CN113674254B - Medical image outlier recognition method, apparatus, electronic device, and storage medium - Google Patents
Medical image outlier recognition method, apparatus, electronic device, and storage medium Download PDFInfo
- Publication number
- CN113674254B CN113674254B CN202110983114.2A CN202110983114A CN113674254B CN 113674254 B CN113674254 B CN 113674254B CN 202110983114 A CN202110983114 A CN 202110983114A CN 113674254 B CN113674254 B CN 113674254B
- Authority
- CN
- China
- Prior art keywords
- image
- outlier
- body part
- abnormal point
- abnormal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 230000002159 abnormal effect Effects 0.000 claims abstract description 153
- 238000013135 deep learning Methods 0.000 claims abstract description 27
- 206010028980 Neoplasm Diseases 0.000 claims description 48
- 238000012549 training Methods 0.000 claims description 17
- 238000013136 deep learning model Methods 0.000 claims description 16
- 230000011218 segmentation Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 12
- 230000002503 metabolic effect Effects 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 8
- 238000013507 mapping Methods 0.000 claims description 6
- 230000036267 drug metabolism Effects 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 7
- 239000003814 drug Substances 0.000 description 7
- 230000008901 benefit Effects 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 6
- 229940079593 drug Drugs 0.000 description 6
- 238000003384 imaging method Methods 0.000 description 6
- 230000003902 lesion Effects 0.000 description 6
- 238000002591 computed tomography Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 238000003709 image segmentation Methods 0.000 description 3
- 239000012216 imaging agent Substances 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000001959 radiotherapy Methods 0.000 description 2
- 239000000243 solution Substances 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 108010001267 Protein Subunits Proteins 0.000 description 1
- 210000001015 abdomen Anatomy 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 210000003484 anatomy Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- WQZGKKKJIJFFOK-UKLRSMCWSA-N dextrose-2-13c Chemical compound OC[C@H]1OC(O)[13C@H](O)[C@@H](O)[C@@H]1O WQZGKKKJIJFFOK-UKLRSMCWSA-N 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000037406 food intake Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 229930182851 human metabolite Natural products 0.000 description 1
- 238000002347 injection Methods 0.000 description 1
- 239000007924 injection Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000004066 metabolic change Effects 0.000 description 1
- 230000004060 metabolic process Effects 0.000 description 1
- 206010061289 metastatic neoplasm Diseases 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000002600 positron emission tomography Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000009885 systemic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10104—Positron emission tomography [PET]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The application relates to a medical image outlier identification method, equipment, an electronic device and a storage medium, wherein the medical image outlier identification method comprises the following steps: acquiring a plurality of body part images of a scanning object based on the scanning image of the scanning object, identifying abnormal points in the body part images according to a preset medical parameter threshold value to obtain first abnormal points, identifying the abnormal points in the plurality of body part images based on a deep learning algorithm to obtain second abnormal points, and determining target abnormal points corresponding to the scanning object according to the first abnormal points and the second abnormal points. The method solves the problem of lower efficiency of manually sketching abnormal points of the medical image in the related technology, and effectively improves the efficiency of identifying the abnormal points of the medical image.
Description
Technical Field
The present application relates to the technical field of medical devices, and in particular, to a method, an apparatus, an electronic device, and a storage medium for identifying abnormal points of medical images.
Background
Positron emission computed tomography (Positron Emission Tomography, abbreviated as PET) systems are imaging systems that reflect the genetic, molecular, metabolic and functional status of lesions. The PET system uses positron nuclide labeled glucose and other human metabolites as imaging agents, and reflects metabolic changes of the imaging agents through the ingestion of focus, so that biological metabolic information of diseases is provided for clinic. An electronic computer tomography (Computed Tomography, abbreviated as CT) system uses X-rays to perform a body layer examination of a human body, so that anatomical structure information of lesions can be clearly obtained. The PET-CT system integrates the PET system and the CT system on one instrument to form a complete imaging system, a patient can obtain CT anatomical images and PET functional metabolic images simultaneously through rapid whole body scanning during examination, and advantages of the two images are complementary, so that a doctor can obtain accurate anatomical positioning while knowing biological metabolic information, and the disease can be comprehensively and accurately judged.
After the PET-CT system inspection, abnormal points in the scanned images are checked as suspected focus to diagnose, and in general, the checking process is manually finished by doctors, and when the scanned images are more, manual checking is time-consuming and labor-consuming, and the efficiency is lower.
At present, an effective solution is not proposed for the problem of lower efficiency of manually drawing abnormal points of medical images in the related art.
Disclosure of Invention
The embodiment of the application provides a medical image abnormal point identification method, equipment, an electronic device and a storage medium, which at least solve the problem of low efficiency in the related art by manually realizing the sketching of the abnormal points of the medical image.
In a first aspect, an embodiment of the present application provides a method for identifying abnormal points in a medical image, including:
acquiring a plurality of body part images of a scanning object based on a scanning image of the scanning object;
identifying abnormal points in the body part image according to a preset medical parameter threshold value to obtain a first abnormal point, wherein the medical parameter threshold value has a corresponding relation with the body part;
Identifying abnormal points in the body part images based on a deep learning algorithm to obtain second abnormal points;
and determining a target abnormal point corresponding to the scanning object according to the first abnormal point and the second abnormal point.
In some embodiments, the determining a target outlier corresponding to the scan object according to the first outlier and the second outlier includes:
acquiring the tumor identification requirement of the scanned object;
And obtaining the target abnormal point of the scanning object according to the type of the tumor identification requirement, the first abnormal point and the second abnormal point.
In some embodiments, the obtaining the target outlier of the scan object according to the type of tumor recognition requirement, the first outlier, and the second outlier comprises:
And under the condition that the type of the tumor recognition requirement is the priority of the tumor recognition false alarm rate, determining the target abnormal point according to the intersection of the first abnormal point and the second abnormal point.
In some embodiments, the obtaining the target outlier of the scan object according to the type of tumor recognition requirement, the first outlier, and the second outlier comprises:
And under the condition that the type of the tumor recognition requirement is that the tumor recognition report missing rate is priority, determining the target abnormal point according to the union of the first abnormal point and the second abnormal point.
In some of these embodiments, the acquiring a plurality of body part images of the scan object based on the scan image of the scan object comprises:
Acquiring a CT image and a PET image of the scanned object;
Acquiring a plurality of body parts of the scanning object in the CT image;
And acquiring a plurality of body part images of the scanning object in the PET image according to the mapping relation between the CT image and the PET image and the plurality of body parts.
In some of these embodiments, after the determining the target outlier corresponding to the scan object, the method further comprises:
Acquiring a preset medical record configuration file;
and generating the electronic medical record of the scanning object according to the parameter information of the target abnormal point and the preset medical record configuration file.
In some of these embodiments, the deep learning algorithm is implemented by a deep learning model, and the training process of the deep learning model includes:
taking the body part image in the scanned image and the body part image marked with the abnormal points as a training set;
training a deep learning algorithm according to the training set until the loss function of the deep learning algorithm converges.
In a second aspect, an embodiment of the present application provides a medical image outlier identification apparatus, including an acquisition module, a segmentation module, an identification module, and a determination module:
the acquisition module is used for acquiring a plurality of body part images of a scanning object based on the scanning image of the scanning object;
The segmentation module is used for identifying abnormal points in the body part image according to a preset medical parameter threshold value to obtain a first abnormal point, wherein the medical parameter threshold value has a corresponding relation with the body part;
The recognition module is used for recognizing abnormal points in the body part images based on a deep learning algorithm to obtain second abnormal points;
the determining module is used for determining a target abnormal point corresponding to the scanning object according to the first abnormal point and the second abnormal point.
In a third aspect, an embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the method for identifying abnormal points of medical images according to the first aspect when the processor executes the computer program.
In a fourth aspect, an embodiment of the present application provides a storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the medical image outlier identification method as described in the first aspect above.
Compared with the related art, the medical image outlier identification method provided by the embodiment of the application has the advantages that the images of the plurality of body parts of the scanned object are acquired based on the scanned image of the scanned object, the outlier in the images of the body parts is identified according to the preset medical parameter threshold value to obtain the first outlier, the outlier in the images of the plurality of body parts is identified based on the deep learning algorithm to obtain the second outlier, the target outlier corresponding to the scanned object is determined according to the first outlier and the second outlier, the problems that the sketching of the outlier of the medical image is manually realized and the efficiency is low in the related art are solved, and the medical image outlier identification efficiency is effectively improved.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the other features, objects, and advantages of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:
FIG. 1 is a schematic view of an application environment of a medical image outlier recognition method according to an embodiment of the present application;
FIG. 2 is a flow chart of a medical image outlier identification method according to an embodiment of the present application;
FIG. 3 is a flow chart of a method of acquiring an image of a body part according to an embodiment of the present application;
Fig. 4 is a hardware structure block diagram of a terminal of a medical image outlier recognition method according to an embodiment of the present application;
Fig. 5 is a block diagram of a medical image outlier recognition apparatus according to an embodiment of the present application.
Detailed Description
The present application will be described and illustrated with reference to the accompanying drawings and examples in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application. All other embodiments, which can be made by a person of ordinary skill in the art based on the embodiments provided by the present application without making any inventive effort, are intended to fall within the scope of the present application. Moreover, it should be appreciated that while such a development effort might be complex and lengthy, it would nevertheless be a routine undertaking of design, fabrication, or manufacture for those of ordinary skill having the benefit of this disclosure, and thus should not be construed as having the benefit of this disclosure.
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is to be expressly and implicitly understood by those of ordinary skill in the art that the described embodiments of the application can be combined with other embodiments without conflict.
Unless defined otherwise, technical or scientific terms used herein should be given the ordinary meaning as understood by one of ordinary skill in the art to which this application belongs. The terms "a," "an," "the," and similar referents in the context of the application are not to be construed as limiting the quantity, but rather as singular or plural. The terms "comprising," "including," "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, article, or apparatus that comprises a list of steps or modules (elements) is not limited to only those steps or elements but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus. The terms "connected," "coupled," and the like in connection with the present application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The term "plurality" as used herein means greater than or equal to two. "and/or" describes an association relationship of an association object, meaning that there may be three relationships, e.g., "a and/or B" may mean: a exists alone, A and B exist together, and B exists alone. The terms "first," "second," "third," and the like, as used herein, are merely distinguishing between similar objects and not representing a particular ordering of objects.
The medical image outlier identification method provided by the application can be applied to an application environment shown in fig. 1, and fig. 1 is a schematic diagram of the application environment of the medical image outlier identification method according to an embodiment of the application, as shown in fig. 1. The medical image scanning system comprises a scanning device 101, a scanning bed 102, a host 103 and a reconstruction machine 104, and a doctor controls the scanning device 101 to scan a patient on the scanning bed 102 through the host 103 to obtain scanning data of the patient. The host 103 sends the acquired scanning data to the reconstruction machine 104 for image reconstruction to finally obtain a scanning image, the host 103 acquires a plurality of body part images of a scanning object based on the scanning image, then identifies abnormal points in the body part images according to a preset medical parameter threshold to obtain a first abnormal point, and identifies the abnormal points in the plurality of body part images based on a deep learning algorithm to obtain a second abnormal point; a target outlier corresponding to the scanned object is determined based on the first outlier and the second outlier.
The embodiment provides a medical image outlier identification method. Fig. 2 is a flowchart of a medical image outlier recognition method according to an embodiment of the present application, as shown in fig. 2, the method including the steps of:
step S210, acquiring a plurality of body part images of the scan subject based on the scan image of the scan subject.
The scan object in this embodiment may be a human body or an animal, and the scan image may be acquired by a medical image scanning system, such as a PET system, a CT system, or a PET-CT system, or the like.
After the scanned image is acquired, the body part in the scanned image may be identified and segmented by an artificial intelligence (ARTIFICIAL INTELLIGENCE, abbreviated as AI) technique, so as to obtain a body part image of the scanned object, such as a head image, a chest-abdomen image, and the like of the scanned object. Specifically, the body part image may be acquired by an image segmentation algorithm using a trained image segmentation model. The image segmentation algorithm is a process of dividing an image into mutually disjoint areas and comprises a segmentation method based on a threshold value, a segmentation method based on an area, a segmentation method based on an edge, a segmentation method based on a specific theory and the like.
Step S220, identifying abnormal points in the body part image according to a preset medical parameter threshold value to obtain a first abnormal point, wherein the medical parameter threshold value has a corresponding relation with the body part.
In the application, abnormal points in the body of a scanned object need to be screened, wherein the abnormal points refer to regions of interest where lesions may occur, such as tumor lesions. During drug metabolism, the abnormal point is different in response to the drug than the surrounding normal body tissue pair. In the present embodiment, the outliers obtained by different ways are divided into a first outlier and a second outlier.
The medical parameter in this embodiment is preferably a drug metabolism index, such as a standard uptake value (Standard Uptake Value, abbreviated as SUV), which is a semi-quantitative index commonly used in the diagnosis of tumor by PET systems, and refers to the ratio of the radioactivity of the imaging agent taken by local tissues to the average systemic injection activity. Accordingly, since the abnormal point is different from the normal body tissue in metabolism of the drug, the threshold can be set according to the metabolic condition of the drug, thereby obtaining the preset medical parameter threshold. Then, a first outlier is identified based on the comparison of the metabolic condition of the drug by the body part with the medical parameter threshold. It should be noted that the metabolic condition of drugs is different for different body parts, and thus, different medical parameter thresholds need to be set for different body parts. In the identification of the first outlier, it is also necessary to identify it based on different medical parameter thresholds.
Specifically, in the case where the medical parameter is SUV, a plurality of SUV thresholds corresponding to the number of body parts are set according to the demand for body parts. Further, for a determined body part, if the SUV threshold is set to 1, then in the body part image, the pixel with SUV smaller than 1 is set to 0, and the pixel with SUV greater than or equal to 1 is set to 1, so that the segmentation of the body part image is realized, and the first abnormal point is identified.
Step S230, identifying abnormal points in the plurality of body part images based on the deep learning algorithm, so as to obtain second abnormal points.
In this embodiment, the abnormal points in the body part image may also be identified by a deep learning algorithm. Specifically, the deep learning algorithm, after training, may identify outliers directly from the input body-part images. In particular, the second outlier may be identified based on differences in drug absorption capacity of different scanned objects, different body parts, and different body parts.
Step S240, determining a target abnormal point corresponding to the scanning object according to the first abnormal point and the second abnormal point.
In the present embodiment, the number of the first outlier, the second outlier, and the target outlier is not limited. The target outlier is the outlier that the scanned object eventually needs to analyze.
Through the steps S210 to S240, the embodiment identifies the abnormal points in the body part image based on different abnormal point identification modes, wherein the abnormal point identification process based on the medical parameter threshold and the abnormal point identification process based on the deep learning algorithm can be completed through the host computer of the medical image scanning system, so that the problem of low efficiency caused by manually implementing the sketching of the abnormal points of the medical image in the related technology is solved, and the efficiency of identifying the abnormal points of the medical image is effectively improved.
In some embodiments, the deep learning algorithm in the present embodiment is implemented by a deep learning model, where the deep learning model in the present embodiment is a trained model, and the training process of the deep learning model includes: the body part image in the scanned image and the body part image marked with the abnormal points are used as training sets, wherein the process of marking the abnormal points in the body part image can be marked in advance by engineers, and then the deep learning algorithm is trained according to the training sets until the loss function of the deep learning algorithm is converged. The loss function is used for evaluating the degree that the predicted value and the true value of the deep learning model are different, and the loss function converges, so that the recognition accuracy of the deep learning model reaches the expected requirement.
Further, in this embodiment, corresponding deep learning models may be trained on different body part images, and for each deep learning model, a serialized file of parameters and model structures of the model is finally obtained, where each deep learning model is stored in the system as a configuration file, and when different body part images are selected for recognition, the corresponding deep learning model is loaded. And finally, assembling the output results of different deep learning models into the recognition result of the second abnormal point of the whole scanning object so as to improve the recognition accuracy of the deep learning algorithm.
In some of these embodiments, the scan subject determines the tumor recognition requirement prior to scanning, so that the final target outlier can be determined based on the tumor recognition requirement of the scan subject. Specifically, a tumor recognition requirement of a scanned object is acquired, a corresponding software program is set according to the tumor recognition requirement to form a software package, then the software package is operated, and a target abnormal point of the scanned object is obtained according to the type of the tumor recognition requirement, the first abnormal point and the second abnormal point. The tumor recognition requirement in this embodiment is mainly used for determining priorities of different recognition types in the tumor recognition process, where the recognition types include a false positive rate of tumor recognition and a false negative rate of tumor recognition, where the false positive rate of tumor recognition has a higher requirement on accuracy of abnormal point recognition, for example, when radiotherapy needs to be performed on a scanned object, accurate screening of lesions related to tumor is required, and the false positive rate of tumor recognition has a higher requirement on integrity of abnormal point recognition, for example, when preliminary screening of tumor needs to be performed on the scanned object, focus missing report needs to be avoided. In this embodiment, the final target outlier is determined according to the tumor recognition requirement, so as to improve the scene adaptability of outlier recognition.
Further, under the condition that the type of the tumor recognition requirement is that the false alarm rate of tumor recognition is priority, determining a target abnormal point according to the intersection of the first abnormal point and the second abnormal point; and determining the target abnormal point according to the union of the first abnormal point and the second abnormal point under the condition that the type of the tumor recognition requirement is that the tumor recognition report missing rate is priority.
In the application, the first abnormal point and the second abnormal point are obtained through different identification modes, so that the first abnormal point and the second abnormal point have overlapped abnormal points and different abnormal points. In this embodiment, when the tumor recognition requirement is that the tumor recognition false alarm rate is prioritized, for example, a radiotherapy task, an abnormal point which is coincident with the first abnormal point and the second abnormal point is determined through intersection as a final target abnormal point, so as to obtain a more accurate target abnormal point, when the tumor recognition requirement is that the tumor recognition false alarm rate is prioritized, for example, a scanning object needs to perform preliminary screening on a tumor, and all the first abnormal point and the second abnormal point are taken as target abnormal points through union, so as to obtain a whole body target abnormal point, for example, a metastatic tumor and a tumor in an organ, so as to avoid false alarm.
In some embodiments, if the medical image outlier recognition is performed based on the PET apparatus or the CT apparatus, the body part positioning and outlier recognition may be performed directly through the corresponding PET image or CT image, and if the medical image outlier recognition is performed based on the PET-CT system, fig. 3 is a flowchart of a method for acquiring the body part image according to an embodiment of the present application, as shown in fig. 3, the method includes the following steps:
In step S310, a CT image and a PET image of the scan object are acquired.
In a PET-CT system, the CT system may perform structural imaging and the PET system may perform functional imaging.
In step S320, a plurality of body parts of the scan object are acquired in the CT image.
After acquiring CT scan images in a digital imaging and communication (DIGITAL IMAGING AND Communications IN MEDICINE, abbreviated as DICOM) format of medicine, the CT scan images can be segmented by an AI segmentation algorithm to obtain segmentation results of different body parts.
Step S330, acquiring a plurality of body part images of the scan object in the PET image according to the mapping relation between the CT image and the PET image and the plurality of body parts.
In the PET-CT system, the mapping relationship between the PET image and the CT image may be stored in advance, and the mapping relationship is specifically a coordinate conversion relationship of pixels in the PET image and the CT image. Therefore, based on the segmentation result of the body part image in the CT image, the body part can be correspondingly positioned in the PET image, and finally a plurality of body part images are obtained.
Through the steps S310 to S330, the body part image with more accurate positioning is obtained based on the correspondence between the CT image and the PET image in the present embodiment.
Further, in the case that the medical parameter is SUV, it is necessary to convert the PET image into an SUV result image first, and on this basis, perform the identification of the first outlier and the second outlier. Specifically, the body part images are segmented based on medical parameter thresholds corresponding to different body parts to obtain first abnormal points, the identified body part images are respectively input into corresponding deep learning models for prediction according to classification of head, neck, chest and abdomen, and finally the identification result of second abnormal points is obtained.
In some embodiments, after determining the target abnormal point corresponding to the scan object, an electronic case of the scan object may be further generated according to the target abnormal point, which specifically includes: firstly, a preset case history configuration file is acquired, wherein a plurality of case history parameters and normal ranges of the case history parameters are set in the case history configuration file, and the case history parameters are such as personal information of a scanning object, a body part of an abnormal point of the scanning object, the number of target abnormal points, a volume of interest (Volume of Interest, VOI for short) and the like. And then generating the electronic medical record of the scanning object according to the parameter information of the target abnormal point and the preset medical record configuration file. The parameter information of the target outlier includes personal information of the scanned object, a body part to which the target outlier belongs, a VOI of the target outlier, and the like, respectively.
Specifically, a diagnosis conclusion corresponding to the target abnormal point can be obtained according to a comparison result of the parameter information of the target abnormal point and the normal range of the medical record parameter, or corresponding parameter information in the target abnormal point can be directly stored according to the medical record parameter information in the medical record configuration file, and finally the electronic medical record is generated. Preferably, the electronic medical record can be stored in a DICOM-format structured report file, a sequence corresponding to the scanned image is formed and stored in a database, so that a doctor can diagnose a target abnormal point of a scanned object conveniently.
Furthermore, the electronic medical record and the scanned image can be loaded and displayed by the advanced application interface, and the medical record can be archived and printed.
It should be noted that the steps illustrated in the above-described flow or flow diagrams of the figures may be performed in a computer system, such as a set of computer-executable instructions, and that, although a logical order is illustrated in the flow diagrams, in some cases, the steps illustrated or described may be performed in an order other than that illustrated herein.
The method embodiment provided by the application can be executed in a terminal, a computer or similar computing device. Taking the operation on the terminal as an example, fig. 4 is a block diagram of the hardware structure of the terminal of the medical image outlier recognition method according to the embodiment of the present application. As shown in fig. 4, the terminal 40 may include one or more (only one is shown in fig. 4) processors 402 (the processor 402 may include, but is not limited to, a microprocessor MCU or a processing device such as a programmable logic device FPGA) and a memory 404 for storing data, and optionally, a transmission device 406 for communication functions and an input-output device 408. It will be appreciated by those skilled in the art that the structure shown in fig. 4 is merely illustrative and is not intended to limit the structure of the terminal. For example, terminal 40 may also include more or fewer components than shown in fig. 4, or have a different configuration than shown in fig. 4.
The memory 404 may be used to store control programs, such as software programs of application software and modules, such as control programs corresponding to the medical image outlier recognition method in the embodiment of the present application, and the processor 402 executes the control programs stored in the memory 404, thereby performing various functional applications and data processing, that is, implementing the above-mentioned method. Memory 404 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, memory 404 may further include memory located remotely from processor 402, which may be connected to terminal 40 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The transmission device 406 is used to receive or transmit data via a network. The specific examples of the network described above may include a wireless network provided by a communication provider of the terminal 40. In one example, the transmission device 406 includes a network adapter (Network Interface Controller, simply referred to as a NIC) that may connect to other network devices through a base station to communicate with the internet. In one example, the transmission device 406 may be a Radio Frequency (RF) module, which is configured to communicate with the internet wirelessly.
The embodiment also provides a medical image abnormal point identifying device, which is used for implementing the above embodiment and the preferred embodiment, and is not described in detail. As used below, the terms "module," "unit," "sub-unit," and the like may be a combination of software and/or hardware that implements a predetermined function. While the means described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.
Fig. 5 is a block diagram of a medical image outlier recognition apparatus according to an embodiment of the present application, and as shown in fig. 5, the apparatus includes an acquisition module 51, a segmentation module 52, a recognition module 53, and a determination module 54:
An acquisition module 51 for acquiring a plurality of body part images of a scan subject based on a scan image of the scan subject;
The segmentation module 52 is configured to identify an abnormal point in the body part image according to a preset medical parameter threshold, so as to obtain a first abnormal point, where the medical parameter threshold has a corresponding relationship with the body part;
the identifying module 53 is configured to identify an abnormal point in the multiple body part images based on a deep learning algorithm, so as to obtain a second abnormal point;
a determining module 54, configured to determine a target outlier corresponding to the scanned object according to the first outlier and the second outlier.
According to the medical image abnormal point identification device, the abnormal points in the body part image are identified based on different abnormal point identification modes through the identification module 53, wherein the abnormal point identification process based on the medical parameter threshold and the abnormal point identification process based on the deep learning algorithm can be completed through the host computer of the medical image scanning system, so that the problem that the efficiency is low due to the fact that the sketching of the abnormal points of the medical image is manually achieved in the related art is solved, and the efficiency of identifying the abnormal points of the medical image is effectively improved.
Further, the determining module 54 is further configured to obtain a tumor recognition requirement of the scan object, and obtain a target outlier of the scan object according to the type of the tumor recognition requirement, the first outlier, and the second outlier, so as to improve scene adaptability of outlier recognition.
Further, in the case that the type of the tumor recognition requirement is that the tumor recognition false alarm rate is prioritized, the determining module 54 determines the target outlier according to the intersection of the first outlier and the second outlier, so as to obtain a more accurate target outlier; in the case that the type of the tumor recognition requirement is that the tumor recognition report missing rate is prioritized, the determining module 54 determines the target outlier according to the union of the first outlier and the second outlier, so as to obtain the target outlier of the whole body.
Further, the acquisition module 51 is further configured to acquire a CT image and a PET image of the scanned object; acquiring a plurality of body parts of a scanning object in a CT image; according to the mapping relation between the CT image and the PET image and the plurality of body parts, a plurality of body part images which are more accurate in positioning of the scanning object are acquired in the PET image.
Further, the medical image outlier identification device further comprises a medical record generation module, wherein the medical record generation module is used for acquiring a preset medical record configuration file; and generating an electronic medical record of the scanning object according to the parameter information of the target abnormal point and a preset medical record configuration file, so that a doctor can diagnose the target abnormal point of the scanning object conveniently.
Further, the deep learning algorithm is realized through a deep learning model, and the medical image outlier recognition device further comprises a training module, wherein the training module is used for taking the body part image in the scanned image and the body part image marked with outliers as a training set; training the deep learning algorithm according to the training set until the loss function of the deep learning algorithm converges.
The above-described respective modules may be functional modules or program modules, and may be implemented by software or hardware. For modules implemented in hardware, the various modules described above may be located in the same processor; or the above modules may be located in different processors in any combination.
The present embodiment also provides an electronic device comprising a memory having stored therein a computer program and a processor arranged to run the computer program to perform the steps of any of the method embodiments described above.
Optionally, the electronic apparatus may further include a transmission device and an input/output device, where the transmission device is connected to the processor, and the input/output device is connected to the processor.
Alternatively, in the present embodiment, the above-described processor may be configured to execute the following steps by a computer program:
s1, acquiring a plurality of body part images of a scanning object based on the scanning image of the scanning object;
S2, identifying abnormal points in the body part image according to a preset medical parameter threshold value to obtain a first abnormal point, wherein the medical parameter threshold value has a corresponding relation with the body part;
s3, identifying abnormal points in the body part images based on a deep learning algorithm to obtain second abnormal points;
s4, determining a target abnormal point corresponding to the scanning object according to the first abnormal point and the second abnormal point.
It should be noted that, specific examples in this embodiment may refer to examples described in the foregoing embodiments and alternative implementations, and this embodiment is not repeated herein.
In addition, in combination with the medical image outlier identification method in the above embodiment, the embodiment of the present application may be implemented by providing a storage medium. The storage medium has a computer program stored thereon; the computer program when executed by a processor implements the method of lesion recognition for medical image outlier recognition in any of the above embodiments.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of protection of the present application is to be determined by the appended claims.
Claims (10)
1. A method for identifying abnormal points in a medical image, comprising:
acquiring a plurality of body part images of a scanning object based on a scanning image of the scanning object;
Identifying abnormal points in the body part image according to a preset medical parameter threshold value to obtain a first abnormal point, wherein the medical parameter threshold value has a corresponding relation with the body part; the medical parameter includes a drug metabolism indicator; identifying an outlier in the body part image according to a preset medical parameter threshold, wherein obtaining a first outlier comprises: identifying a first outlier based on a comparison of the metabolic condition of the body part to the medical parameter threshold;
Identifying abnormal points in the body part images based on a deep learning algorithm to obtain second abnormal points;
and determining a target abnormal point corresponding to the scanning object according to the first abnormal point and the second abnormal point.
2. The method of claim 1, wherein determining a target outlier corresponding to the scan object from the first outlier and the second outlier comprises:
acquiring the tumor identification requirement of the scanned object;
And obtaining the target abnormal point of the scanning object according to the type of the tumor identification requirement, the first abnormal point and the second abnormal point.
3. The method of claim 2, wherein the obtaining the target outlier of the scan object from the type of tumor recognition requirement, the first outlier, and the second outlier comprises:
And under the condition that the type of the tumor recognition requirement is the priority of the tumor recognition false alarm rate, determining the target abnormal point according to the intersection of the first abnormal point and the second abnormal point.
4. The method of claim 2, wherein the obtaining the target outlier of the scan object from the type of tumor recognition requirement, the first outlier, and the second outlier comprises:
And under the condition that the type of the tumor recognition requirement is that the tumor recognition report missing rate is priority, determining the target abnormal point according to the union of the first abnormal point and the second abnormal point.
5. The method of claim 1, wherein the acquiring a plurality of body part images of the scan subject based on the scan image of the scan subject comprises:
Acquiring a CT image and a PET image of the scanned object;
Acquiring a plurality of body parts of the scanning object in the CT image;
And acquiring a plurality of body part images of the scanning object in the PET image according to the mapping relation between the CT image and the PET image and the plurality of body parts.
6. The medical image outlier identification method according to claim 1, wherein after said determining a target outlier corresponding to the scan object, the method further comprises:
Acquiring a preset medical record configuration file;
and generating the electronic medical record of the scanning object according to the parameter information of the target abnormal point and the preset medical record configuration file.
7. The method for identifying abnormal points of medical images according to any one of claims 1 to 6, wherein the deep learning algorithm is implemented by a deep learning model, and the training process of the deep learning model includes:
taking the body part image in the scanned image and the body part image marked with the abnormal points as a training set;
training a deep learning algorithm according to the training set until the loss function of the deep learning algorithm converges.
8. The medical image abnormal point identification device is characterized by comprising an acquisition module, a segmentation module, an identification module and a determination module:
the acquisition module is used for acquiring a plurality of body part images of a scanning object based on the scanning image of the scanning object;
The segmentation module is used for identifying abnormal points in the body part image according to a preset medical parameter threshold value to obtain a first abnormal point, wherein the medical parameter threshold value has a corresponding relation with the body part; the medical parameter includes a drug metabolism indicator; identifying an outlier in the body part image according to a preset medical parameter threshold, wherein obtaining a first outlier comprises: identifying a first outlier based on a comparison of the metabolic condition of the body part to the medical parameter threshold;
The recognition module is used for recognizing abnormal points in the body part images based on a deep learning algorithm to obtain second abnormal points;
the determining module is used for determining a target abnormal point corresponding to the scanning object according to the first abnormal point and the second abnormal point.
9. An electronic device comprising a memory and a processor, wherein the memory has stored therein a computer program, the processor being arranged to run the computer program to perform the medical image outlier identification method of any one of claims 1 to 7.
10. A storage medium, characterized in that the storage medium has stored therein a computer program, wherein the computer program is arranged to perform the steps of the medical image outlier recognition method according to any one of claims 1 to 7 when run.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110983114.2A CN113674254B (en) | 2021-08-25 | 2021-08-25 | Medical image outlier recognition method, apparatus, electronic device, and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110983114.2A CN113674254B (en) | 2021-08-25 | 2021-08-25 | Medical image outlier recognition method, apparatus, electronic device, and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113674254A CN113674254A (en) | 2021-11-19 |
| CN113674254B true CN113674254B (en) | 2024-05-14 |
Family
ID=78546350
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110983114.2A Active CN113674254B (en) | 2021-08-25 | 2021-08-25 | Medical image outlier recognition method, apparatus, electronic device, and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113674254B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114820591B (en) * | 2022-06-06 | 2023-02-21 | 北京医准智能科技有限公司 | Image processing method, image processing apparatus, electronic device, and medium |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006119340A2 (en) * | 2005-05-04 | 2006-11-09 | Imquant, Inc. | Dynamic tumor diagnostic and treatment system |
| CN109509183A (en) * | 2018-10-31 | 2019-03-22 | 上海联影医疗科技有限公司 | Screening technique, device, medical supply, Medical Imaging System and the storage medium of medical image |
| CN110310257A (en) * | 2019-05-31 | 2019-10-08 | 上海联影医疗科技有限公司 | Medical image processing method, device, computer equipment and storage medium |
| CN111583184A (en) * | 2020-04-14 | 2020-08-25 | 上海联影智能医疗科技有限公司 | Image analysis method, network, computer device, and storage medium |
| WO2020182033A1 (en) * | 2019-03-08 | 2020-09-17 | 腾讯科技（深圳）有限公司 | Image region positioning method and apparatus and medical image processing device |
| CN111904379A (en) * | 2020-07-13 | 2020-11-10 | 上海联影医疗科技有限公司 | Scanning method and device of multi-modal medical equipment |
| CN111951278A (en) * | 2020-07-31 | 2020-11-17 | 上海联影智能医疗科技有限公司 | Method for segmenting medical images and computer-readable storage medium |
| CN112037147A (en) * | 2020-09-02 | 2020-12-04 | 上海联影医疗科技有限公司 | Medical image noise reduction method and device |
| WO2021004157A1 (en) * | 2019-07-09 | 2021-01-14 | 上海联影医疗科技有限公司 | Medical scanning imaging method and apparatus, storage medium, and computer device |
| WO2021030629A1 (en) * | 2019-08-14 | 2021-02-18 | Genentech, Inc. | Three dimensional object segmentation of medical images localized with object detection |
| CN112508086A (en) * | 2019-12-20 | 2021-03-16 | 上海联影智能医疗科技有限公司 | System and method for classifying medical images |
| WO2021081841A1 (en) * | 2019-10-30 | 2021-05-06 | 未艾医疗技术(深圳)有限公司 | Splenic tumor recognition method based on vrds 4d medical image, and related apparatus |
| CN113177928A (en) * | 2021-05-18 | 2021-07-27 | 数坤(北京)网络科技股份有限公司 | Image identification method and device, electronic equipment and storage medium |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040122702A1 (en) * | 2002-12-18 | 2004-06-24 | Sabol John M. | Medical data processing system and method |
| US7412280B2 (en) * | 2003-07-01 | 2008-08-12 | Ge Medical Systems Global Technology Company, Llc | Systems and methods for analyzing an abnormality of an object |
| US8238624B2 (en) * | 2007-01-30 | 2012-08-07 | International Business Machines Corporation | Hybrid medical image processing |
| US10282588B2 (en) * | 2016-06-09 | 2019-05-07 | Siemens Healthcare Gmbh | Image-based tumor phenotyping with machine learning from synthetic data |
| US11443433B2 (en) * | 2018-02-10 | 2022-09-13 | The Trustees Of The University Of Pennsylvania | Quantification and staging of body-wide tissue composition and of abnormal states on medical images via automatic anatomy recognition |
| US10929973B2 (en) * | 2018-10-02 | 2021-02-23 | Siemens Healtcare Gmbh | Medical image pre-processing at the scanner for facilitating joint interpretation by radiologists and artificial intelligence algorithms |
| US11010938B2 (en) * | 2019-04-03 | 2021-05-18 | Uih America, Inc. | Systems and methods for positron emission tomography image reconstruction |
-
2021
- 2021-08-25 CN CN202110983114.2A patent/CN113674254B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006119340A2 (en) * | 2005-05-04 | 2006-11-09 | Imquant, Inc. | Dynamic tumor diagnostic and treatment system |
| CN109509183A (en) * | 2018-10-31 | 2019-03-22 | 上海联影医疗科技有限公司 | Screening technique, device, medical supply, Medical Imaging System and the storage medium of medical image |
| WO2020182033A1 (en) * | 2019-03-08 | 2020-09-17 | 腾讯科技（深圳）有限公司 | Image region positioning method and apparatus and medical image processing device |
| CN110310257A (en) * | 2019-05-31 | 2019-10-08 | 上海联影医疗科技有限公司 | Medical image processing method, device, computer equipment and storage medium |
| WO2021004157A1 (en) * | 2019-07-09 | 2021-01-14 | 上海联影医疗科技有限公司 | Medical scanning imaging method and apparatus, storage medium, and computer device |
| WO2021030629A1 (en) * | 2019-08-14 | 2021-02-18 | Genentech, Inc. | Three dimensional object segmentation of medical images localized with object detection |
| WO2021081841A1 (en) * | 2019-10-30 | 2021-05-06 | 未艾医疗技术(深圳)有限公司 | Splenic tumor recognition method based on vrds 4d medical image, and related apparatus |
| CN112508086A (en) * | 2019-12-20 | 2021-03-16 | 上海联影智能医疗科技有限公司 | System and method for classifying medical images |
| CN111583184A (en) * | 2020-04-14 | 2020-08-25 | 上海联影智能医疗科技有限公司 | Image analysis method, network, computer device, and storage medium |
| CN111904379A (en) * | 2020-07-13 | 2020-11-10 | 上海联影医疗科技有限公司 | Scanning method and device of multi-modal medical equipment |
| CN111951278A (en) * | 2020-07-31 | 2020-11-17 | 上海联影智能医疗科技有限公司 | Method for segmenting medical images and computer-readable storage medium |
| CN112037147A (en) * | 2020-09-02 | 2020-12-04 | 上海联影医疗科技有限公司 | Medical image noise reduction method and device |
| CN113177928A (en) * | 2021-05-18 | 2021-07-27 | 数坤(北京)网络科技股份有限公司 | Image identification method and device, electronic equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 心电信号异常识别分类研究与实现;沈艺珊;《硕士电子期刊》;20191231(第07期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113674254A (en) | 2021-11-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111904379B (en) | Scanning method and device for multi-mode medical equipment | |
| CN111325714B (en) | Method for processing region of interest, computer device and readable storage medium | |
| CN111161270A (en) | Blood vessel segmentation method for medical image, computer device and readable storage medium | |
| CN110717905B (en) | Brain image detection method, computer device, and storage medium | |
| CN111080584A (en) | Quality control method for medical image, computer device and readable storage medium | |
| CN107123095A (en) | A kind of PET image reconstruction method, imaging system | |
| CN111462115B (en) | Medical image display method and device and computer equipment | |
| KR102202398B1 (en) | Image processing apparatus and image processing method thereof | |
| CN111887878B (en) | PET scanning method | |
| US20210374452A1 (en) | Method and device for image processing, and elecrtonic equipment | |
| CN110853082A (en) | Medical image registration method and device, electronic equipment and computer storage medium | |
| CN111128345B (en) | Medical image acquisition method, medical scanning device and computer storage medium | |
| CN112037147B (en) | Medical image noise reduction method and device | |
| CN111612689B (en) | Medical image processing method, medical image processing device, computer equipment and readable storage medium | |
| CN111445550B (en) | Iterative reconstruction method, device and computer readable storage medium for PET image | |
| CN113674254B (en) | Medical image outlier recognition method, apparatus, electronic device, and storage medium | |
| CN110136076A (en) | Medical scanning imaging method, device, storage medium and computer equipment | |
| KR101885562B1 (en) | Method for mapping region of interest in first medical image onto second medical image and apparatus using the same | |
| CN112102235A (en) | Human body part recognition method, computer device, and storage medium | |
| US11010893B2 (en) | Image identifying method and image identifying device | |
| CN112200780B (en) | Bone tissue positioning method, device, computer equipment and storage medium | |
| CN113160199B (en) | Image recognition method and device, computer equipment and storage medium | |
| CN111544022B (en) | PET attenuation correction method and device and computer equipment | |
| CN116630239A (en) | Image analysis method, device and computer equipment | |
| CN112669450B (en) | Human body model construction method and personalized human body model construction method |
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 |