CN114201459A - An improved radiotherapy structure standardized nomenclature method, device and storage medium - Google Patents
An improved radiotherapy structure standardized nomenclature method, device and storage medium Download PDFInfo
- Publication number
- CN114201459A CN114201459A CN202111297814.2A CN202111297814A CN114201459A CN 114201459 A CN114201459 A CN 114201459A CN 202111297814 A CN202111297814 A CN 202111297814A CN 114201459 A CN114201459 A CN 114201459A
- Authority
- CN
- China
- Prior art keywords
- radiotherapy
- radiotherapy structure
- interest
- region
- standard
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
- G06F16/16—File or folder operations, e.g. details of user interfaces specifically adapted to file systems
- G06F16/164—File meta data generation
- G06F16/166—File name conversion
-
- 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
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/40—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mechanical, radiation or invasive therapies, e.g. surgery, laser therapy, dialysis or acupuncture
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Human Computer Interaction (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Surgery (AREA)
- Public Health (AREA)
- Urology & Nephrology (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiation-Therapy Devices (AREA)
Abstract
本发明公开了一种改进的放疗结构标准化命名方法、装置及存储介质。本发明通过将原始放疗结构文件中的原始放疗结构命名转化为标准放疗结构命名，能够对放疗结构命名本身进行标准化处理，保证放疗结构命名的一致性，通过将所有标准放疗结构文件分为训练集和测试集，根据训练集中的放疗结构勾画轮廓及其特征参数建立知识库和放疗结构分类模型，应用知识库和放疗结构分类模型对通过位置验证的测试集进行结构验证，能够对标准放疗结构命名与放疗结构勾画轮廓进行一致性验证，保证标准放疗结构名称与放疗结构勾画轮廓相互匹配，从而进一步实现放疗结构标准化命名，确保内容一致，提高放疗安全性。
The invention discloses an improved radiotherapy structure standardized naming method, device and storage medium. By transforming the original radiotherapy structure naming in the original radiotherapy structure file into the standard radiotherapy structure naming, the present invention can standardize the radiotherapy structure naming itself and ensure the consistency of the radiotherapy structure naming. All standard radiotherapy structure files are divided into training sets. And the test set, according to the outline of the radiotherapy structure in the training set and its characteristic parameters to establish a knowledge base and a radiotherapy structure classification model, apply the knowledge base and the radiotherapy structure classification model to verify the structure of the test set that has passed the location verification, and can name the standard radiotherapy structure. Consistency verification is carried out with the outline of the radiotherapy structure to ensure that the standard radiotherapy structure name and the outline of the radiotherapy structure match each other, so as to further realize the standardized naming of the radiotherapy structure, ensure the consistency of the content, and improve the safety of radiotherapy.
Description
Technical Field
The invention relates to the technical field of computer medical treatment, in particular to an improved radiotherapy structure standardized naming method, device and storage medium.
Background
In the field of tumor radiotherapy, more and more institutions place more importance on standardization of radiotherapy structure nomenclature. The existing radiotherapy structure standardization naming method mainly considers the standardization of radiotherapy structure naming per se to avoid confusion caused by inconsistent or inappropriate naming, and can be applied to multiple cancer types and multiple centers. However, in the design and implementation process of the radiotherapy plan, due to manual operation, the radiotherapy structure naming and the radiotherapy structure delineating contour may not be matched, which easily causes omission and confusion, is difficult to further realize the radiotherapy structure standardized naming, and cannot better improve the radiotherapy safety.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides an improved radiotherapy structure standardized naming method, a device and a storage medium, which can further realize the radiotherapy structure standardized naming, ensure the content consistency and improve the radiotherapy safety.
In order to solve the above technical problems, in a first aspect, an embodiment of the present invention provides an improved method for standardizing naming of radiotherapy structures, including:
acquiring a plurality of original radiotherapy structure files, and converting the original radiotherapy structure name in each original radiotherapy structure file into a standard radiotherapy structure name to obtain a plurality of standard radiotherapy structure files;
dividing all the standard radiotherapy structure files into a training set and a testing set; the training set only comprises the standard radiotherapy structure file with a normal radiotherapy structure, and the test set comprises the standard radiotherapy structure file with a normal radiotherapy structure and the standard radiotherapy structure file with an abnormal radiotherapy structure;
respectively extracting a radiotherapy structure delineation contour in each standard radiotherapy structure file in the training set as a first region of interest, acquiring characteristic parameters of the first region of interest, and establishing a knowledge base and a radiotherapy structure classification model according to the acquired characteristic parameters of all the first region of interest and all the first region of interest;
traversing each standard radiotherapy structure file in the test set, extracting a radiotherapy structure delineation contour in the current standard radiotherapy structure file as a second region of interest, acquiring position information and characteristic parameters of the second region of interest, performing structure verification on the second region of interest and the characteristic parameters of the second region of interest sequentially through the knowledge base and the radiotherapy structure classification model when the position information of the second region of interest passes the position verification, and taking the current standard radiotherapy structure file as a target radiotherapy structure file if the verification result is that the radiotherapy structure is normal.
Further, the improved radiotherapy structure standardized naming method further comprises the following steps:
when the position information of the second region of interest does not pass the position verification, modifying the current standard radiotherapy structure file to obtain a first standard radiotherapy structure file, extracting a radiotherapy structure delineation contour in the first standard radiotherapy structure file to serve as a third region of interest, acquiring the position information and characteristic parameters of the third region of interest, when the position information of the third region of interest passes the position verification, sequentially passing through the knowledge base and the radiotherapy structure classification model to perform structure verification on the third region of interest and the characteristic parameters of the third region of interest, and if the verification result is that the radiotherapy structure is normal, using the first standard radiotherapy structure file as the target radiotherapy structure file.
Further, the improved radiotherapy structure standardized naming method further comprises the following steps:
if the verification result is that the radiotherapy structure is abnormal, modifying the current standard radiotherapy structure file to obtain a second standard radiotherapy structure file, extracting a radiotherapy structure delineation contour in the second standard radiotherapy structure file to serve as a fourth region of interest, acquiring position information and characteristic parameters of the fourth region of interest, performing structure verification on the fourth region of interest and the characteristic parameters of the fourth region of interest sequentially through the knowledge base and the radiotherapy structure classification model when the position information of the fourth region of interest passes the position verification, and if the verification result is that the radiotherapy structure is normal, taking the second standard radiotherapy structure file as the target radiotherapy structure file.
Further, the converting the original radiotherapy structure name in each original radiotherapy structure file into a standard radiotherapy structure name to obtain a plurality of standard radiotherapy structure files specifically includes:
and respectively reading the original radiotherapy structure names in each original radiotherapy structure file, and converting the original radiotherapy structure names in each original radiotherapy structure file into standard radiotherapy structure names according to a predefined radiotherapy structure name conversion table to obtain a plurality of standard radiotherapy structure files.
Further, establishing a knowledge base and a radiotherapy structure classification model according to the obtained characteristic parameters of all the first regions of interest and all the first regions of interest, specifically:
and performing data analysis on all the first interested areas and the characteristic parameters of all the first interested areas to obtain a normal characteristic parameter range of the radiotherapy structure, establishing the knowledge base by combining the normal characteristic parameter range, and establishing the classification model of the radiotherapy structure according to the knowledge base.
Further, after extracting a radiotherapy structure delineation contour in the current standard radiotherapy structure file as a second region of interest in each standard radiotherapy structure file in the traversal test set, and acquiring position information and characteristic parameters of the second region of interest, before performing structure verification on the second region of interest and the characteristic parameters of the second region of interest sequentially through the knowledge base and the radiotherapy structure classification model when the position information of the second region of interest passes the position verification, the method further includes:
and judging whether the position information of the second region of interest is correct or not according to a predefined position verification rule, and if so, passing position verification on the position information of the second region of interest.
Further, the structural verification of the second region of interest and the characteristic parameters of the second region of interest sequentially through the knowledge base and the radiotherapy structural classification model specifically includes:
and comparing the characteristic parameters of the second region of interest and the second region of interest with the data in the knowledge base, and judging whether the radiotherapy structure of the standard radiotherapy structure is abnormal or not according to the characteristic parameters of the second region of interest and the second region of interest through a radiotherapy structure classification model after the comparison is successful.
In a second aspect, an embodiment of the present invention provides an improved apparatus for standardizing naming of radiotherapy structures, including:
the system comprises a naming conversion module, a data processing module and a data processing module, wherein the naming conversion module is used for acquiring a plurality of original radiotherapy structure files, converting the original radiotherapy structure naming in each original radiotherapy structure file into a standard radiotherapy structure naming, and obtaining a plurality of standard radiotherapy structure files;
the file dividing module is used for dividing all the standard radiotherapy structure files into a training set and a testing set; the training set only comprises the standard radiotherapy structure file with a normal radiotherapy structure, and the test set comprises the standard radiotherapy structure file with a normal radiotherapy structure and the standard radiotherapy structure file with an abnormal radiotherapy structure;
the model base establishing module is used for respectively extracting the radiotherapy structure delineation contour in each standard radiotherapy structure file in the training set as a first region of interest, acquiring the characteristic parameters of the first region of interest, and establishing a knowledge base and a radiotherapy structure classification model according to the acquired characteristic parameters of all the first region of interest and all the first region of interest;
and the file verification module is used for traversing each standard radiotherapy structure file in the test set, extracting the outline of the radiotherapy structure in the current standard radiotherapy structure file to be used as a second region of interest, acquiring the position information and the characteristic parameters of the second region of interest, performing structure verification on the second region of interest and the characteristic parameters of the second region of interest sequentially through the knowledge base and the radiotherapy structure classification model when the position information of the second region of interest passes the position verification, and if the verification result is that the radiotherapy structure is normal, using the current standard radiotherapy structure file as a target radiotherapy structure file.
In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, which includes a stored computer program, where the computer program, when running, controls an apparatus in which the computer-readable storage medium is located to execute the above-mentioned improved radiotherapy structure standardized naming method.
The embodiment of the invention has the following beneficial effects:
obtaining a plurality of standard radiotherapy structure files by acquiring a plurality of original radiotherapy structure files and converting the original radiotherapy structure names in each original radiotherapy structure file into standard radiotherapy structure names, and dividing all the standard radiotherapy structure files into a training set and a test set; the training set only comprises standard radiotherapy structure files with normal radiotherapy structures, the testing set comprises the standard radiotherapy structure files with the normal radiotherapy structures and the standard radiotherapy structure files with abnormal radiotherapy structures, the radiotherapy structure delineation contour in each standard radiotherapy structure file in the training set is respectively extracted to serve as a first region of interest, characteristic parameters of the first region of interest are obtained, a knowledge base and a radiotherapy structure classification model are established according to the obtained characteristic parameters of all the first region of interest and all the first region of interest, each standard radiotherapy structure file in the testing set is traversed, the radiotherapy structure delineation contour in the current standard radiotherapy structure file is extracted to serve as a second region of interest, position information and characteristic parameters of the second region of interest are obtained, and when the position information of the second region of interest passes position verification, the characteristic parameters of the second region of interest and the second region of interest sequentially pass through the knowledge base and the radiotherapy structure classification model And performing structure verification, and if the verification result is that the radiotherapy structure is normal, taking the current standard radiotherapy structure file as a target radiotherapy structure file, so that the design and implementation of a radiotherapy plan can be performed by using the target radiotherapy structure. Compared with the prior art, the embodiment of the invention can carry out standardized processing on the radiotherapy structure names by converting the original radiotherapy structure names in the original radiotherapy structure files into the standard radiotherapy structure names, ensures the consistency of the radiotherapy structure names, divides all the standard radiotherapy structure files into the training set and the test set, establishing a knowledge base and a radiotherapy structure classification model according to the radiotherapy structure delineation contour in the training set and characteristic parameters thereof, applying the knowledge base and the radiotherapy structure classification model to carry out structure verification on the test set passing the position verification, being capable of carrying out consistency verification on the standard radiotherapy structure name and the radiotherapy structure delineation contour, ensuring that the standard radiotherapy structure name is matched with the radiotherapy structure delineation contour, thereby further realizing the standardized naming of the radiotherapy structure, ensuring the content consistency and improving the safety of the radiotherapy.
Drawings
FIG. 1 is a schematic flow chart of an improved method for standardized naming of radiotherapy structures according to a first embodiment of the present invention;
FIG. 2 is a schematic flow chart of an improved method for standardized naming of radiotherapy structures according to a first embodiment of the present invention;
fig. 3 is a schematic structural diagram of an improved radiotherapy structure standardization naming device in a second embodiment of the present invention.
Detailed Description
The technical solutions in the present invention will be described clearly and completely with reference to the accompanying drawings, and it is obvious that the described embodiments are only some embodiments of the present invention, not all 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.
It should be noted that, the step numbers in the text are only for convenience of explanation of the specific embodiments, and do not serve to limit the execution sequence of the steps. The method provided by the embodiment can be executed by the relevant terminal device, and the following description takes a processor as an execution subject as an example.
As shown in fig. 1, the first embodiment provides an improved method for standardizing the naming of radiotherapy structures, which includes steps S1-S4:
s1, acquiring a plurality of original radiotherapy structure files, and converting the original radiotherapy structure names in each original radiotherapy structure file into standard radiotherapy structure names to obtain a plurality of standard radiotherapy structure files;
s2, dividing all standard radiotherapy structure files into a training set and a testing set; the training set only comprises standard radiotherapy structure files with normal radiotherapy structures, and the test set comprises standard radiotherapy structure files with normal radiotherapy structures and standard radiotherapy structure files with abnormal radiotherapy structures;
s3, respectively extracting the radiotherapy structure delineation contour in each standard radiotherapy structure file in the training set as a first region of interest, acquiring characteristic parameters of the first region of interest, and establishing a knowledge base and a radiotherapy structure classification model according to all the obtained first regions of interest and the characteristic parameters of all the first regions of interest;
s4, traversing each standard radiotherapy structure file in the test set, extracting a radiotherapy structure delineation contour in the current standard radiotherapy structure file as a second region of interest, acquiring position information and characteristic parameters of the second region of interest, performing structure verification on the second region of interest and the characteristic parameters of the second region of interest sequentially through a knowledge base and a radiotherapy structure classification model when the position information of the second region of interest passes the position verification, and if the verification result is that the radiotherapy structure is normal, taking the current standard radiotherapy structure file as a target radiotherapy structure file.
It is noted that the radiotherapy construct includes the target and the organs at risk.
As shown in fig. 2, 855 original radiotherapy Structure files of nasopharyngeal carcinoma patients, i.e. DICOM RT Structure files, containing 15 clinical abnormal structures, of which 625 men and 230 women all received IMRT treatment with the average age of (47 ± 12) years, are obtained exemplarily in step S1. All patients adopt CT or CT simulation (CT-Sim) to obtain CT images, the scanning range is from the upper edge of the frontal sinus to the lower edge of the clavicle head by 2cm, the scanning mode is flat scanning and enhanced, the layer thickness is 3mm, and the body positions are all in the first advanced supine position. After the primary target area and the organs at risk are delineated by the main doctor, the CT image of each patient is examined or modified and perfected by the main or subsidiary main physicians, and then a radiotherapy plan design is carried out by using a Treatment Planning System (TPS), wherein the image, the structure, the plan and the dose are standard DICOM files.
In order to facilitate data cleaning and classification analysis, the names of organs at risk in original radiotherapy structure files of 855 cases of nasopharyngeal carcinoma patients are subjected to semantic-Based name standardization (SBSN), namely MATLAB software is used for analyzing the original radiotherapy structure files of patients, original radiotherapy structure name fields in the original radiotherapy structure files are read and stored, a corresponding nasopharyngeal carcinoma radiotherapy structure name transformation table is formulated according to an AAPM-263 radiotherapy structure name standardization principle, the original radiotherapy structure names in the original radiotherapy structure files are transformed into standard radiotherapy structure names through SBSN, standard radiotherapy structure files are obtained, and the successfully transformed standard radiotherapy structure files are reintroduced into a radiotherapy planning system for result verification.
At step S2, 855 standard radiotherapy profiles for nasopharyngeal carcinoma patients were divided into training and testing sets after SBSN. The training set is 719 randomly selected from 840 standard radiotherapy structure files of nasopharyngeal carcinoma patients except 15 clinical abnormal structures, and the testing set is the remaining 121 and 15 clinical abnormal structures in 840 standard radiotherapy structure files of nasopharyngeal carcinoma patients.
In step S3, for each standard radiotherapy structure file in the training set, the standard radiotherapy structure file is analyzed by the mabab software, a radiotherapy structure delineation contour in the standard radiotherapy structure file is extracted as a first Region of Interest (ROI), and a characteristic parameter of the first Region of Interest is obtained: volume, number of layers, ct (hu) mean, 5 Gray features (i.e., Gray minimum, Gray maximum, Gray standard deviation, Gray skewness, and Gray kurtosis) based on first-order Gray histogram, and 5 texture features (i.e., contrast, homogeneity, cross-correlation, energy, and entropy) based on Gray-level Co-occurrence Matrix (GLCM).
It can be understood that the gray level co-occurrence matrix texture feature description proposed by Haralick et al is widely applied to the medical field, the gray level of an original CT image is mapped to 16 levels, the distance d is 1, the gray level co-occurrence matrices in the directions of 0 °, 45 °, 90 ° and 135 ° are solved, and 5 mutually independent second-order texture feature quantities, namely contrast, homogeneity, cross-correlation, energy and entropy, are calculated based on the GLCM in 4 directions.
And performing data analysis on the obtained characteristic parameters of all the first interested areas and all the first interested areas, aiming at identification such as missing of a contour structure or abnormal volume of a nasopharyngeal carcinoma endangered organ, determining a normal characteristic parameter range from P5 to P95 by using a percentile, and establishing a knowledge base by using < P5 or > P95 as a standard for judging an abnormal structure.
And establishing a radiotherapy structure classification model based on the characteristic parameter values obtained by the knowledge base, and further identifying and correcting the incorrectly marked abnormal structure. And establishing a radiotherapy structure classification model by using Fisher discriminant analysis, and evaluating the performance of the model by adopting self verification, interactive verification and external verification.
It is understood that self-validation refers to self-generation validation using the training set 719 cases of patient data; interactive verification means that 719 cases of patient data in a training set are divided into 10 sets, one set is randomly selected to be used as a verification set each time, the other 9 sets are used as the training sets, and the verification sets can be exchanged with 10 possibilities; external validation refers to validation of patient data using test set 136.
In step S4, traversing the standard radiotherapy structure file in the test set, analyzing the currently traversed standard radiotherapy structure file by using the mabab software, extracting a radiotherapy structure delineation contour in the standard radiotherapy structure file as a second Region of Interest (ROI), and acquiring position information of the second Region of Interest and characteristic parameters of the second Region of Interest: volume, number of layers, ct (hu) mean, 5 Gray features (i.e., Gray minimum, Gray maximum, Gray standard deviation, Gray skewness, and Gray kurtosis) based on first-order Gray histogram, and 5 texture features (i.e., contrast, homogeneity, cross-correlation, energy, and entropy) based on Gray-level Co-occurrence Matrix (GLCM).
And carrying out position verification on the position information of the second region of interest, wherein the position verification mainly aims at identifying the nasopharyngeal carcinoma endangered organ level position abnormality and the left and right position abnormality in the symmetrical structure. According to the body position information of the patient, based on a predefined position verification rule, judging whether the position of an Organ At Risk (OAR) in a standard radiotherapy structure is correct or not according to corresponding fields in the standard radiotherapy structure file, specifically, reading fields such as Patientposition and ImagePositionsPatent in the standard radiotherapy structure file to determine the three-dimensional space positioning of the patient, and establishing the transformation of world coordinates and image grid voxel coordinates. The position verification mainly identifies 4 types of abnormal structures: (1) judging the position abnormality of the opposite layer; (2) judging a standard of left and right positions; (3) detecting whether the OARs have non-unique structures; (4) garbage point: abnormal structures outside the outer contour are detected.
And when the position information of the second region of interest passes the position verification, sequentially passing through the knowledge base and the radiotherapy structure classification model to perform structure verification on the second region of interest and the characteristic parameters of the second region of interest, and if the verification result shows that the radiotherapy structure is normal, taking the current standard radiotherapy structure file as a target radiotherapy structure file.
The content-based radiotherapy structure standardization naming processing (CBSN) mainly relates to CBSN position verification, a CBSN knowledge base and a CBSN radiotherapy structure classification model. And finally, after CBSN verification, the standard radiotherapy structure file which judges the radiotherapy structure to be normal enters a plan design stage, and the standard radiotherapy structure file which judges the radiotherapy structure to be abnormal returns to check and modify.
The abnormal structure identification is an important basis for evaluating the accuracy of the CBSN, 121 cases of patient data derived simulation abnormal structures in the test set are used as CBSN simulation test evaluation, and according to 15 cases of clinical abnormal structure conditions, the 121 cases of patient data derived simulation abnormal structures in the test set are respectively used for CBSN position verification and evaluation of a CBSN knowledge base and a CBSN danger organ classification model: (1) simulating 40 position abnormal structures such as the right Eye named as 'Eye _ L' or the structure named as 'Len _ R' positioned on the top of the head level for CBSN position verification evaluation; (2) simulating 70 abnormal organs at risk such as volume and the like for CBSN knowledge base evaluation; (3) 277 crystals named as 'OpticNerve' or abnormal structures such as brainstem named as 'SpinalCord' are simulated, and the CBSN organs-at-risk classification model is used for distinguishing and detecting.
This embodiment turns into standard radiotherapy structure naming through original radiotherapy structure naming among the original radiotherapy structure file, can carry out standardized processing to radiotherapy structure naming itself, guarantee the uniformity of radiotherapy structure naming, through divide into training set and test set with all standard radiotherapy structure files, establish knowledge base and radiotherapy structure classification model according to the radiotherapy structure delineation profile and the characteristic parameter in the training set, use knowledge base and radiotherapy structure classification model to carry out the structure verification to the test set through position verification, can carry out the uniformity verification to standard radiotherapy structure naming and radiotherapy structure delineation profile, guarantee that it is matchd between standard radiotherapy structure name and the radiotherapy structure delineation profile, thereby further the radiotherapy realizes the standardized naming of radiotherapy structure, ensure the content is unanimous, improve the radiotherapy security.
In a preferred embodiment, the improved radiotherapy structure standardized naming method further comprises: when the position information of the second region of interest does not pass the position verification, modifying the current standard radiotherapy structure file to obtain a first standard radiotherapy structure file, extracting a radiotherapy structure delineation contour in the first standard radiotherapy structure file to serve as a third region of interest, acquiring the position information and characteristic parameters of the third region of interest, when the position information of the third region of interest passes the position verification, sequentially passing through a knowledge base and a radiotherapy structure classification model to perform structure verification on the third region of interest and the characteristic parameters of the third region of interest, and if the verification result is that the radiotherapy structure is normal, using the first standard radiotherapy structure file as a target radiotherapy structure file.
Illustratively, when the radiotherapy structure delineation contour in the currently traversed standard radiotherapy structure file, that is, the second region of interest does not pass the position verification, the standard radiotherapy structure name in the current standard radiotherapy structure file is considered to be inconsistent with the radiotherapy structure delineation contour, and the current standard radiotherapy structure file needs to be returned and modified to obtain a new standard radiotherapy structure file, so as to perform the position verification and the structure verification again.
In a preferred embodiment, the improved radiotherapy structure standardized naming method further comprises: if the verification result is that the radiotherapy structure is abnormal, modifying the current standard radiotherapy structure file to obtain a second standard radiotherapy structure file, extracting a radiotherapy structure delineation contour in the second standard radiotherapy structure file to serve as a fourth region of interest, acquiring position information and characteristic parameters of the fourth region of interest, performing structure verification on the fourth region of interest and the characteristic parameters of the fourth region of interest sequentially through a knowledge base and a radiotherapy structure classification model when the position information of the fourth region of interest passes the position verification, and if the verification result is that the radiotherapy structure is normal, taking the second standard radiotherapy structure file as a target radiotherapy structure file.
As an example, when the radiotherapy structure delineation contour in the currently traversed standard radiotherapy structure file, that is, the second region of interest and the feature parameters of the second region of interest do not pass through the structure verification of the knowledge base and the radiotherapy structure classification model, it is considered that the standard radiotherapy structure name in the current standard radiotherapy structure file is inconsistent with the radiotherapy structure delineation contour, and the current standard radiotherapy structure file needs to be returned and modified to obtain a new standard radiotherapy structure file for performing the position verification and the structure verification again.
In a preferred embodiment, the original radiotherapy structure name in each original radiotherapy structure file is converted into a standard radiotherapy structure name to obtain a plurality of standard radiotherapy structure files, specifically: and respectively reading the original radiotherapy structure names in each original radiotherapy structure file, and converting the original radiotherapy structure names in each original radiotherapy structure file into standard radiotherapy structure names according to a predefined radiotherapy structure name conversion table to obtain a plurality of standard radiotherapy structure files.
Illustratively, MATLAB software is utilized to analyze an original radiotherapy structure file of a patient, an original radiotherapy structure naming field in the original radiotherapy structure file is read and stored, a corresponding nasopharyngeal carcinoma radiotherapy structure naming conversion table is formulated according to an AAPM TG-263 radiotherapy structure naming standardization principle, and an original radiotherapy structure naming in the original radiotherapy structure file is converted into a standard radiotherapy structure naming through SBSN to obtain a standard radiotherapy structure file.
In a preferred embodiment, the establishing a knowledge base and a radiotherapy structure classification model according to the obtained characteristic parameters of all the first regions of interest and all the first regions of interest specifically includes: and performing data analysis on all the first interested areas and the characteristic parameters of all the first interested areas to obtain a normal characteristic parameter range of the radiotherapy structure, establishing a knowledge base by combining the normal characteristic parameter range, and establishing a classification model of the radiotherapy structure according to the knowledge base.
Illustratively, by performing data analysis on all the obtained first interested areas and characteristic parameters of all the first interested areas, aiming at identification such as missing contour structure or abnormal volume of nasopharyngeal carcinoma organs at risk, determining normal characteristic parameter ranges from P5 to P95 by using percentiles, and establishing a knowledge base by taking < P5 or > P95 as a standard for judging abnormal structures.
And establishing a radiotherapy structure classification model based on the characteristic parameter values obtained by the knowledge base, and further identifying and correcting the incorrectly marked abnormal structure. And establishing a radiotherapy structure classification model by using Fisher discriminant analysis, and evaluating the performance of the model by adopting self verification, interactive verification and external verification.
In a preferred embodiment, after extracting, in each standard radiotherapy structure file in the traversal test set, a radiotherapy structure delineation contour in a current standard radiotherapy structure file as a second region of interest, and acquiring location information and characteristic parameters of the second region of interest, when the location information of the second region of interest passes location verification, before performing structure verification on the second region of interest and the characteristic parameters of the second region of interest sequentially through a knowledge base and a radiotherapy structure classification model, the method further includes: and judging whether the position information of the second region of interest is correct or not according to a predefined position verification rule, and if so, passing the position verification of the position information of the second region of interest.
Illustratively, the location verification is primarily directed to the identification of nasopharyngeal carcinoma endangered organ level location abnormalities and left-right location abnormalities in symmetric structures. According to the body position information of the patient, based on a predefined position verification rule, judging whether the position of an Organ At Risk (OAR) in a standard radiotherapy structure is correct or not according to corresponding fields in the standard radiotherapy structure file, specifically, reading fields such as Patientposition and ImagePositionsPatent in the standard radiotherapy structure file to determine the three-dimensional space positioning of the patient, and establishing the transformation of world coordinates and image grid voxel coordinates. In addition, the critical organ delineation contour information in the standard radiotherapy structure file is stored in fields such as ROIContourSequence and the like, the position information of the critical organ contour is analyzed and read, and world coordinate and image grid voxel coordinate conversion is established for subsequent analysis.
The CBSN position verification mainly identifies 4 types of abnormal structures:
(1) and establishing judgment of the relative layer position abnormality of the nasopharyngeal darcinoma OARs according to the three-dimensional space positioning information and the ROI position field. Converting pixel points in each layer of contour of the organ at risk from a world coordinate system to a grid voxel coordinate system, simultaneously acquiring specific numerical values of image layers corresponding to the contour in the Z-axis direction, establishing an empirical range value of the contour layer of the organ at risk in the Z-axis direction, and forming judgment, for example: and the organs at risk named as 'Eye _ L' are delineated at the vertex level of the Z-axis numerical value and belong to relative level position abnormity.
(2) All patients with nasopharyngeal carcinoma are in the First advanced Supine position (HFS), and the judgment standard of the left and right positions of the nasopharyngeal carcinoma is established according to the field information such as body position and the like. First, the left and right position criteria on the two-dimensional image are determined from the patient position information, dividing the patient outline (generally named "body" or "External" structure) into two regions in the middle: left and right. Converting the pixel points in each layer of contour of the organ at risk from a world coordinate system to a grid voxel coordinate system, reading the positions of all the pixel points on the two-dimensional image, and judging the region to which the pixel points belong, such as: a threat organ named right lens "Len _ R" is located in the left region belonging to left and right position abnormalities.
(3) And reading fields such as ROIContourSequence and the like to detect whether the organ structure at risk is non-unique.
(4) Garbage point: and reading fields such as ROIContourSequence and the like, and detecting abnormal structures outside the outline. After reading the outline of the patient (generally named as a "body" or "External" structure) as an area of interest, detecting whether other sketched outlines (garbage points) exist outside the outline, judging by identifying the number of outlines appearing in the same image plane, and if the number of outlines appearing in the same plane is more than 2, determining that the sketched outlines exist in the image background area.
In a preferred embodiment, the structure verification is performed on the second region of interest and the characteristic parameters of the second region of interest sequentially through a knowledge base and a radiotherapy structure classification model, specifically: and comparing the characteristic parameters of the second region of interest and the second region of interest with the data in the knowledge base, and judging whether the radiotherapy structure of the current standard radiotherapy structure is abnormal or not through the radiotherapy structure classification model after the comparison is successful according to the characteristic parameters of the second region of interest and the second region of interest.
The CBSN knowledge base aims at identifying nasopharyngeal carcinoma organs at abnormal volumes and the like, and the CBSN organs at risk classification model can further predict and identify and correct wrongly marked abnormal structures.
In the embodiment, the characteristic parameters of the second region of interest and the second region of interest are compared with the data in the knowledge base, and then the radiotherapy structure classification model is continuously passed through after the comparison is successful, and whether the radiotherapy structure of the current standard radiotherapy structure is abnormal is judged according to the characteristic parameters of the second region of interest and the second region of interest, so that the consistency verification of the name of the standard radiotherapy structure and the outline of the radiotherapy structure can be performed at multiple levels, the accuracy of the verification result is ensured, the standardized naming of the radiotherapy structure is further realized, the consistency of the content is ensured, and the safety of the radiotherapy is improved.
Based on the same inventive concept as the first embodiment, the second embodiment provides an improved radiotherapy structure standardized naming device as shown in fig. 3, comprising: the name conversion module 21 is configured to obtain a plurality of original radiotherapy structure files, and convert an original radiotherapy structure name in each original radiotherapy structure file into a standard radiotherapy structure name to obtain a plurality of standard radiotherapy structure files; the file dividing module 22 is used for dividing all standard radiotherapy structure files into a training set and a test set; the training set only comprises standard radiotherapy structure files with normal radiotherapy structures, and the test set comprises standard radiotherapy structure files with normal radiotherapy structures and standard radiotherapy structure files with abnormal radiotherapy structures; the model base establishing module 23 is configured to extract the radiotherapy structure delineation contour in each standard radiotherapy structure file in the training set as a first region of interest, acquire characteristic parameters of the first region of interest, and establish a knowledge base and a radiotherapy structure classification model according to the acquired characteristic parameters of all the first regions of interest and all the first regions of interest; the file verification module 24 is configured to traverse each standard radiotherapy structure file in the test set, extract a radiotherapy structure delineation contour in the current standard radiotherapy structure file as a second region of interest, acquire position information and characteristic parameters of the second region of interest, perform structure verification on the second region of interest and the characteristic parameters of the second region of interest sequentially through the knowledge base and the radiotherapy structure classification model when the position information of the second region of interest passes the position verification, and if the verification result is that the radiotherapy structure is normal, take the current standard radiotherapy structure file as a target radiotherapy structure file.
In a preferred embodiment, the file verification module 24 is further configured to modify the current standard radiotherapy structure file to obtain a first standard radiotherapy structure file when the position information of the second region of interest does not pass the position verification, extract a radiotherapy structure delineation contour in the first standard radiotherapy structure file as a third region of interest, obtain position information and characteristic parameters of the third region of interest, perform structure verification on the third region of interest and the characteristic parameters of the third region of interest sequentially through the knowledge base and the radiotherapy structure classification model when the position information of the third region of interest passes the position verification, and if the verification result is that the radiotherapy structure is normal, take the first standard radiotherapy structure file as the target radiotherapy structure file.
In a preferred embodiment, the file verification module 24 is further configured to modify the current standard radiotherapy structure file to obtain a second standard radiotherapy structure file if the verification result is that the radiotherapy structure is abnormal, extract a radiotherapy structure delineation contour in the second standard radiotherapy structure file as a fourth region of interest, obtain position information and characteristic parameters of the fourth region of interest, perform structure verification on the fourth region of interest and the characteristic parameters of the fourth region of interest sequentially through the knowledge base and the radiotherapy structure classification model when the position information of the fourth region of interest passes the position verification, and take the second standard radiotherapy structure file as a target radiotherapy structure file if the verification result is that the radiotherapy structure is normal.
In a preferred embodiment, the original radiotherapy structure name in each original radiotherapy structure file is converted into a standard radiotherapy structure name to obtain a plurality of standard radiotherapy structure files, specifically: and respectively reading the original radiotherapy structure names in each original radiotherapy structure file, and converting the original radiotherapy structure names in each original radiotherapy structure file into standard radiotherapy structure names according to a predefined radiotherapy structure name conversion table to obtain a plurality of standard radiotherapy structure files.
In a preferred embodiment, the establishing a knowledge base and a radiotherapy structure classification model according to the obtained characteristic parameters of all the first regions of interest and all the first regions of interest specifically includes: and performing data analysis on all the first interested areas and the characteristic parameters of all the first interested areas to obtain a normal characteristic parameter range of the radiotherapy structure, establishing a knowledge base by combining the normal characteristic parameter range, and establishing a classification model of the radiotherapy structure according to the knowledge base.
In a preferred embodiment, after extracting, in each standard radiotherapy structure file in the traversal test set, a radiotherapy structure delineation contour in a current standard radiotherapy structure file as a second region of interest, and acquiring location information and characteristic parameters of the second region of interest, when the location information of the second region of interest passes location verification, before performing structure verification on the second region of interest and the characteristic parameters of the second region of interest sequentially through a knowledge base and a radiotherapy structure classification model, the method further includes: and judging whether the position information of the second region of interest is correct or not according to a predefined position verification rule, and if so, passing the position verification of the position information of the second region of interest.
In a preferred embodiment, the structure verification is performed on the second region of interest and the characteristic parameters of the second region of interest sequentially through a knowledge base and a radiotherapy structure classification model, specifically: and comparing the characteristic parameters of the second region of interest and the second region of interest with the data in the knowledge base, and judging whether the radiotherapy structure of the current standard radiotherapy structure is abnormal or not through the radiotherapy structure classification model after the comparison is successful according to the characteristic parameters of the second region of interest and the second region of interest.
Based on the same inventive concept as the first embodiment, a third embodiment provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program runs, the apparatus on which the computer-readable storage medium is located is controlled to execute the improved radiotherapy structure standardized naming method as described in the first embodiment, and the same beneficial effects can be achieved.
In summary, the embodiment of the present invention has the following advantages:
obtaining a plurality of standard radiotherapy structure files by acquiring a plurality of original radiotherapy structure files and converting the original radiotherapy structure names in each original radiotherapy structure file into standard radiotherapy structure names, and dividing all the standard radiotherapy structure files into a training set and a test set; the training set only comprises standard radiotherapy structure files with normal radiotherapy structures, the testing set comprises the standard radiotherapy structure files with the normal radiotherapy structures and the standard radiotherapy structure files with abnormal radiotherapy structures, the radiotherapy structure delineation contour in each standard radiotherapy structure file in the training set is respectively extracted to serve as a first region of interest, characteristic parameters of the first region of interest are obtained, a knowledge base and a radiotherapy structure classification model are established according to the obtained characteristic parameters of all the first region of interest and all the first region of interest, each standard radiotherapy structure file in the testing set is traversed, the radiotherapy structure delineation contour in the current standard radiotherapy structure file is extracted to serve as a second region of interest, position information and characteristic parameters of the second region of interest are obtained, and when the position information of the second region of interest passes position verification, the characteristic parameters of the second region of interest and the second region of interest sequentially pass through the knowledge base and the radiotherapy structure classification model And performing structure verification, and if the verification result is that the radiotherapy structure is normal, taking the current standard radiotherapy structure file as a target radiotherapy structure file, so that the design and implementation of a radiotherapy plan can be performed by using the target radiotherapy structure. According to the embodiment of the invention, the original radiotherapy structure naming in the original radiotherapy structure file is converted into the standard radiotherapy structure naming, the radiotherapy structure naming can be subjected to standardized processing, the consistency of the radiotherapy structure naming is ensured, all standard radiotherapy structure files are divided into the training set and the testing set, the knowledge base and the radiotherapy structure classification model are established according to the radiotherapy structure delineation contour and the characteristic parameters thereof in the training set, the knowledge base and the radiotherapy structure classification model are applied to carry out structure verification on the testing set which passes the position verification, the consistency verification can be carried out on the standard radiotherapy structure naming and the radiotherapy structure delineation contour, and the matching between the standard radiotherapy structure naming and the radiotherapy structure delineation contour is ensured, so that the radiotherapy structure standardized naming is further realized, the content consistency is ensured, and the safety is improved.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
It will be understood by those skilled in the art that all or part of the processes of the above embodiments may be implemented by hardware related to instructions of a computer program, and the computer program may be stored in a computer readable storage medium, and when executed, may include the processes of the above embodiments. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
Claims (9)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111297814.2A CN114201459B (en) | 2021-11-03 | 2021-11-03 | An improved radiotherapy structure standardized naming method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111297814.2A CN114201459B (en) | 2021-11-03 | 2021-11-03 | An improved radiotherapy structure standardized naming method, device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114201459A true CN114201459A (en) | 2022-03-18 |
| CN114201459B CN114201459B (en) | 2024-06-25 |
Family
ID=80646884
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111297814.2A Active CN114201459B (en) | 2021-11-03 | 2021-11-03 | An improved radiotherapy structure standardized naming method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114201459B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117095798A (en) * | 2023-07-28 | 2023-11-21 | 广州中医药大学第一附属医院 | Deep learning data preprocessing system and electronic equipment for automatic sketching of radiotherapy |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070198935A1 (en) * | 2006-02-22 | 2007-08-23 | Mckesson Information Solutions Llc | Visual document navigation scheme |
| US20080301174A1 (en) * | 2007-03-30 | 2008-12-04 | Albert Mons | Data structure, system and method for knowledge navigation and discovery |
| CN108877951A (en) * | 2018-05-24 | 2018-11-23 | 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） | A kind of radiotherapy structure naming standard method, apparatus, equipment and medium |
| CN109448858A (en) * | 2018-10-30 | 2019-03-08 | 中山大学肿瘤防治中心 | Colorectal cancer medical data processing method and processing device, storage medium, electronic equipment |
| US20200203022A1 (en) * | 2018-01-19 | 2020-06-25 | Beijing Linking Medical Technology Co., Ltd. | Standardized Cloud Radiotherapy Planning Method, Storage Medium, and System |
| US20210103756A1 (en) * | 2019-10-08 | 2021-04-08 | Samsung Life Public Welfare Foundation | System and method for segmenting normal organ and/or tumor structure based on artificial intelligence for radiation treatment planning |
-
2021
- 2021-11-03 CN CN202111297814.2A patent/CN114201459B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070198935A1 (en) * | 2006-02-22 | 2007-08-23 | Mckesson Information Solutions Llc | Visual document navigation scheme |
| US20080301174A1 (en) * | 2007-03-30 | 2008-12-04 | Albert Mons | Data structure, system and method for knowledge navigation and discovery |
| US20200203022A1 (en) * | 2018-01-19 | 2020-06-25 | Beijing Linking Medical Technology Co., Ltd. | Standardized Cloud Radiotherapy Planning Method, Storage Medium, and System |
| CN108877951A (en) * | 2018-05-24 | 2018-11-23 | 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） | A kind of radiotherapy structure naming standard method, apparatus, equipment and medium |
| CN109448858A (en) * | 2018-10-30 | 2019-03-08 | 中山大学肿瘤防治中心 | Colorectal cancer medical data processing method and processing device, storage medium, electronic equipment |
| US20210103756A1 (en) * | 2019-10-08 | 2021-04-08 | Samsung Life Public Welfare Foundation | System and method for segmenting normal organ and/or tumor structure based on artificial intelligence for radiation treatment planning |
Non-Patent Citations (3)
| Title |
|---|
| 杨心悦;肖晓娟;卢宝兰;余深平;: "DCE-MRI定量灌注参数鉴别直肠癌病理分级的Meta分析", 中国循证医学杂志, no. 05, 25 May 2017 (2017-05-25) * |
| 罗邦雨;伍亚军;周一兵;钱金栋;赵利荣;崔天祥;钟良志;李光辉;孙建国;: "一款基于深度学习的放疗轮廓勾画软件的研发", 中国医疗设备, no. 07, 10 July 2020 (2020-07-10) * |
| 郑智满;张焜毅;高丽娟;林承光;苏阳妹;: "AAPM TG 263放射治疗结构命名的标准化", 医疗装备, no. 17, 15 September 2018 (2018-09-15) * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117095798A (en) * | 2023-07-28 | 2023-11-21 | 广州中医药大学第一附属医院 | Deep learning data preprocessing system and electronic equipment for automatic sketching of radiotherapy |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114201459B (en) | 2024-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10076299B2 (en) | Systems and methods for determining hepatic function from liver scans | |
| US20160239969A1 (en) | Methods, systems, and computer readable media for automated detection of abnormalities in medical images | |
| CN113706435A (en) | Chest enhanced CT image processing method based on traditional image omics | |
| CN114450716B (en) | Image processing for stroke characterization | |
| Paganelli et al. | Quantification of organ motion based on an adaptive image‐based scale invariant feature method | |
| CN114757953B (en) | Medical ultrasonic image recognition method, equipment and storage medium | |
| Zhuang et al. | Generic and robust method for automatic segmentation of PET images using an active contour model | |
| CN118444829B (en) | Intelligent display method, system, medium and device for tumor images | |
| CN113243914A (en) | Brain image-based assessment method and neuropsychiatric disease assessment system | |
| CN114201459A (en) | An improved radiotherapy structure standardized nomenclature method, device and storage medium | |
| TWI797767B (en) | Device and method of angiography for cerebrovascular obliteration | |
| CN115100155A (en) | Method and system for establishing radiation pneumonitis prediction model | |
| Xu et al. | Clinical-inspired framework for automatic kidney stone recognition and analysis on transverse CT images | |
| CN118570161B (en) | Medical image quality assessment method, system and computer readable medium | |
| CN107845106B (en) | Medical Image Registration Method Using Improved NNDR Strategy | |
| Zeng et al. | Abnormality detection via iterative deformable registration and basis-pursuit decomposition | |
| IL262027A (en) | System and methods for diagnostic image analysis and image quality assessment | |
| JP2024161515A (en) | Automated qualitative description of anatomical changes in radiation therapy. | |
| CN112633405A (en) | Model training method, medical image analysis device, medical image analysis equipment and medical image analysis medium | |
| Malinda et al. | Lumbar vertebrae synthetic segmentation in computed tomography images using hybrid deep generative adversarial networks | |
| CN114494132B (en) | Disease classification system based on deep learning and spatial statistical analysis of fiber tracts | |
| Werner et al. | Assessing accuracy of non-linear registration in 4D image data using automatically detected landmark correspondences | |
| CN116563270A (en) | Emphysema quantitative analysis method based on CT image random area | |
| US20230061960A1 (en) | Pulmonary function identifying method | |
| CN111383218B (en) | Medical image-based diagnostic information processing method 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |