CN114003745A - Medical image labeling multi-version management method - Google Patents
Medical image labeling multi-version management method Download PDFInfo
- Publication number
- CN114003745A CN114003745A CN202111189659.2A CN202111189659A CN114003745A CN 114003745 A CN114003745 A CN 114003745A CN 202111189659 A CN202111189659 A CN 202111189659A CN 114003745 A CN114003745 A CN 114003745A
- Authority
- CN
- China
- Prior art keywords
- labeling
- tree object
- object file
- medical image
- generating
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/51—Indexing; Data structures therefor; Storage structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/55—Clustering; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Document Processing Apparatus (AREA)
Abstract
The invention discloses a medical image labeling multi-version management method, which comprises the following steps: receiving a first medical image data set uploaded by a first user; performing full-process labeling processing on a first medical image data set, converting data generated after each process node is completed into tree object files, calculating the storage address and the file content of each tree object file by combining a preset SHA-1 algorithm, generating one or more hash values corresponding to each process node, and configuring a corresponding version number for each hash value; and according to the input operation instruction, performing version management on the first medical image data set by combining the hash value and the version number corresponding to each process node. The medical image labeling method stores the medical image labeling process, and can trace back to any version in the labeling process through the hash value serving as the unique identifier, so that the problem is accurately positioned, and the loss of the image labeling record is avoided.
Description
Technical Field
The invention relates to the field of medical image annotation, in particular to a medical image annotation multi-version management method.
Background
In medicine, clinical diagnosis does not intuitively determine diseases, but obtains effective labeling data of focuses and related organs by means of labeled medical images, and jointly determines the diseases and the severity of patients by combining preliminary diagnosis of doctors and inquiry information of the patients. The medical image effectively labeled provides important quantitative information for doctors in the aspects of disease diagnosis, disease condition evaluation, development trend prediction, treatment strategy making and the like.
At present, a large number of data sets judged to be correct labeled information are reserved as data evidences in AI auxiliary diagnosis, a large number of data sets judged to be wrong labeled information are reserved as subject bases in teaching and training, data sources and labeled information between the data sets and the labeled information are split, and only one type of available data sets are reserved in version management of a labeling system.
Most data set labeling version management systems retain a labeling result data set through version control or version iteration management, cannot store revisions of all files, and lack the function that the version management system can conveniently look up revisions of specific versions. The version control system Git manages all files in the library through the version library, including various states such as modification, deletion and the like, but lacks a function that a user can select to share part of codes of a project and can set the progress improvement condition and resource management authority of the whole team. The data crowdsourcing system distributes projects through mass data acquisition, intelligently retrieves the projects before the projects are marked by staff, intelligently audits the projects after the staff are marked, and finally audits the data set marking results manually to generate a data set marking result and finish the process, but does not manage the data set marking version formed by a plurality of projects. Data annotation software for image algorithms, such as LabelMe, Vott and the like, mostly uses a client, and although a plurality of tasks are simultaneously annotated and a new annotation task can be added through configuration, multi-version management formed by a plurality of task sets is lacked.
Disclosure of Invention
The invention provides a medical image labeling multi-version management method, which aims to solve the technical problems that the labeling record is easy to lose, the labeling efficiency is low and a version management system cannot look up the revision progress condition of a specific version in the conventional medical image labeling version management method.
In order to solve the above technical problem, an embodiment of the present invention provides a method for multi-version management of medical image annotation, including:
receiving a first medical image data set uploaded by a first user; wherein the first medical image data set comprises a plurality of medical images to be annotated;
performing full-process labeling processing on the first medical image data set, converting data generated after each process node is completed into tree object files, calculating the storage address and the file content of each tree object file by combining a preset SHA-1 algorithm, generating one or more hash values corresponding to each process node, and configuring a corresponding version number for each hash value; wherein, the full-process labeling process comprises the following process nodes: the system comprises an initial data storage node, an annotation task dividing node, an annotation result uploading node, an annotation acceptance node, an annotation result summarizing node and a project progress updating node;
and according to the input operation instruction, combining the hash value and the version number corresponding to each process node, and performing version management on the first medical image data set.
Further, when the full-flow labeling processing is performed at the initial data storage node, the processing operations executed are specifically:
receiving a first medical image data set uploaded by a first user, and converting the first medical image data set into a first binary large object;
calculating the storage address and the file content of the first binary large object by combining the SHA-1 algorithm, generating an initial dataset hash value corresponding to the first binary large object, and setting an initial dataset version number corresponding to the first binary large object;
establishing a plurality of different project branches according to a first medical image data set uploaded by a first user and a preset classification mode, and generating a corresponding first tree object file according to all the project branches;
calculating the storage address and the file content of the first tree object file through a preset SHA-1 algorithm, generating an initial item hash value corresponding to the first tree object file, and setting an initial item version number corresponding to the first tree object file.
Further, when the full-flow annotation processing is at the annotation task partition node, the processing operation executed is specifically:
generating a plurality of labeling tasks corresponding to the first medical image data set according to a preset labeling task division rule, and generating a corresponding second tree object file according to all labeling tasks; wherein, one labeling task corresponds to one second tree object file;
and calculating the storage address and the file content of the second tree object file by combining the SHA-1 algorithm, generating an initial task hash value corresponding to the second tree object file, and setting an initial task version number corresponding to the second tree object file.
Further, when the full-flow annotation processing is at the annotation result upload node, the processing operation executed is specifically:
responding to the task completion operation of a plurality of users on any one same labeling task, respectively receiving the labeling results of the plurality of users, storing the labeling results as a plurality of corresponding JSON files, and generating a plurality of corresponding third tree object files; the marking result of one user corresponds to one JSON file, and one JSON file corresponds to one third tree object file;
and respectively calculating the storage address and the file content of each third tree object file by combining the SHA-1 algorithm, generating a personal labeling result hash value corresponding to each third tree object file, and respectively setting the personal labeling result version number corresponding to each third tree object file.
Further, when the full-flow annotation processing is at the annotation acceptance node, the processing operation executed is specifically:
according to the JSON files corresponding to the marking results of the users, performing content comparison on the marking results of the users, and judging whether the marking results of the users are consistent;
if the labeling results are consistent, generating a fourth tree object file according to the labeling result of any one labeling task; wherein, the labeling result of one labeling task corresponds to one fourth tree object file;
respectively calculating the storage address and the file content of each fourth tree object file by combining the SHA-1 algorithm, generating a first acceptance result hash value corresponding to each fourth tree object file, and respectively setting a first acceptance result version number corresponding to each fourth tree object file;
and if not, generating a corresponding acceptance version according to the auditing result of the labeling committee.
Further, the generating of the corresponding acceptance version according to the review result of the labeling committee specifically includes:
receiving an auditing result of an annotation committee;
if the verification results are consistent, generating corresponding seventh tree object files according to the labeling results of any one same labeling task, respectively calculating the storage address and the file content of each seventh tree object file by combining the SHA-1 algorithm, generating a second acceptance result hash value corresponding to each seventh tree object file, and respectively setting a second acceptance result version number corresponding to each seventh tree object file;
and if the audit results are inconsistent, generating corresponding eighth tree object files according to the audit results of the labeling committees, respectively calculating the storage address and the file content of each eighth tree object file by combining the SHA-1 algorithm, generating a committee audit result hash value corresponding to each eighth tree object file, respectively setting a committee audit result version number corresponding to each eighth tree object file, and then generating a corresponding acceptance version according to the labeling results of the labeling experts.
Further, the generating of the corresponding acceptance version according to the labeling result of the labeling expert specifically includes:
receiving the labeling result of the labeling expert and generating a corresponding ninth tree object file;
and respectively calculating the storage address and the file content of each ninth tree object file by combining the SHA-1 algorithm, generating a third acceptance result hash value corresponding to each ninth tree object file, and respectively setting a third acceptance result version number corresponding to each ninth tree object file.
Further, when the full-flow annotation processing is at the annotation result summary node, the processing operation executed is specifically:
responding to task completion operation of each user on the plurality of labeling tasks, respectively receiving a plurality of labeling results corresponding to each labeling task, and generating a plurality of fifth tree object files; wherein, a plurality of marking results of one marking task correspond to a fifth tree object file;
and respectively calculating the storage address and the file content of each fifth tree object file by combining the SHA-1 algorithm, generating a task marking result hash value corresponding to each fifth tree object file, and respectively setting the task marking result version number corresponding to each fifth tree object file.
Further, when the full-flow annotation processing is at the project progress update node, the processing operation executed is specifically:
responding to task completion operation of each user on the plurality of labeling tasks, and generating a corresponding sixth tree object file according to a project for completing any one labeling task; wherein, the project which completes one labeling task corresponds to a sixth tree object file;
and respectively calculating the storage address and the file content of each sixth tree object file by combining the SHA-1 algorithm, generating a project progress hash value corresponding to each sixth tree object file, and respectively setting a project progress version number corresponding to each sixth tree object file.
Further, the version management of the first medical image data set specifically includes:
determining a first flow node to be managed according to an input operation instruction;
inquiring a hash value corresponding to the first process node, and determining a version number to be managed of the first medical image data set according to the inquired hash value;
and calling the medical image data of the first medical image data set under the version number to be managed according to the version number to be managed.
Compared with the prior art, the embodiment of the invention has the following beneficial effects:
the invention provides a medical image labeling multi-version management method, which is characterized in that a unique hash value and a version number are generated through an SHA-1 algorithm to identify a medical image data set and a plurality of items, tasks and other objects corresponding to the medical image data set, so that a multi-version system is formed. The unique hash value and the version number which are used as the identification are utilized to realize the control and management of the medical image data set and a plurality of projects and tasks corresponding to the medical image data set, so that the image labeling process can trace back to any version in the modification process through the identification, the modification details of the file are compared, and a modifier of the file is checked, thereby accurately positioning and finding out the reasons of various problems, simultaneously avoiding the loss of image labeling records, improving the image labeling efficiency, and providing a high-quality data base for the innovation driving development of 'AI + medical treatment'.
Drawings
FIG. 1: a flow diagram of a method for labeling multi-version management of a medical image according to an embodiment of the present invention;
FIG. 2: the structural schematic diagram of the full-process annotation processing of the medical image annotation multi-version management method provided by the embodiment of the invention is shown;
FIG. 3: the method for managing the medical image labeling multiple versions is a flow diagram for performing full-life-cycle version control management on a medical image labeling process;
FIG. 4: the method for medical image annotation multi-version management provided by the embodiment of the invention is a flow diagram for checking and accepting the personal annotation result;
FIG. 5: the method for labeling multiple versions of medical images provided by an embodiment of the present invention is a flowchart illustrating version control management of a medical image data set.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The first embodiment is as follows:
referring to fig. 1, a method for multi-version management of medical image annotation provided by an embodiment of the present invention includes:
s101: receiving a first medical image data set uploaded by a first user; wherein the first medical image data set comprises a number of medical images to be annotated.
S102: performing full-process labeling processing on a first medical image data set, converting data generated after each process node is completed into tree object files, calculating the storage address and the file content of each tree object file by combining a preset SHA-1 algorithm, generating one or more hash values corresponding to each process node, and configuring a corresponding version number for each hash value; referring to fig. 2, the full process labeling process includes the following process nodes: the system comprises an initial data storage node, an annotation task dividing node, an annotation result uploading node, an annotation acceptance node, an annotation result summarizing node and a project progress updating node.
S103: and according to the input operation instruction, performing version management on the first medical image data set by combining the hash value and the version number corresponding to each process node.
In this embodiment, the operation procedure for performing the full-procedure labeling processing on the first medical image data set specifically includes:
determining a first flow node to be managed according to an input operation instruction;
inquiring a hash value corresponding to the first process node, and determining a version number to be managed of the first medical image data set according to the inquired hash value;
and calling the medical image data of the first medical image data set under the version number to be managed according to the version number to be managed.
Referring to fig. 3, when the full-flow labeling processing is performed at the initial data storage node, the processing operations executed are specifically:
s201: receiving a first medical image data set uploaded by a first user, and converting the first medical image data set into a first binary large object;
and calculating the storage address and the file content of the first binary large object by combining an SHA-1 algorithm, generating an initial data set hash value corresponding to the first binary large object, and setting an initial data set version number corresponding to the first binary large object.
S202: establishing a plurality of different project branches according to a first medical image data set uploaded by a first user and a preset classification mode, and generating a corresponding first tree object file according to all the project branches;
and calculating the storage address and the file content of the first tree object file through a preset SHA-1 algorithm, generating an initial item hash value corresponding to the first tree object file, and setting an initial item version number corresponding to the first tree object file.
Referring to fig. 3, when the full-flow annotation process is at the annotation task partition node, the processing operations executed are specifically:
s203: generating a plurality of labeling tasks corresponding to the first medical image data set according to a preset labeling task division rule, and generating a corresponding second tree object file according to all the labeling tasks; wherein, one labeling task corresponds to one second tree object file;
and calculating the storage address and the file content of the second tree object file by combining an SHA-1 algorithm, generating an initial task hash value corresponding to the second tree object file, and setting an initial task version number corresponding to the second tree object file.
S204: and receiving and responding to the auditing results of the ethics committee.
In this embodiment, if the result of the audit is rejected, S205 is executed, and if the result of the audit is approved, S206 is executed.
S205: and ending a plurality of labeling tasks corresponding to the first medical image data set.
Referring to fig. 3, when the full-flow annotation process is at the annotation result upload node, the processing operations executed are specifically:
s206: responding to the completion operation of a plurality of users on any task of the same labeling task, respectively receiving the labeling results of the plurality of users, storing the labeling results as a plurality of corresponding JSON files, and generating a plurality of corresponding third tree object files; the marking result of one user corresponds to one JSON file, and one JSON file corresponds to one third tree object file;
and respectively calculating the storage address and the file content of each third tree object file by combining an SHA-1 algorithm, generating a personal labeling result hash value corresponding to each third tree object file, and respectively setting the personal labeling result version number corresponding to each third tree object file.
Referring to fig. 3, when the full-flow annotation process is at the annotation acceptance node, the processing operations executed are specifically:
s207: and (3) checking and accepting the labeling results of a plurality of users for any one labeling task, and generating a corresponding checking and accepting version, please refer to fig. 4, wherein the specific operation is as follows:
s301: and comparing the contents of the marking results of the users according to the JSON files corresponding to the marking results of the users, and judging whether the marking results of the users are consistent.
In this embodiment, if the annotation results of the multiple users are consistent, step S302 is executed, and if the annotation results of the multiple users are not consistent, step S303 is executed.
S302: generating a fourth tree object file according to the labeling result of any one labeling task; wherein, the labeling result of one labeling task corresponds to one fourth tree object file;
and respectively calculating the storage address and the file content of each fourth tree object file by combining an SHA-1 algorithm, generating a first acceptance result hash value corresponding to each fourth tree object file, and respectively setting a first acceptance result version number corresponding to each fourth tree object file.
S303: and receiving and responding to the auditing result of the labeling committee.
In this embodiment, if the audit results are consistent, S304 is executed, and if the audit results are inconsistent, S305 is executed.
S304: and generating corresponding seventh tree object files according to the labeling results of any one same labeling task, respectively calculating the storage address and the file content of each seventh tree object file by combining an SHA-1 algorithm, generating a second acceptance result hash value corresponding to each seventh tree object file, and respectively setting the second acceptance result version number corresponding to each seventh tree object file.
S305: and generating corresponding eighth tree object files according to the auditing results of the labeling committee, respectively calculating the storage address and the file content of each eighth tree object file by combining an SHA-1 algorithm, generating committee auditing result hash values corresponding to each eighth tree object file, and respectively setting committee auditing result version numbers corresponding to each eighth tree object file.
S306: receiving the labeling result of the labeling expert and generating a corresponding ninth tree object file;
and respectively calculating the storage address and the file content of each ninth tree object file by combining an SHA-1 algorithm, generating a third acceptance result hash value corresponding to each ninth tree object file, and respectively setting a third acceptance result version number corresponding to each ninth tree object file.
Referring to fig. 3, when the full-flow annotation processing is at the annotation result summary node, the processing operations executed are specifically:
s208: responding to task completion operation of each user on a plurality of labeling tasks, respectively receiving a plurality of labeling results corresponding to each labeling task, and generating a plurality of fifth tree object files; wherein, a plurality of marking results of one marking task correspond to a fifth tree object file;
and respectively calculating the storage address and the file content of each fifth tree object file by combining an SHA-1 algorithm, generating a task marking result hash value corresponding to each fifth tree object file, and respectively setting the task marking result version number corresponding to each fifth tree object file.
Referring to fig. 3, when the full-flow annotation process is at the project progress update node, the processing operations executed are specifically:
s209: responding to task completion operation of each user on a plurality of labeling tasks, and generating a corresponding sixth tree object file according to a project for completing any one labeling task; wherein, the project which completes one labeling task corresponds to a sixth tree object file;
and respectively calculating the storage address and the file content of each sixth tree object file by combining the SHA-1 algorithm, generating a project progress hash value corresponding to each sixth tree object file, and respectively setting a project progress version number corresponding to each sixth tree object file.
Referring to fig. 5, in the present embodiment, the performing the full-process annotation process on the first medical image data set further includes version management on the medical image data set, specifically:
s401: and according to the first medical image data set uploaded by the first user, performing editing operation of adding, modifying or deleting data set contents to generate a corresponding second medical image data set.
S402: and converting the second medical image data set after the editing operation into a second binary large object.
S403: and calculating the storage address and the file content of the second large binary object by combining an SHA-1 algorithm, generating a second data set hash value corresponding to the second large binary object, and setting a second data set version number corresponding to the second large binary object.
Compared with the prior art, the embodiment of the invention has the following beneficial effects:
the invention provides a medical image labeling multi-version management method, which is characterized in that a unique hash value and a version number are generated through an SHA-1 algorithm to identify a medical image data set and a plurality of items, tasks and other objects corresponding to the medical image data set, so that a multi-version system is formed. The unique hash value and the version number which are used as the identification are utilized to realize the control and management of the medical image data set and a plurality of projects and tasks corresponding to the medical image data set, so that the image labeling process can trace back to any version in the modification process through the identification, the modification details of the file are compared, and a modifier of the file is checked, thereby accurately positioning and finding out the reasons of various problems, simultaneously avoiding the loss of image labeling records, improving the image labeling efficiency, and providing a high-quality data base for the innovation driving development of 'AI + medical treatment'.
The above-mentioned embodiments are provided to further explain the objects, technical solutions and advantages of the present invention in detail, and it should be understood that the above-mentioned embodiments are only examples of the present invention and are not intended to limit the scope of the present invention. It should be understood that any modifications, equivalents, improvements and the like, which come within the spirit and principle of the invention, may occur to those skilled in the art and are intended to be included within the scope of the invention.
Claims (10)
1. A method for multi-version management of medical image annotation, comprising:
receiving a first medical image data set uploaded by a first user; wherein the first medical image data set comprises a plurality of medical images to be annotated;
performing full-process labeling processing on the first medical image data set, converting data generated after each process node is completed into tree object files, calculating the storage address and the file content of each tree object file by combining a preset SHA-1 algorithm, generating one or more hash values corresponding to each process node, and configuring a corresponding version number for each hash value; wherein, the full-process labeling process comprises the following process nodes: the system comprises an initial data storage node, an annotation task dividing node, an annotation result uploading node, an annotation acceptance node, an annotation result summarizing node and a project progress updating node;
and according to the input operation instruction, combining the hash value and the version number corresponding to each process node, and performing version management on the first medical image data set.
2. The method for multi-version management of medical image annotation according to claim 1, wherein when the full-process annotation process is in the initial data storage node, the processing operations executed are specifically:
receiving a first medical image data set uploaded by a first user, and converting the first medical image data set into a first binary large object;
calculating the storage address and the file content of the first binary large object by combining the SHA-1 algorithm, generating an initial dataset hash value corresponding to the first binary large object, and setting an initial dataset version number corresponding to the first binary large object;
establishing a plurality of different project branches according to a first medical image data set uploaded by a first user and a preset classification mode, and generating a corresponding first tree object file according to all the project branches;
calculating the storage address and the file content of the first tree object file through a preset SHA-1 algorithm, generating an initial item hash value corresponding to the first tree object file, and setting an initial item version number corresponding to the first tree object file.
3. The method for multi-version management of medical image annotation according to claim 1, wherein when the full-flow annotation process is at the annotation task division node, the processing operations executed are specifically:
generating a plurality of labeling tasks corresponding to the first medical image data set according to a preset labeling task division rule, and generating a corresponding second tree object file according to all labeling tasks; wherein, one labeling task corresponds to one second tree object file;
and calculating the storage address and the file content of the second tree object file by combining the SHA-1 algorithm, generating an initial task hash value corresponding to the second tree object file, and setting an initial task version number corresponding to the second tree object file.
4. The method for multi-version management of medical image annotation according to claim 1, wherein when the full-flow annotation process is at the annotation result uploading node, the processing operations executed are specifically:
responding to the task completion operation of a plurality of users on any one same labeling task, respectively receiving the labeling results of the plurality of users, storing the labeling results as a plurality of corresponding JSON files, and generating a plurality of corresponding third tree object files; the marking result of one user corresponds to one JSON file, and one JSON file corresponds to one third tree object file;
and respectively calculating the storage address and the file content of each third tree object file by combining the SHA-1 algorithm, generating a personal labeling result hash value corresponding to each third tree object file, and respectively setting the personal labeling result version number corresponding to each third tree object file.
5. The method for multi-version management of medical image annotation according to claim 1, wherein when the full-flow annotation process is at the annotation acceptance node, the processing operations executed are specifically:
according to a plurality of JSON files corresponding to the marking results of a plurality of users, performing content comparison on the marking results of the plurality of users, and judging whether the marking results of the plurality of users are consistent;
if the labeling results are consistent, generating a fourth tree object file according to the labeling result of any one labeling task; wherein, the labeling result of one labeling task corresponds to one fourth tree object file;
respectively calculating the storage address and the file content of each fourth tree object file by combining the SHA-1 algorithm, generating a first acceptance result hash value corresponding to each fourth tree object file, and respectively setting a first acceptance result version number corresponding to each fourth tree object file;
and if not, generating a corresponding acceptance version according to the auditing result of the labeling committee.
6. The method for multi-version management of medical image annotation of claim 5, wherein the generating of the corresponding acceptance version according to the review result of the annotation committee comprises:
receiving an auditing result of an annotation committee;
if the verification results are consistent, generating corresponding seventh tree object files according to the labeling results of any one same labeling task, respectively calculating the storage address and the file content of each seventh tree object file by combining the SHA-1 algorithm, generating a second acceptance result hash value corresponding to each seventh tree object file, and respectively setting a second acceptance result version number corresponding to each seventh tree object file;
and if the audit results are inconsistent, generating corresponding eighth tree object files according to the audit results of the labeling committees, respectively calculating the storage address and the file content of each eighth tree object file by combining the SHA-1 algorithm, generating a committee audit result hash value corresponding to each eighth tree object file, respectively setting a committee audit result version number corresponding to each eighth tree object file, and then generating a corresponding acceptance version according to the labeling results of the labeling experts.
7. The method for multi-version management of medical image annotation according to claim 6, wherein the generating of the corresponding acceptance version according to the annotation result of the annotation expert comprises:
receiving the labeling result of the labeling expert and generating a corresponding ninth tree object file;
and respectively calculating the storage address and the file content of each ninth tree object file by combining the SHA-1 algorithm, generating a third acceptance result hash value corresponding to each ninth tree object file, and respectively setting a third acceptance result version number corresponding to each ninth tree object file.
8. The method for multi-version management of medical image annotation according to claim 1, wherein when the full-flow annotation process is at the annotation result summarizing node, the processing operations executed are specifically:
responding to task completion operation of each user on the plurality of labeling tasks, respectively receiving a plurality of labeling results corresponding to each labeling task, and generating a plurality of fifth tree object files; wherein, a plurality of marking results of one marking task correspond to a fifth tree object file;
and respectively calculating the storage address and the file content of each fifth tree object file by combining the SHA-1 algorithm, generating a task marking result hash value corresponding to each fifth tree object file, and respectively setting the task marking result version number corresponding to each fifth tree object file.
9. The method for multi-version management of medical image annotation according to claim 1, wherein when the full-process annotation process is at the project progress update node, the processing operations executed are specifically:
responding to task completion operation of each user on the plurality of labeling tasks, and generating a corresponding sixth tree object file according to a project for completing any one labeling task; wherein, the project which completes one labeling task corresponds to a sixth tree object file;
and respectively calculating the storage address and the file content of each sixth tree object file by combining the SHA-1 algorithm, generating a project progress hash value corresponding to each sixth tree object file, and respectively setting a project progress version number corresponding to each sixth tree object file.
10. The method for multi-version management of medical image annotation according to any one of claims 1 to 9, wherein the version management of the first medical image data set specifically comprises:
determining a first flow node to be managed according to an input operation instruction;
inquiring a hash value corresponding to the first process node, and determining a version number to be managed of the first medical image data set according to the inquired hash value;
and calling the medical image data of the first medical image data set under the version number to be managed according to the version number to be managed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111189659.2A CN114003745A (en) | 2021-10-12 | 2021-10-12 | Medical image labeling multi-version management method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111189659.2A CN114003745A (en) | 2021-10-12 | 2021-10-12 | Medical image labeling multi-version management method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114003745A true CN114003745A (en) | 2022-02-01 |
Family
ID=79922776
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111189659.2A Pending CN114003745A (en) | 2021-10-12 | 2021-10-12 | Medical image labeling multi-version management method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114003745A (en) |
-
2021
- 2021-10-12 CN CN202111189659.2A patent/CN114003745A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109634679B (en) | Block chain configuration updating method, device, equipment and storage medium | |
| CN107301214B (en) | Data migration method and device in HIVE and terminal equipment | |
| CN107016047A (en) | Document query, document storing method and device | |
| US9378260B2 (en) | Methods and apparatus for synchronizing closed heterogenous systems | |
| CN111061733B (en) | Data processing method, device, electronic equipment and computer readable storage medium | |
| CN110275889B (en) | Feature processing method and device suitable for machine learning | |
| CN114385652A (en) | Data blood relationship construction method and system, electronic device and storage medium | |
| CN111435367A (en) | Knowledge graph construction method, system, equipment and storage medium | |
| US10417234B2 (en) | Data flow modeling and execution | |
| CN113641651A (en) | Business data management method, system and computer storage medium | |
| US11157247B2 (en) | Cryptographic digital twin segregation system for development operations | |
| US9092472B1 (en) | Data merge based on logical segregation | |
| CN114003745A (en) | Medical image labeling multi-version management method | |
| US20060069679A1 (en) | System for creating associative records | |
| CN108427675A (en) | Build the method and apparatus of index | |
| US20130110730A1 (en) | Integration of computerized project planning and project diagramming | |
| JP2019101829A (en) | Software component management system, computor, and method | |
| CN111813880B (en) | Homeland space planning project management method, system and storage medium | |
| JP2024505238A (en) | Data processing system with operations on logical dataset groups | |
| JP6397657B2 (en) | Requirements calculation system, requirements calculation method and program | |
| CN116959656B (en) | Medical main index extraction merging method and system based on ES | |
| CN113377604A (en) | Data processing method, device, equipment and storage medium | |
| CN117827252B (en) | Function demonstration method, server and storage medium for system update | |
| KR102453579B1 (en) | System for providing cooperation solution for collaboration between workers | |
| CN113590213B (en) | Component maintenance method, electronic device and storage medium |
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 |