CN117274244B

CN117274244B - Medical imaging inspection method, system and medium based on three-dimensional image recognition processing

Info

Publication number: CN117274244B
Application number: CN202311534962.0A
Authority: CN
Inventors: 王嘉宏
Original assignee: Aidipu Technology Co ltd
Current assignee: Aidipu Technology Co ltd
Priority date: 2023-11-17
Filing date: 2023-11-17
Publication date: 2024-02-20
Anticipated expiration: 2043-11-17
Also published as: CN117274244A

Abstract

Medical imaging inspection methods, systems, and media based on three-dimensional image recognition processing are provided. The method comprises the following steps: acquiring a medical detection image set of a user, extracting characteristic data set, processing to obtain image characteristic data of an abnormal part, collecting, processing to obtain an abnormal lesion risk assessment index, correcting by combining a plurality of measurement coefficients of historical similar imaging focus samples to obtain an abnormal lesion risk assessment correction index, evaluating and obtaining a focus risk judgment index of the abnormal marking part according to the acquired health detection report data set of the abnormal part and the shared record data of the user, comparing with the abnormal lesion risk assessment correction index to obtain a lesion assessment matching degree coefficient of the part, and verifying a lesion assessment result of the imaging marking part through threshold comparison; therefore, the risk condition of the abnormal lesion part is analyzed according to three-dimensional imaging identification, and the reliability of the imaging identification detection technology is compared and judged by combining the report and the risk evaluation result of shared information processing.

Description

Medical imaging inspection method, system and medium based on three-dimensional image recognition processing

Technical Field

The present application relates to the field of imaging processing and medical inspection techniques, and in particular, to a medical imaging inspection method, system, and medium based on three-dimensional image recognition processing.

Background

Medical imaging is a technology for acquiring internal tissue images of a human body or a part in a non-invasive manner, and comprises medical imaging (medical imaging) and medical imaging (medical imaging), wherein medical imaging is an image forming process and comprises research on imaging mechanisms, medical imaging is a technology for carrying out further processing on an acquired image to restore or highlight certain characteristic information in the image or carrying out classification analysis on the image, and the like, the application of the existing medical three-dimensional imaging technology is a technology for obtaining tissue image contrast or structure scanning image through contrast technology, further carrying out observation and analysis identification, and not carrying out identification tracking and analysis judgment on lesion tissues or lesion structures without symptoms, and lacks a means for carrying out auxiliary verification on the disease risk result of the imaging identification judgment.

In view of the above problems, an effective technical solution is currently needed.

Disclosure of Invention

The invention aims to provide a medical imaging inspection method, a system and a medium based on three-dimensional image recognition processing, which can realize the analysis of the risk condition of an abnormal lesion part according to three-dimensional imaging recognition, and compare and judge the reliability of an imaging recognition detection technology by combining the risk evaluation results of report and shared information processing so as to realize imaging effect inspection.

The application also provides a medical imaging inspection method based on three-dimensional image recognition processing, which comprises the following steps:

collecting a plurality of medical detection image sets of a user in a preset period for a plurality of time periods, and acquiring health detection report information of the user in the preset period and user sharing record information;

performing image feature pickup according to the medical detection image sets in each time period to obtain medical detection image information sets, and extracting corresponding medical detection image feature data sets;

extracting a health detection report data set according to the health detection report information, wherein the health detection report data set comprises abnormal part dissimilarity index data, abnormal part lesion index data, energy sub-health index data and physical and mental activity index data, and extracting user sharing record data according to the user sharing record information;

processing and identifying the corresponding medical detection image characteristic data set of the user in each time period to obtain abnormal visual detection image characteristic data, and marking the corresponding abnormal visual detection identification part as an abnormal mark tracking part;

collecting a plurality of abnormal visual inspection image characteristic data corresponding to the abnormal mark tracking position in the preset period to obtain an abnormal mark tracking position characteristic data set, and processing to obtain an abnormal lesion risk assessment index;

Performing similarity comparison through a preset focus imaging characteristic database according to the abnormal vision inspection image characteristic data of the abnormal mark tracking part to obtain a plurality of matched historical similar imaging focus samples and corresponding part lesion identification evaluation coefficients, and performing correction processing on the abnormal lesion risk evaluation index to obtain an abnormal lesion risk evaluation correction index;

according to the health detection report data set corresponding to the abnormal mark tracking position in each time period and the user shared record data, evaluating and processing through a preset focus risk data detection model, and obtaining a focus risk judgment index of the abnormal mark position;

and comparing the abnormal mark part focus risk judging index with the abnormal lesion risk evaluating correction index to obtain a part lesion evaluating matching degree coefficient of the abnormal mark tracking part, and comparing the abnormal mark part focus risk judging index with a preset focus evaluating matching degree threshold value to verify a lesion evaluating result of the imaging mark part.

Optionally, in the medical imaging inspection method based on three-dimensional image recognition processing described in the present application, the performing image feature pickup according to the medical detection image set in each time period to obtain a medical detection image information set, and extracting a corresponding medical detection image feature data set includes:

Picking up image characteristic information through a preset medical three-dimensional imaging detection recognition model according to the medical detection image set in each time period, and obtaining a medical detection image information set in each time period;

and extracting a corresponding medical detection image characteristic data set according to the medical detection image information set, wherein the medical detection image characteristic data set comprises medical detection image characteristic data corresponding to a plurality of medical detection images of a plurality of imaging detection parts.

Optionally, in the medical imaging inspection method based on three-dimensional image recognition processing described in the present application, the extracting a health detection report data set according to the health detection report information, including abnormal part dissimilarity index data and abnormal part lesion index data, and energy sub-health index data and physical and mental activity index data, extracting user sharing record data according to the user sharing record information includes:

extracting a health detection report data set according to the health detection report information, wherein the health detection report data set comprises abnormal part dissimilarity index data and abnormal part lesion index data of each abnormal report part in each time period, and energy sub-health index data and physical and mental activity index data;

and extracting user sharing record data according to the user sharing record information, wherein the user sharing record data comprises professional age attribute compensation data, medical history and medical history data and work-rest diet yield data.

Optionally, in a medical imaging inspection method based on three-dimensional image recognition processing described in the present application, the processing and recognizing the corresponding medical detection image feature data set of the user in each time period to obtain abnormal visual detection image feature data, and marking the corresponding abnormal visual detection identification part as an abnormal mark tracking part includes:

processing and identifying the corresponding medical detection image characteristic data set of the user in each time period through a preset medical imaging characteristic detection and identification model to obtain abnormal visual detection image characteristic data;

the abnormal visual inspection image characteristic data comprise abnormal part identification data, abnormal tissue outline size data, abnormal association structure data and abnormal tissue morphology description data;

and acquiring the abnormal visual inspection identification part corresponding to the identification according to the abnormal part identification data, and marking the abnormal visual inspection identification part as an abnormal mark tracking part.

Optionally, in the medical imaging inspection method based on three-dimensional image recognition processing described in the present application, the collecting the plurality of feature data of the abnormal visual inspection image corresponding to the abnormal marker tracing position in the preset period to obtain a feature data set of the abnormal marker tracing position, and processing the feature data set to obtain an abnormal lesion risk assessment index includes:

Collecting a plurality of abnormal visual inspection image characteristic data corresponding to the abnormal mark tracking position in the preset period to obtain an abnormal mark tracking position characteristic data set;

processing the feature data set of the abnormal mark tracking part through a preset abnormal lesion risk assessment model to obtain an abnormal lesion risk assessment index of the abnormal mark tracking part;

the calculation formula of the abnormal lesion risk assessment index is as follows:

；

wherein,assessing an index for risk of abnormal lesions, < >>、/>、/>Abnormal tissue outline size data, abnormal connective structure data and abnormal tissue morphology description data in the ith time period respectively +_>Presetting a lesion risk coefficient for an abnormal part, wherein n is the number of time periods in a preset period, and +.>、/>、/>Is a preset characteristic coefficient.

Optionally, in the medical imaging inspection method based on three-dimensional image recognition processing described in the present application, the performing similarity comparison according to the feature data of the abnormal visual inspection image of the abnormal mark tracking part through a preset focus imaging feature data information base to obtain a plurality of conformed historical similar imaging focus samples and corresponding part lesion identification evaluation coefficients, and performing correction processing on the abnormal lesion risk evaluation index to obtain an abnormal lesion risk evaluation correction index, including:

Performing similar contrast processing through a preset focus imaging feature data base according to the abnormal visual inspection image feature data of the abnormal mark tracking part to obtain a plurality of historical similar imaging focus samples meeting the similar contrast requirement;

extracting corresponding part lesion identification evaluation coefficients of final detection of a sample part according to the plurality of historical similar imaging focus samples;

correcting the abnormal lesion risk assessment index according to a plurality of the part lesion identification assessment coefficients to obtain an abnormal lesion risk assessment correction index;

the correction calculation formula of the abnormal lesion risk evaluation correction index is as follows:

；

wherein,correction index for evaluating risk of abnormal lesions, +.>Identifying evaluation coefficients for the part lesions of the jth historical similar imaging lesion sample, and performing +.>For the risk assessment index of abnormal lesions, m is the number of historical similar imaging lesion samples,preset characteristic coefficients of a focus sample are imaged for the j-th history similarity.

Optionally, in a medical imaging inspection method based on three-dimensional image recognition processing described in the present application, the evaluating processing is performed by a preset focus risk data detection model according to the health detection report data set corresponding to the abnormal marking and detecting part in each time period and the user sharing record data, so as to obtain a focus risk judgment index of the abnormal marking part, including:

Evaluating and processing through a preset focus risk data detection model according to the abnormal part dissimilarity index data and abnormal part lesion index data corresponding to the abnormal mark tracking part in each time period, and the energy sub-health index data and the physical and mental activity index data to obtain abnormal mark part focus risk identification data;

correcting according to the abnormal marking part focus risk identification data in combination with the job age attribute compensation data, the medical history and medical history data and the work-rest food yield data to obtain an abnormal marking part focus risk judgment index of the user in the preset period;

the calculation formula of the abnormal marking part focus risk identification data is as follows:

；

wherein,identification data of focus risk of abnormal marked part, +.>、/>、/>、/>Respectively, abnormal part dissimilation index data, abnormal part lesion index data, energy sub-health index data, physical and mental activity index data and +.>、/>、/>、/>Is a preset characteristic coefficient;

the correction calculation formula of the abnormal marking part focus risk judgment index is as follows:

；

wherein,judging index for risk of abnormal marked part focus>、/>、/>Respectively compensating data of professional age attribute, medical history data of medical history, food and drink yield data of work and rest, and +. >Identification data of focus risk of abnormal marked part, +.>、/>、/>、/>Is a preset characteristic coefficient.

Optionally, in the medical imaging inspection method based on three-dimensional image recognition processing described in the present application, the comparing the abnormal marking part focus risk judging index with the abnormal lesion risk evaluating and correcting index to obtain a part lesion evaluating and matching degree coefficient of the abnormal marking and detecting part, and comparing the part lesion evaluating and matching degree coefficient with a preset focus evaluating and matching degree threshold to verify a lesion evaluating result of the imaging marking part, including:

comparing the abnormal marked part focus risk judging index of the user with the abnormal lesion risk evaluating and correcting index to obtain a part lesion evaluating matching degree coefficient of the abnormal marked part;

threshold comparison is carried out according to the lesion assessment matching degree coefficient of the part and a preset lesion assessment matching degree threshold value, and the lesion assessment result of the imaging marked part is verified according to the threshold comparison result;

the calculation formula of the matching degree coefficient of the lesion evaluation of the part is as follows:

；

wherein,evaluating the matching degree coefficient for the lesions of the part, < >>The risk judgment index for the abnormal marked part focus, Correction index for evaluating risk of abnormal lesions, +.>、/>Is a preset characteristic coefficient.

In a second aspect, the present application provides a medical imaging inspection system based on three-dimensional image recognition processing, the system comprising: the medical imaging inspection system comprises a memory and a processor, wherein the memory comprises a program of a medical imaging inspection method based on three-dimensional image recognition processing, and the program of the medical imaging inspection method based on the three-dimensional image recognition processing realizes the following steps when being executed by the processor:

In a third aspect, the present application also provides a computer-readable storage medium, in which a medical imaging inspection method program based on a three-dimensional image recognition process is included, which when executed by a processor, implements the steps of the medical imaging inspection method based on a three-dimensional image recognition process as described in any one of the above.

As can be seen from the above, the three-dimensional image recognition processing-based medical imaging inspection method, system and medium provided by the application acquire a user medical detection image set and extract medical detection image feature data set to obtain abnormal vision inspection image feature data of an abnormal mark tracking part, integrate the processing to obtain an abnormal lesion risk assessment index, then combine a plurality of part lesion identification assessment coefficients of historical similar imaging lesion samples to correct the abnormal lesion risk assessment index, acquire an abnormal mark part lesion risk judgment index according to the acquired abnormal mark tracking part corresponding health detection report data set and combine the user shared record data assessment processing to obtain a part lesion assessment matching degree coefficient by comparing the abnormal mark part lesion risk assessment index with the abnormal lesion risk assessment correction index, and verify the lesion assessment result of the imaging mark part through threshold comparison; therefore, the risk condition of the abnormal lesion part is analyzed according to three-dimensional imaging identification, and the reliability of the imaging identification detection technology is compared and judged by combining the report and the risk evaluation result of shared information processing.

Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application. The objects and other advantages of the present application may be realized and attained by the structure particularly pointed out in the written description and drawings.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a medical imaging inspection method based on three-dimensional image recognition processing provided in an embodiment of the present application;

FIG. 2 is a flowchart of acquiring a medical detection image feature dataset according to a medical imaging inspection method based on three-dimensional image recognition processing provided in an embodiment of the present application;

FIG. 3 is a flowchart of acquiring a health detection report data set and user shared record data according to a medical imaging inspection method based on three-dimensional image recognition processing according to an embodiment of the present application;

Fig. 4 is a schematic structural diagram of a medical imaging inspection system based on three-dimensional image recognition processing according to an embodiment of the present application.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. The components of the embodiments of the present application, which are generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, as provided in the accompanying drawings, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, are intended to be within the scope of the present application.

It should be noted that like reference numerals and letters refer to like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Meanwhile, in the description of the present application, the terms "first", "second", and the like are used only to distinguish the description, and are not to be construed as indicating or implying relative importance.

Referring to fig. 1, fig. 1 is a flow chart of a medical imaging inspection method based on three-dimensional image recognition processing in some embodiments of the present application. The medical imaging inspection method based on three-dimensional image recognition processing is used in terminal equipment, such as a computer, a mobile phone terminal and the like. The medical imaging inspection method based on three-dimensional image recognition processing comprises the following steps:

s101, collecting a plurality of medical detection image sets of a user in a preset period for a plurality of time periods, and acquiring health detection report information of the user in the preset period and user sharing record information;

s102, performing image feature pickup according to the medical detection image sets in each time period to obtain medical detection image information sets, and extracting corresponding medical detection image feature data sets;

s103, extracting a health detection report data set according to the health detection report information, wherein the health detection report data set comprises abnormal part dissimilarity index data, abnormal part lesion index data, energy sub-health index data and physical and mental activity index data, and extracting user sharing record data according to the user sharing record information;

s104, processing and identifying the corresponding medical detection image characteristic data set of the user in each time period to obtain abnormal visual detection image characteristic data, and marking the corresponding abnormal visual detection identification part as an abnormal mark tracking part;

S105, collecting a plurality of abnormal visual inspection image characteristic data corresponding to the abnormal mark tracking position in the preset period to obtain an abnormal mark tracking position characteristic data set, and processing to obtain an abnormal lesion risk assessment index;

s106, performing similarity comparison through a preset focus imaging feature database according to the abnormal vision inspection image feature data of the abnormal mark tracking part to obtain a plurality of matched historical similar imaging focus samples and corresponding part lesion identification evaluation coefficients, and performing correction processing on the abnormal lesion risk evaluation index to obtain an abnormal lesion risk evaluation correction index;

s107, evaluating and processing through a preset focus risk data detection model according to the health detection report data set corresponding to the abnormal mark tracking position in each time period and the user shared record data to obtain a focus risk judgment index of the abnormal mark position;

s108, comparing the abnormal mark part focus risk judging index with the abnormal lesion risk evaluating and correcting index to obtain a part lesion evaluating matching degree coefficient of the abnormal mark tracking part, and comparing the abnormal mark part focus risk judging index with a preset focus evaluating matching degree threshold value to verify a lesion evaluating result of the imaging mark part.

It should be noted that, in order to realize that feature data is obtained by performing information recognition on a three-dimensional imaged image and judging the risk situation result of a lesion site or a tissue structure where an abnormal condition is likely to exist by analyzing, and to verify that the risk evaluation result obtained by combining a report and shared information processing is compared with the risk evaluation result obtained by performing imaging recognition evaluation, the reliability of the imaging recognition detection technology is judged to realize the technology of imaging effect inspection, a medical detection image information set is obtained by collecting a plurality of medical detection image sets obtained by a user in a plurality of time periods, a corresponding medical detection image feature data set is obtained by performing image feature pickup, then the abnormal visual detection image feature data of the abnormal visual detection site is obtained by performing recognition, and the abnormal visual detection site is marked as a marker following part, namely, the lesion site where the abnormal condition is present or the abnormal tissue structure is obtained by performing information data pickup and recognition processing on the imaging image, and the abnormal condition is obtained by performing set evaluation on a plurality of image feature data of the sites of the abnormal marker tracking monitoring site in the time period, namely, a lesion site is obtained by performing image matching with the actual condition history feature matching evaluation by performing image matching history feature matching, and obtaining a lesion feature matching history database, and obtaining a similarity feature matching database is obtained by performing image matching evaluation, and then carrying out correction processing on the abnormal lesion risk assessment index through an assessment coefficient of a sample to obtain a more accurate assessment correction result of the abnormal lesion risk, in order to verify the assessment accuracy of the lesion risk of the abnormal part judged by the identification of imaging information data, carrying out assessment by combining the data of a health detection report in a period of time with the data of a user sharing record through a detection model to obtain a judgment result of the lesion risk of the abnormal part, carrying out contrast processing on the obtained abnormal mark part lesion risk assessment index and the abnormal lesion risk assessment correction index to obtain a part lesion assessment matching degree coefficient of the abnormal mark part, namely obtaining a risk assessment result of the abnormal lesion part identified by imaging through the contrast of the part lesion risk assessment of imaging data and the result of the lesion risk judgment of the report record data, and finally carrying out threshold contrast with a preset lesion identification matching degree threshold value to verify the reliability of the identification assessment of the lesion assessment result of the imaging mark part.

Referring to fig. 2, fig. 2 is a flowchart of acquiring a medical detection image feature dataset according to a medical imaging inspection method based on three-dimensional image recognition processing in some embodiments of the present application. According to the embodiment of the invention, the image feature pickup is performed according to the medical detection image set in each time period to obtain a medical detection image information set, and the corresponding medical detection image feature data set is extracted, specifically:

s201, picking up image characteristic information through a preset medical three-dimensional imaging detection recognition model according to the medical detection image set in each time period, and obtaining a medical detection image information set in each time period;

s202, extracting a corresponding medical detection image characteristic data set according to the medical detection image information set, wherein medical detection image characteristic data corresponding to a plurality of medical detection images comprising a plurality of imaging detection parts.

It should be noted that, in order to identify a lesion site where an abnormal lesion may occur through information data of an imaging image, firstly, an image information extraction is required to be performed on a medical detection image set synthesized by medical detection images in each time period acquired in a certain period, so as to further obtain feature data of the image, and an image feature data set is obtained by collecting the image feature data of each time period, according to the medical detection image set in each time period, image feature information is picked up through a preset medical three-dimensional imaging detection recognition model, and integrated into the medical detection image feature information set of each time period, where the medical three-dimensional imaging detection recognition model is an image feature recognition model obtained through a third-party medical three-dimensional image detection platform and used for picking up image features of a medical three-dimensional image, such as an image drawing outline, a size, a structure, a morphology, a coordinate, a structure and the like in the three-dimensional image can be extracted through the model, and then a corresponding medical detection image feature data set is extracted according to the medical detection image feature information set, including medical detection image feature data corresponding to a plurality of imaging detection sites such as local tissues, organs and a plurality of medical detection images of the system in each time period, and the feature data is the image feature data reflecting the image feature data describing the outline, the tissue, the structure and the feature data.

Referring to fig. 3, fig. 3 is a flowchart of acquiring a health detection report data set and user shared record data according to a medical imaging inspection method based on three-dimensional image recognition processing in some embodiments of the present application. According to the embodiment of the invention, the health detection report data set is extracted according to the health detection report information, and comprises abnormal part dissimilarity index data, abnormal part lesion index data, energy sub-health index data and physical and mental activity index data, and user sharing record data is extracted according to the user sharing record information, specifically:

s301, extracting a health detection report data set according to the health detection report information, wherein the health detection report data set comprises abnormal part dissimilarity index data and abnormal part lesion index data of each abnormal report part in each time period, and energy sub-health index data and physical and mental activity index data;

s302, extracting user sharing record data according to the user sharing record information, wherein the user sharing record data comprises professional age attribute compensation data, medical history and medical history data and work-rest diet yield data.

It should be noted that, to verify the accuracy of the relevant evaluation results of the lesion risk of the part identified and judged by the imaging information data of the user, the data of the health detection report of the user in the preset period is further processed in combination with the data of the shared record of the user, the risk judgment of the lesion condition of the part is obtained by processing the physicochemical detection results of the detection report in combination with the investigation registration information of the user, the risk evaluation results of the abnormal lesion site identified by imaging are assisted and verified, the health detection report data set is extracted according to the health detection report information in the period of the user, including the part dissimilarity index and the part lesion index of each part detected by the various abnormal reports reported in each period, and the sub-health detection index and the physical and mental activity detection index of the energy health condition of the user in each period, and the shared record information of the shared record of the user is collected at the same time, and the shared record data of the user is extracted, including the attribute compensation coefficient data of the occupation and age of the user, i.e. the compensation coefficient reflecting the attribute of the occupation and age of the user, the data of the family medical history and the health record and the good rule of the information.

According to the embodiment of the invention, the processing and identifying the corresponding medical detection image feature data set of the user in each time period to obtain the abnormal visual detection image feature data, and marking the corresponding abnormal visual detection identification part as an abnormal mark tracking part comprises the following specific steps:

After obtaining the medical detection image characteristic data of a plurality of imaging detection parts, processing and identifying the corresponding medical detection image characteristic data set of a user in each time period through a preset medical imaging characteristic detection and identification model to obtain the abnormal vision detection image characteristic data, wherein the medical imaging characteristic detection and identification model is obtained through training a large number of characteristic data of historical imaging samples and samples of the abnormal characteristic data, the abnormal image characteristic data in the imaging image can be identified through the model, and comprises delineating identification data of the abnormal part, outline size data of the abnormal tissue, description data of an abnormal connective structure and morphological description data of tissue structure, phase and spectrum of the abnormal change tissue, and the abnormality identification part captured by the identification corresponding to the characteristic data is marked to trace the abnormal marking part, so that the identification and subsequent tracking are convenient.

According to an embodiment of the present invention, the collecting the feature data of the plurality of abnormal visual inspection images corresponding to the abnormal mark tracking position in the preset period to obtain the feature data set of the abnormal mark tracking position, and processing the feature data set to obtain the risk assessment index of the abnormal lesion specifically includes:

；

wherein,assessing an index for risk of abnormal lesions, < >>、/>、/>Abnormal tissue outline size data, abnormal connective structure data and abnormal tissue morphology description data in the ith time period respectively +_>Presetting a lesion risk coefficient for an abnormal part, wherein n is the number of time periods in a preset period, and +.>、/>、/>Is a preset characteristic coefficient (the lesion risk coefficient and the characteristic coefficient are obtained through query of a preset lesion imaging characteristic data information base).

In order to evaluate and identify the lesion risk condition of the identified abnormal identification part, the method performs integration and evaluation processing on a plurality of image feature data of the abnormal mark tracking monitoring part in each period to obtain an abnormal lesion risk evaluation index of the abnormal part in the period, that is, a lesion risk evaluation result of lesion assembly of the abnormal mark part is obtained through calculation and evaluation on the feature data of the abnormal part, and a calculation formula of a preset abnormal lesion risk evaluation model is used to calculate and process the integration of the feature data of the abnormal mark tracking part in each period to obtain an abnormal lesion risk evaluation index reflecting the lesion condition of the abnormal part.

According to the embodiment of the invention, the abnormal vision inspection image characteristic data of the abnormal mark tracking part is subjected to similar comparison through a preset focus imaging characteristic data information base to obtain a plurality of matched historical similar imaging focus samples and corresponding part lesion identification evaluation coefficients, and the abnormal lesion risk evaluation index is subjected to correction processing to obtain an abnormal lesion risk evaluation correction index, wherein the method specifically comprises the following steps:

；

wherein,correction index for evaluating risk of abnormal lesions, +.>Identifying evaluation coefficients for the part lesions of the jth historical similar imaging lesion sample, and performing +.>For the risk assessment index of abnormal lesions, m is the number of historical similar imaging lesion samples,lesion sample for jth history similar imagingIs obtained by querying a preset lesion imaging feature database.

It should be noted that, further, in order to improve the accuracy of lesion risk assessment of the anomaly identification tracking monitoring part, the risk assessment index is corrected by using the final lesion identification assessment coefficient result of the anomaly part similarity historical sample obtained by similarity query, so as to improve the accuracy of the anomaly part risk assessment, similarity comparison is performed by using a preset lesion imaging feature database according to the anomaly vision image feature data of the anomaly identification tracking part, the similarity comparison can be cosine similarity comparison or Euclidean distance similarity comparison, a plurality of historical similar imaging lesion samples meeting the preset similarity comparison requirement are obtained by using the similarity comparison, and corresponding part lesion identification assessment coefficients, namely, a plurality of similar samples meeting the requirement and actual lesion assessment coefficients are found by using the feature data of the anomaly part through a historical database, and then correction calculation is performed on the anomaly lesion risk assessment index according to the assessment coefficients of the similar samples, so as to obtain a more accurate assessment correction result reflecting the lesion risk of the anomaly part.

According to the embodiment of the invention, the health detection report data set corresponding to the abnormal mark tracking position in each time period is combined with the user shared record data to be evaluated and processed through a preset focus risk data detection model, so as to obtain a focus risk judgment index of the abnormal mark position, which is specifically as follows:

；

wherein,identification data of focus risk of abnormal marked part, +.>、/>、/>、/>Respectively, abnormal part dissimilation index data, abnormal part lesion index data, energy sub-health index data, physical and mental activity index data and +. >、/>、/>、/>Is a preset characteristic coefficient;

；

wherein,judging index for risk of abnormal marked part focus>、/>、/>Respectively compensating data of professional age attribute, medical history data of medical history, food and drink yield data of work and rest, and +.>For the risk identification data of the abnormal marking part focus,、/>、/>、/>is a preset characteristic coefficient (the characteristic coefficient is obtained through inquiring a preset focus imaging characteristic data base).

In order to verify the accuracy of the evaluation of the lesion risk of the abnormal part judged by the identification of the imaging information data of the three-dimensional image, the physical and chemical detection report data of the health detection report of the user in the period is combined with the relevant information data of the user registration sharing record, the judgment result of the focus risk of the abnormal part is obtained by calculation through a preset detection model, the focus risk identification data of the abnormal marked part is obtained by calculation according to a plurality of health detection report data corresponding to the abnormal marked tracking part in each time period, and then the focus risk identification data of the abnormal marked part is corrected and calculated with the shared record data of the user, so that the focus risk judgment index of the user in the preset period is obtained, and the technology for verifying the result data of imaging detection and identification through the detection report and the data registered by the user is realized.

According to the embodiment of the invention, the comparing treatment is performed according to the abnormal marking part focus risk judging index and the abnormal lesion risk evaluating correction index to obtain the part lesion evaluating matching degree coefficient of the abnormal marking tracking part, and the threshold comparing verification is performed with the preset focus evaluating matching degree threshold to verify the lesion evaluating result of the imaging marking part, specifically:

；

wherein,evaluating the matching degree coefficient for the lesions of the part, < >>The risk judgment index for the abnormal marked part focus,correction index for evaluating risk of abnormal lesions, +.>、/>Is a preset characteristic coefficient (the characteristic coefficient is obtained by inquiring a preset focus imaging characteristic data base And (3) obtaining).

Finally, comparing the obtained lesion risk judging index of the abnormal mark position with the abnormal lesion risk evaluating and correcting index to obtain a lesion evaluation matching degree coefficient of the abnormal mark position, namely, comparing the lesion risk evaluating result of the abnormal mark position judged in the medical test image of the user with a lesion risk identifying result of the corresponding position obtained by evaluating the detection data information of the physicochemical detection report and the registration record of the user through a preset calculation formula to obtain a lesion evaluation matching degree result, comparing the lesion evaluation matching degree coefficient of the abnormal mark position with a preset lesion evaluation matching degree threshold value according to the lesion evaluation matching degree coefficient of the abnormal mark position, and verifying the lesion evaluation result of the imaging mark position according to the threshold value comparison result, thereby verifying the authenticity of the lesion of the abnormal mark position identified in the medical test image and the accuracy of the evaluation result of the real lesion risk condition of the abnormal mark position.

As shown in fig. 4, the invention also discloses a medical imaging inspection system 4 based on three-dimensional image recognition processing, which comprises a memory 41 and a processor 42, wherein the memory comprises a medical imaging inspection method program based on three-dimensional image recognition processing, and the medical imaging inspection method program based on three-dimensional image recognition processing realizes the following steps when being executed by the processor:

According to the embodiment of the invention, the image feature pickup is performed according to the medical detection image set in each time period to obtain a medical detection image information set, and the corresponding medical detection image feature data set is extracted, specifically:

According to the embodiment of the invention, the health detection report data set is extracted according to the health detection report information, and comprises abnormal part dissimilarity index data, abnormal part lesion index data, energy sub-health index data and physical and mental activity index data, and user sharing record data is extracted according to the user sharing record information, specifically:

；

wherein,correction index for evaluating risk of abnormal lesions, +.>Identifying evaluation coefficients for the part lesions of the jth historical similar imaging lesion sample, and performing +.>For the risk assessment index of abnormal lesions, m is the number of historical similar imaging lesion samples,and (3) the preset characteristic coefficient of the jth historical similar imaging focus sample (the characteristic coefficient is obtained through inquiring a preset focus imaging characteristic database).

；

wherein,identification data of focus risk of abnormal marked part, +.>、/>、/>、/>Respectively, abnormal part dissimilation index data, abnormal part lesion index data, energy sub-health index data, physical and mental activity index data and +. >、/>、/>、/>Is a preset characteristic coefficient; />

；

wherein,judging index for risk of abnormal marked part focus>、/>、/>Respectively compensating data of professional age attribute, medical history data of medical history, food and drink yield data of work and rest, and +.>Identification data of focus risk of abnormal marked part, +.>、/>、/>、/>Is a preset characteristic coefficient (the characteristic coefficient is obtained through inquiring a preset focus imaging characteristic data base).

；

wherein,evaluating the matching degree coefficient for the lesions of the part, < >>The risk judgment index for the abnormal marked part focus,is an abnormal lesionRisk assessment correction index,/->、/>Is a preset characteristic coefficient (the characteristic coefficient is obtained through inquiring a preset focus imaging characteristic data base).

A third aspect of the present invention provides a readable storage medium having embodied therein a medical imaging verification method program based on a three-dimensional image recognition process, which when executed by a processor, implements the steps of the medical imaging verification method based on a three-dimensional image recognition process as set forth in any one of the above.

The invention discloses a medical imaging inspection method, a system and a medium based on three-dimensional image recognition processing, which are characterized in that a user medical detection image set is obtained, medical detection image feature data set is extracted to process to obtain abnormal vision inspection image feature data of an abnormal mark pursuit part, the abnormal lesion risk assessment index is obtained through integrated processing, then a plurality of part lesion identification assessment coefficients of historical similar imaging lesion samples are combined to correct to obtain an abnormal lesion risk assessment correction index, the user shared recordation data assessment processing is combined to obtain an abnormal mark part lesion risk judgment index according to the obtained abnormal mark pursuit part corresponding health detection report data set, then the abnormal mark focus assessment match coefficient is obtained through comparison with the abnormal lesion risk assessment correction index, and the lesion assessment result of the imaging mark part is verified through threshold comparison; therefore, the risk condition of the abnormal lesion part is analyzed according to three-dimensional imaging identification, and the reliability of the imaging identification detection technology is compared and judged by combining the report and the risk evaluation result of shared information processing.

In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.

The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units; can be located in one place or distributed to a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated in one processing unit, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units.

Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware related to program instructions, and the foregoing program may be stored in a readable storage medium, where the program, when executed, performs steps including the above method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk or an optical disk, or the like, which can store program codes.

Alternatively, the above-described integrated units of the present invention may be stored in a readable storage medium if implemented in the form of software functional modules and sold or used as separate products. Based on such understanding, the technical solution of the embodiments of the present invention may be embodied in essence or a part contributing to the prior art in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, ROM, RAM, magnetic or optical disk, or other medium capable of storing program code.

Claims

1. The medical imaging inspection method based on three-dimensional image recognition processing is characterized by comprising the following steps of:

performing image feature pickup on the medical detection image sets in each time period to obtain medical detection image information sets, and extracting corresponding medical detection image feature data sets;

performing similarity comparison on the abnormal visual inspection image characteristic data of the abnormal mark tracking part and data in a preset focus imaging characteristic data base to obtain a plurality of historical similar imaging focus samples meeting the requirements and corresponding part lesion identification evaluation coefficients, and performing correction processing on the abnormal lesion risk evaluation index to obtain an abnormal lesion risk evaluation correction index;

Evaluating the health detection report data set and the user shared record data corresponding to the abnormal mark tracking position in each time period through a preset focus risk data detection model to obtain a focus risk judgment index of the abnormal mark position;

and comparing the abnormal mark part focus risk judging index with the abnormal lesion risk evaluating correction index to obtain a part lesion evaluating matching degree coefficient of the abnormal mark tracking part, and comparing the part lesion evaluating matching degree coefficient with a preset focus evaluating matching degree threshold value to verify a lesion evaluating result of the imaging mark part.

2. The three-dimensional image recognition processing-based medical imaging inspection method according to claim 1, wherein the performing image feature pickup on the medical detection image sets in each period of time to obtain medical detection image information sets and extracting corresponding medical detection image feature data sets includes:

picking up image characteristic information of the medical detection image set in each time period through a preset medical three-dimensional imaging detection recognition model to obtain a medical detection image information set in each time period;

3. The three-dimensional image recognition processing-based medical imaging examination method according to claim 2, wherein the extracting a health detection report data set including abnormal region dissimilarity index data and abnormal region lesion index data, and vigor sub-health index data and physical and mental activity index data according to the health detection report information, extracting user sharing record data according to the user sharing record information, comprises:

and extracting user sharing record data according to the user sharing record information, wherein the user sharing record data comprises professional age attribute compensation data, family medical history data and work-rest diet yield data.

4. A medical imaging inspection method based on three-dimensional image recognition processing according to claim 3, wherein said processing and recognizing the corresponding medical detection image feature data set of the user in each time period to obtain abnormal visual detection image feature data, and marking the corresponding abnormal visual detection identification part as an abnormal mark tracking part comprises:

5. The method according to claim 4, wherein the step of collecting the plurality of abnormality detection image feature data corresponding to the abnormality marker following region in the predetermined period to obtain an abnormality marker following region feature data set, and processing the same to obtain an abnormality lesion risk assessment index, comprises:

；

wherein,assessing an index for risk of abnormal lesions, < >>、/>、/>Abnormal tissue outline size data and abnormality in the ith time period respectivelyAssociative structural data, allo-tissue morphology data,>presetting a lesion risk coefficient for an abnormal part, wherein n is the number of time periods in a preset period, and +.>、/>、/>Is a preset characteristic coefficient.

6. The three-dimensional image recognition processing-based medical imaging inspection method according to claim 5, wherein the performing similar comparison between the abnormal visual inspection image characteristic data of the abnormal mark tracking part and the data in the preset focus imaging characteristic data information base to obtain a plurality of history similar imaging focus samples and corresponding part lesion identification evaluation coefficients meeting the requirements, and performing correction processing on the abnormal lesion risk evaluation index to obtain an abnormal lesion risk evaluation correction index comprises:

performing similar comparison processing on the abnormal visual inspection image characteristic data of the abnormal mark tracking part and data in a preset focus imaging characteristic data information base to obtain a plurality of historical similar imaging focus samples meeting the similar comparison requirement;

；

wherein,correction index for evaluating risk of abnormal lesions, +.>Identifying evaluation coefficients for the part lesions of the jth historical similar imaging lesion sample, and performing +.>For the risk assessment index of abnormal lesions, m is the number of historical similar imaging lesion samples, and +.>Preset characteristic coefficients of a focus sample are imaged for the j-th history similarity.

7. The medical imaging inspection method based on three-dimensional image recognition processing according to claim 6, wherein the evaluating the health detection report data set and the user shared record data corresponding to the abnormal marking and detecting part in each time period by a preset focus risk data detection model to obtain a focus risk judgment index of the abnormal marking part comprises:

evaluating the abnormal part dissimilarity index data and abnormal part lesion index data corresponding to the abnormal mark tracking part in each time period and the energy sub-health index data and physical and mental activity index data through a preset focus risk data detection model to obtain focus risk identification data of the abnormal mark part;

Correcting according to the abnormal marking part focus risk identification data, the job age attribute compensation data, the family medical history data and the work and rest food yield data to obtain an abnormal marking part focus risk judgment index of the user in the preset period;

；

wherein,judging index for risk of abnormal marked part focus>、/>、/>Respectively compensating data of professional age attribute, medical history data of family medical history, food and drink yield data of work and rest, and +.>Identification data of focus risk of abnormal marked part, +.>、/>、/>、/>Is a preset characteristic coefficient.

8. The three-dimensional image recognition processing-based medical imaging inspection method according to claim 7, wherein comparing the abnormal marking part focus risk judgment index with the abnormal lesion risk evaluation correction index to obtain a part lesion evaluation matching degree coefficient of the abnormal marking test part, and comparing the part lesion evaluation matching degree coefficient with a preset focus evaluation matching degree threshold value, and verifying a lesion evaluation result of the imaging marking part comprises:

Comparing the abnormal marked part focus risk judging index of the user with the abnormal lesion risk evaluating and correcting index to obtain a part lesion evaluating matching degree coefficient of the abnormal marked and detected part;

threshold comparison is carried out on the lesion assessment matching degree coefficient of the part and a preset lesion assessment matching degree threshold value, and the lesion assessment result of the imaging marked part is verified according to the threshold comparison result;

；

wherein,evaluating the matching degree coefficient for the lesions of the part, < >>Judging index for risk of abnormal marked part focus>Correction index for evaluating risk of abnormal lesions, +.>、/>Is a preset characteristic coefficient.

9. A medical imaging verification system based on three-dimensional image recognition processing, the system comprising: the medical imaging inspection system comprises a memory and a processor, wherein the memory comprises a program of a medical imaging inspection method based on three-dimensional image recognition processing, and the program of the medical imaging inspection method based on the three-dimensional image recognition processing realizes the following steps when being executed by the processor:

10. A computer-readable storage medium, characterized in that a medical imaging examination method program based on a three-dimensional image recognition process is included in the computer-readable storage medium, which when executed by a processor, implements the steps of the medical imaging examination method based on a three-dimensional image recognition process according to any one of claims 1 to 8.