparatus | |
| CN111627531A (en) | Medical image classification processing system based on artificial intelligence | |
| JP2005537101A (en) | Patient identification method for X-ray film user interface | |
| CN109284702A (en) | A kind of answer sheet based on image model to point and marking system | |
| CN111368929B (en) | Picture marking method | |
| CN113902684A (en) | Image segmentation method and device, electronic device and storage medium | |
| CN113972009A (en) | Medical examination consultation system based on clinical examination medical big data | |
| Mononen et al. | DigiWeb-a workflow environment for quality assurance of transcription in digitization of natural history collections | |
| CN113254752A (en) | Lesson preparation method and device based on big data and storage medium | |
| Lima et al. | patologIA: online platform for improving image-based diagnostics applied to medical students | |
| CN111126502B (en) | DICOM medical image sequence classification method for artificial intelligence auxiliary diagnosis | |
| Antani et al. | Medical validation and CBIR of spine x-ray images over the Internet | |
| EP4354452A1 (en) | Medical image search and retrieval | |
| US20100121846A1 (en) | Filter by example | |
| CN111462913B (en) | Automatic segmentation method and device for disease diagnosis in case document |
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 |