CN113822917A - Accurate registration method for liver cancer imaging omics images - Google Patents

Accurate registration method for liver cancer imaging omics images Download PDF

Info

Publication number
CN113822917A
CN113822917A CN202111145960.3A CN202111145960A CN113822917A CN 113822917 A CN113822917 A CN 113822917A CN 202111145960 A CN202111145960 A CN 202111145960A CN 113822917 A CN113822917 A CN 113822917A
Authority
CN
China
Prior art keywords
image
registered
images
registration
feature information
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.)
Withdrawn
Application number
CN202111145960.3A
Other languages
Chinese (zh)
Inventor
陈超
王武杰
王维
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Second Hospital of Shandong University
Original Assignee
Second Hospital of Shandong University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Second Hospital of Shandong University filed Critical Second Hospital of Shandong University
Priority to CN202111145960.3A priority Critical patent/CN113822917A/en
Publication of CN113822917A publication Critical patent/CN113822917A/en
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
    • G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
    • 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
    • G06T7/0014—Biomedical image inspection using an image reference approach
    • 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
    • G06T2207/10088—Magnetic resonance imaging [MRI]
    • 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
    • G06T2207/10—Image acquisition modality
    • G06T2207/10116—X-ray image
    • 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
    • G06T2207/10132—Ultrasound image
    • 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
    • G06T2207/20024—Filtering details
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30056—Liver; Hepatic
    • 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)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical 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)
  • Measuring And Recording Apparatus For Diagnosis (AREA)

Abstract

The invention is suitable for the technical field of medical images, and provides a precise registration method of liver cancer imaging omics images, which is used for assisting a doctor in distinguishing and analyzing the liver cancer imaging omics images and solving the problems that key information is adopted for feature matching in the prior art, the coverage area is narrow, and tiny lesions cannot be found; the method comprises the following steps: acquiring characteristic information of an image to be registered; extracting feature information of an image to be registered, and determining a set of salient feature information points; the method comprises the steps of extracting the set of the significant characteristic information points, inputting the significant characteristic information points into an image accurate registration model to obtain an accurate registration result, and inputting the significant characteristic information points into the image accurate registration model to obtain the accurate registration result.

Description

Accurate registration method for liver cancer imaging omics images
Technical Field
The invention belongs to the technical field of medical images, and particularly relates to a precise registration method of liver cancer imaging omics images.
Background
When medical image analysis is performed, several images of the same patient are often put together for analysis, so that comprehensive information of multiple aspects of the patient is obtained, the level of medical diagnosis and treatment is improved, quantitative analysis is performed on several different images, and firstly, the problem of strict alignment of the several images is solved, namely, the registration of the images. Medical image registration refers to the search for a (or a series of) spatial transformation for one medical image to spatially coincide with a corresponding point on the other medical image, where the coincidence is that the same anatomical point on the body has the same spatial position on the two matching images, and the result of the registration is that all anatomical points, or at least all points of diagnostic interest and points of surgical interest, on the two images are matched.
With the continuous updating and development of medical imaging equipment, for the same patient, pathological cell structure information of the patient can be collected in multiple angles and multiple aspects through multiple imaging technologies such as CT, MRI and the like, in the liver cancer treatment process, the patient relapse condition can be discovered as early as possible, the disease condition can be predicted, the life safety and the health of the patient can be guaranteed, for the traditional diagnosis means in the past, doctors with abundant experience are often needed, and through observation and comprehensive analysis of different images in multiple aspects and combination of subjective experience and space imagination, the relevant prediction results can be made.
Disclosure of Invention
The invention provides an accurate registration method of liver cancer imaging omics images, and aims to solve the problems that key information is adopted for feature matching, the coverage area is narrow, and tiny lesions cannot be found in the prior art.
The invention is realized in this way, a liver cancer imaging omics image accurate registration method is used for assisting a doctor in distinguishing and analyzing the liver cancer imaging omics image, and comprises the following steps:
acquiring characteristic information of an image to be registered;
extracting the feature information of the image to be registered, comparing the feature information of the image to be registered with a preset standard reference image at a plurality of feature information points, and determining a set of significant feature information points;
and extracting the set of the salient feature information points, inputting the salient feature information points into an accurate image registration model to obtain an accurate registration result, and feeding back the registration result.
Preferably, the image feature information to be registered includes: the liver appearance, the liver texture, the liver gray scale texture characteristic data and the liver intensity characteristic data presented by the magnetic anatomical imaging, the CT image, the PET image, the X-ray image and the ultrasonic image.
Preferably, the specific step of acquiring the feature information of the image to be registered includes:
detecting an image acquisition instruction, and acquiring a liver photo;
uploading the liver photo, converting the liver photo into an electronic image according to an initial image conversion model, and extracting the feature information of the image to be registered;
and uploading the feature information of the image to be registered.
Preferably, the implementation method of the initial image conversion model comprises:
acquiring a large number of liver photo samples;
and training by using the picture of the liver as input and the characteristic information of the liver image as output to obtain an initial image conversion model.
Preferably, the extracting the feature information of the image to be registered, comparing the feature information of the image to be registered with a preset standard reference image at a plurality of feature information points, and determining the set of significant feature information points includes:
extracting characteristic information of the image to be registered, reconstructing the image to be registered, and storing the image to be registered;
extracting a standard reference image and an image to be registered, performing denoising and de-graying on the image to be registered through a Gaussian filtering algorithm to obtain a simulated image to be registered, performing three-dimensional image standardization on the simulated image to be registered, and comparing the simulated image to be registered with the standard reference image;
and determining whether the simulated image to be registered meets the standard deviation of the standardized data according to the standard reference image, outputting the salient feature information points if the simulated image to be registered meets the standard deviation of the standardized data, and storing the salient feature information point set.
Preferably, the specific step of comparing the standard reference image comprises:
determining a three-dimensional image salient registration point according to the determined simulation image to be registered and the standard reference image, and comparing the three-dimensional image salient registration point with the standard reference image to obtain a data difference to be registered;
acquiring a mean deviation value between a data difference to be registered and a standard deviation of standardized data of different three-dimensional image salient registration points as an interference set;
and traversing all data in the interference set, meeting the standard deviation of the standardized data when the data in the interference set is smaller than a preset threshold value of a standard reference image, and outputting the information points of the salient features.
Preferably, the specific implementation steps of extracting the salient feature information point set, inputting the salient feature information points into an image accurate registration model to obtain an accurate registration result, and feeding back the registration result include:
extracting a set of salient feature information points;
establishing a precise registration model of the salient feature information point input image;
and inputting the salient feature information points into the image accurate registration model, calculating registration loss according to the image accurate registration model, judging whether the registration loss is less than a standard loss threshold value, and if so, performing accurate registration.
Preferably, the method for establishing the accurate image registration model comprises the following steps:
acquiring a plurality of standard training images;
inputting each group of standard training images into an initial image registration network, respectively performing one-to-one registration to obtain a plurality of image deformation fields, and selecting five groups of image deformation fields with large difference as candidate images;
and calculating the image deviation degree of the candidate image and the reference image in the image registration network, judging whether the image deviation degree meets a preset condition, and if the image deviation degree meets the preset condition, determining the selected image as a plurality of images to be registered for accurate registration.
Preferably, the initial image conversion model implementation method further includes:
and obtaining an initial image conversion model, wherein the initial image conversion method takes photos presented in different directions and angles as variables, screens and calculates the photos through an RANSAC algorithm to obtain a feature point set with image feature points, and extracts an image block set with overlapped feature information through the feature point set to form the initial image conversion model.
Preferably, the method for establishing the accurate image registration model further comprises the following steps:
if the standard training images do not meet the preset conditions, acquiring a plurality of standard training images, inputting each group of standard training images into an initial image registration network, respectively performing one-to-one registration to obtain a plurality of image deformation fields, and selecting five groups of image deformation fields with large difference as candidate images;
and calculating the constraint deviation of each group of deformation fields, comparing the constraint deviation to be registered with the standard constraint deviation preset by the initial image registration network, and if the constraint deviation to be registered meets the preset conditions, determining the selected image as a plurality of images to be registered for accurate registration.
Compared with the prior art, the embodiment of the application mainly has the following beneficial effects:
according to the liver cancer imaging omics image accurate registration method provided by the invention, the obvious characteristic information points are input into the image accurate registration model to obtain an accurate registration result, so that the problems that key information is adopted for characteristic matching, the coverage area is narrow and tiny lesions cannot be found in the prior art are solved, the diagnosis efficiency of doctors is improved, and the treatment work of patients is facilitated.
Drawings
Fig. 1 is a schematic structural diagram of a precise registration method for hepatoma imaging omics images provided by the present invention.
Fig. 2 is a schematic view of an implementation process for acquiring feature information of an image to be registered according to the present invention.
Fig. 3 is a schematic flow chart of an implementation process of the initial image conversion model provided by the present invention.
Fig. 4 is a schematic diagram of an implementation process for determining a set of salient feature information points according to comparison between feature information of an image to be registered and a preset standard reference image at a plurality of feature information points, and the implementation process for extracting feature information of the image to be registered is provided by the present invention.
Fig. 5 is a schematic diagram of an implementation process of comparing standard reference images provided by the present invention.
Fig. 6 is a schematic view of an implementation flow for obtaining an accurate registration result and feeding back the registration result according to the present invention.
Fig. 7 is a schematic flow chart of the implementation of establishing the accurate image registration model provided by the invention.
Detailed Description
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "including" and "having," and any variations thereof, in the description and claims of this application and the description of the above figures are intended to cover non-exclusive inclusions. The terms "first," "second," and the like in the description and claims of this application or in the above-described drawings are used for distinguishing between different objects and not for describing a particular order.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase 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 explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
The embodiment of the invention provides a precise registration method of liver cancer imagery omics images, and figure 1 shows an implementation flow of the precise registration method of the liver cancer imagery omics images, and the precise registration method of the liver cancer imagery omics images comprises the following specific steps:
step S100, acquiring characteristic information of an image to be registered;
step S200, extracting the feature information of the image to be registered, comparing the feature information of the image to be registered with a preset standard reference image at a plurality of feature information points, and determining a set of significant feature information points;
and step S300, extracting the set of the salient feature information points, inputting the salient feature information points into an image accurate registration model to obtain an accurate registration result, and feeding back the registration result.
In an embodiment of the present invention, the information about the feature of the image to be registered includes: the method comprises the steps of obtaining liver appearance, liver texture, liver gray scale texture feature data and liver intensity feature data presented by nuclear magnetic anatomical imaging, CT imaging, PET imaging, X-ray imaging and ultrasonic imaging, wherein the nuclear magnetic anatomical imaging, the CT imaging, the PET imaging, the X-ray imaging and the ultrasonic imaging are all in the prior art, and rapidly collecting feature images through the nuclear magnetic anatomical imaging, the CT imaging, the PET imaging, the X-ray imaging and the ultrasonic imaging.
The obvious characteristic information points are input into the accurate image registration model to obtain an accurate registration result, so that the problems that key information is adopted for characteristic matching, the coverage area is narrow and tiny lesions cannot be found in the prior art are solved, the diagnosis efficiency of doctors is improved, and the treatment work of patients is facilitated.
Fig. 2 shows an implementation process of obtaining feature information of an image to be registered, where the specific steps of obtaining the feature information of the image to be registered are as follows:
step S101, detecting an image acquisition instruction, and acquiring a liver picture;
step S102, uploading the liver photo, converting the liver photo into an electronic image according to an initial image conversion model, and extracting the feature information of the image to be registered;
and step S103, uploading the characteristic information of the image to be registered.
In an embodiment of the invention, when the feature information of the image to be registered is uploaded, the feature information of the image needs to be encrypted, wherein the encryption mode is a dynamic password or a fixed password mode, and in addition, a patient can look up a self-detection image and detect pathology at a client at any time in a face recognition or fingerprint recognition mode, so that the patient can see a doctor conveniently.
Fig. 3 shows an implementation flow of an initial image conversion model, and the implementation method of the initial image conversion model includes the following specific steps:
step S1021, acquiring a large number of liver photo samples;
step S1022, training is performed by using the liver image as input and the feature information of the liver image as output, so as to obtain an initial image conversion model.
In one embodiment of the invention, the channel for obtaining the liver photo sample is through a hospital database and a patient diagnosis record, the number of the samples is not less than 1000, and the ratio of the test group to the training group in the samples is 2:3, so that the accuracy of the obtained initial image conversion model is ensured.
Fig. 4 shows an implementation process of extracting feature information of an image to be registered, determining a set of significant feature information points by comparing the feature information of the image to be registered with a preset standard reference image at a plurality of feature information points, wherein the specific steps of extracting the feature information of the image to be registered, determining the set of significant feature information points by comparing the feature information of the image to be registered with the preset standard reference image at the plurality of feature information points are as follows:
step S201, extracting characteristic information of an image to be registered, reconstructing the image to be registered, and storing the image to be registered;
step S202, extracting a standard reference image and an image to be registered, carrying out denoising and de-graying on the image to be registered through a Gaussian filter algorithm to obtain a simulated image to be registered, carrying out three-dimensional image standardization on the simulated image to be registered, and comparing the simulated image to be registered with the standard reference image;
step S203, determining whether the simulated image to be registered meets the standard deviation of the standardized data according to the standard reference image, outputting the salient feature information points if the simulated image to be registered meets the standard deviation of the standardized data, and storing the salient feature information point set.
Fig. 5 shows an implementation process of comparing standard reference images, which specifically includes the following steps:
step S2021, determining a three-dimensional image salient registration point according to the determined simulated image to be registered and the standard reference image, and comparing the three-dimensional image salient registration point with the standard reference image to obtain a data difference to be registered;
step S2022, acquiring a deviation mean value between the data difference to be registered of the salient registration points of different three-dimensional images and the standard deviation of the standardized data as an interference set;
step S2023, traversing all data in the interference set, meeting the standard deviation of the standardized data when the data in the interference set is smaller than a preset threshold value of a standard reference image, and outputting the salient feature information points.
In one embodiment of the invention, the mean value of the deviation between the data difference to be registered of the salient registration points of different three-dimensional images and the standard deviation of the standardized data is obtained as A0Obtaining the deviation value between the data difference to be registered of the salient registration points of different three-dimensional images and the standard deviation of the standardized data as A1，A2，A3，，，，An(ii) a Then A is mixed1，A2，A3，，，，AnThe maximum and minimum numerical values in the three-dimensional image are filtered, the average value of n-2 data samples is calculated, and a deviation interference set between the data difference to be registered of the salient registration points of different three-dimensional images and the standard deviation of the standardized data is obtained.
Fig. 6 shows the specific steps of extracting the salient feature information point set, inputting the salient feature information points into an image accurate registration model to obtain an accurate registration result, feeding back the registration result, extracting the salient feature information point set, inputting the salient feature information points into the image accurate registration model to obtain an accurate registration result, and feeding back the registration result, as follows:
step S301, extracting a salient feature information point set;
step S302, establishing a salient feature information point input image accurate registration model;
step S303, inputting the salient feature information points into an image accurate registration model, calculating registration loss according to the image accurate registration model, judging whether the registration loss is less than a standard loss threshold value, and if so, performing accurate registration.
Fig. 7 shows an implementation process of establishing the image accurate registration model, where the specific steps of establishing the image accurate registration model are as follows:
step S3031, acquiring a plurality of standard training images;
step S3032, inputting each group of standard training images into an initial image registration network, respectively carrying out one-to-one registration to obtain a plurality of image deformation fields, and selecting five groups of image deformation fields with large difference as candidate images;
step S3033, calculating the image deviation degree of the candidate image and the reference image in the image registration network, judging whether the image deviation degree meets the preset condition, and if the image deviation degree meets the preset condition, determining the selected image as a plurality of images to be registered for accurate registration.
In an embodiment of the present invention, the image deviation degree between the candidate image and the reference image in the image registration network is calculated, whether the image deviation degree meets a preset condition is determined, if the image deviation degree does not meet the preset condition, the step S3031 is executed again, and if the image deviation degree does not meet the preset condition, the selected image is determined as a plurality of images to be registered for accurate registration, the operation is terminated, and the operation record is stored.
In an embodiment of the present invention, the method for implementing the initial image conversion model further includes:
and obtaining an initial image conversion model, wherein the initial image conversion method takes photos presented in different directions and angles as variables, screens and calculates the photos through an RANSAC algorithm to obtain a feature point set with image feature points, and extracts an image block set with overlapped feature information through the feature point set to form the initial image conversion model.
Meanwhile, the method for establishing the accurate image registration model further comprises the following steps:
if the standard training images do not meet the preset conditions, acquiring a plurality of standard training images, inputting each group of standard training images into an initial image registration network, respectively performing one-to-one registration to obtain a plurality of image deformation fields, and selecting five groups of image deformation fields with large difference as candidate images;
and calculating the constraint deviation of each group of deformation fields, comparing the constraint deviation to be registered with the standard constraint deviation preset by the initial image registration network, and if the constraint deviation to be registered meets the preset conditions, determining the selected image as a plurality of images to be registered for accurate registration.
The embodiment of the invention also provides an implementation process of the liver cancer imaging omics image accurate registration system, and the liver cancer imaging omics image accurate registration system comprises the following steps:
the characteristic information acquisition module is used for acquiring the characteristic information of the image to be registered;
the information point comparison module is used for extracting the characteristic information of the image to be registered, comparing the characteristic information of the image to be registered with a preset standard reference image at a plurality of characteristic information points and determining a set of significant characteristic information points;
and the registration result acquisition module is used for extracting the salient feature information point set, inputting the salient feature information points into an image accurate registration model to obtain an accurate registration result and feeding back the registration result.
In conclusion, the invention provides a precise registration method for liver cancer imaging omics images, which is characterized in that the precise registration result is obtained by inputting the significant feature information points into a precise registration model of images, so that the problems that the prior art adopts key information to perform feature matching, the coverage area is narrow, and tiny lesions cannot be found are solved, the diagnosis efficiency of doctors is improved, and the method is beneficial to the treatment work of patients.
It should be noted that, for simplicity of description, the above-mentioned embodiments are described as a series of acts, but those skilled in the art will recognize that the present invention is not limited by the order of acts, as some steps may occur in other orders or concurrently in accordance with the invention. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required by the invention.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus may be implemented in other manners. For example, the above-described embodiments of the apparatus are merely illustrative, and for example, the above-described division of the units is only one type of division of logical functions, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or communication connection may be an indirect coupling or communication connection between devices or units through some interfaces, and may be in a telecommunication or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
The above examples are only used to illustrate the technical solutions of the present invention, and do not limit the scope of the present invention. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from these embodiments without making any inventive step, fall within the scope of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art may still make various combinations, additions, deletions or other modifications of the features of the embodiments of the present invention according to the situation without conflict, so as to obtain different technical solutions without substantially departing from the spirit of the present invention, and these technical solutions also fall within the protection scope of the present invention.

Claims (10)

1. A liver cancer imaging omics image accurate registration method is used for assisting a doctor in distinguishing and analyzing liver cancer imaging omics images and is characterized by comprising the following steps:
acquiring characteristic information of an image to be registered;
extracting the feature information of the image to be registered, comparing the feature information of the image to be registered with a preset standard reference image at a plurality of feature information points, and determining a set of significant feature information points;
and extracting the set of the salient feature information points, inputting the salient feature information points into an accurate image registration model to obtain an accurate registration result, and feeding back the registration result.
2. The method for accurately registering liver cancer omics images as set forth in claim 1, wherein the characteristic information of the images to be registered comprises: the liver appearance, the liver texture, the liver gray scale texture characteristic data and the liver intensity characteristic data presented by the magnetic anatomical imaging, the CT image, the PET image, the X-ray image and the ultrasonic image.
3. The accurate registration method of hepatoma omics images as set forth in claim 2, wherein said specific step of obtaining the characteristic information of the images to be registered comprises:
detecting an image acquisition instruction, and acquiring a liver photo;
uploading the liver photo, converting the liver photo into an electronic image according to an initial image conversion model, and extracting the feature information of the image to be registered;
and uploading the feature information of the image to be registered.
4. The method for accurate registration of hepatoma omics images as defined in claim 3, wherein said method for implementing said initial image transformation model comprises:
acquiring a large number of liver photo samples;
and training by using the picture of the liver as input and the characteristic information of the liver image as output to obtain an initial image conversion model.
5. The accurate registration method of hepatoma omics images as defined in claim 3, wherein the specific steps of extracting the feature information of the image to be registered, comparing the feature information of the image to be registered with a preset standard reference image at a plurality of feature information points, and determining a set of significant feature information points comprise:
extracting characteristic information of the image to be registered, reconstructing the image to be registered, and storing the image to be registered;
extracting a standard reference image and an image to be registered, performing denoising and de-graying on the image to be registered through a Gaussian filtering algorithm to obtain a simulated image to be registered, performing three-dimensional image standardization on the simulated image to be registered, and comparing the simulated image to be registered with the standard reference image;
and determining whether the simulated image to be registered meets the standard deviation of the standardized data according to the standard reference image, outputting the salient feature information points if the simulated image to be registered meets the standard deviation of the standardized data, and storing the salient feature information point set.
6. The method of claim 5, wherein the step of comparing the standard reference images comprises:
determining a three-dimensional image salient registration point according to the determined simulation image to be registered and the standard reference image, and comparing the three-dimensional image salient registration point with the standard reference image to obtain a data difference to be registered;
acquiring a mean deviation value between a data difference to be registered and a standard deviation of standardized data of different three-dimensional image salient registration points as an interference set;
and traversing all data in the interference set, meeting the standard deviation of the standardized data when the data in the interference set is smaller than a preset threshold value of a standard reference image, and outputting the information points of the salient features.
7. The accurate registration method for hepatoma imaging omics images as defined in claim 6, wherein the specific implementation steps of extracting the set of significant characteristic information points, inputting the significant characteristic information points into an accurate registration model for images to obtain an accurate registration result, and feeding back the registration result comprise:
extracting a set of salient feature information points;
establishing a precise registration model of the salient feature information point input image;
and inputting the salient feature information points into the image accurate registration model, calculating registration loss according to the image accurate registration model, judging whether the registration loss is less than a standard loss threshold value, and if so, performing accurate registration.
8. The method for accurately registering liver cancer omics images as set forth in claim 7, wherein the method for establishing the accurate registration model comprises the following steps:
acquiring a plurality of standard training images;
inputting each group of standard training images into an initial image registration network, respectively performing one-to-one registration to obtain a plurality of image deformation fields, and selecting five groups of image deformation fields with large difference as candidate images;
and calculating the image deviation degree of the candidate image and the reference image in the image registration network, judging whether the image deviation degree meets a preset condition, and if the image deviation degree meets the preset condition, determining the selected image as a plurality of images to be registered for accurate registration.
9. The method of claim 4, wherein the initial image transformation model implementation method further comprises:
and obtaining an initial image conversion model, wherein the initial image conversion method takes photos presented in different directions and angles as variables, screens and calculates the photos through an RANSAC algorithm to obtain a feature point set with image feature points, and extracts an image block set with overlapped feature information through the feature point set to form the initial image conversion model.
10. The method for accurate registration of hepatoma omics images as defined in claim 8, wherein said method for establishing an accurate registration of images further comprises:
if the standard training images do not meet the preset conditions, acquiring a plurality of standard training images, inputting each group of standard training images into an initial image registration network, respectively performing one-to-one registration to obtain a plurality of image deformation fields, and selecting five groups of image deformation fields with large difference as candidate images;
and calculating the constraint deviation of each group of deformation fields, comparing the constraint deviation to be registered with the standard constraint deviation preset by the initial image registration network, and if the constraint deviation to be registered meets the preset conditions, determining the selected image as a plurality of images to be registered for accurate registration.
CN202111145960.3A 2021-09-28 2021-09-28 Accurate registration method for liver cancer imaging omics images Withdrawn CN113822917A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111145960.3A CN113822917A (en) 2021-09-28 2021-09-28 Accurate registration method for liver cancer imaging omics images

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111145960.3A CN113822917A (en) 2021-09-28 2021-09-28 Accurate registration method for liver cancer imaging omics images

Publications (1)

Publication Number Publication Date
CN113822917A true CN113822917A (en) 2021-12-21

Family

ID=78915761

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111145960.3A Withdrawn CN113822917A (en) 2021-09-28 2021-09-28 Accurate registration method for liver cancer imaging omics images

Country Status (1)

Country Link
CN (1) CN113822917A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116309751A (en) * 2023-03-15 2023-06-23 北京医准智能科技有限公司 Image processing method, device, electronic equipment and medium

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116309751A (en) * 2023-03-15 2023-06-23 北京医准智能科技有限公司 Image processing method, device, electronic equipment and medium

Similar Documents

Publication Publication Date Title
CN110996789B (en) Systems and methods for performing screening, diagnostic, or other image-based analysis tasks
AU2008211786B2 (en) Tools for aiding in the diagnosis of neurodegenerative diseases
Mazziotta et al. A four-dimensional probabilistic atlas of the human brain
Deng et al. Deep learning-based HCNN and CRF-RRNN model for brain tumor segmentation
CN110379492A (en) A kind of completely new AI+PACS system and its audit report construction method
US8335359B2 (en) Systems, apparatus and processes for automated medical image segmentation
KR101919866B1 (en) Method for aiding determination of presence of bone metastasis from bone scan image and apparatus using the same
Liu Symmetry and asymmetry analysis and its implications to computer-aided diagnosis: A review of the literature
CN110598722B (en) Multi-modal neuroimaging data automatic information fusion system
KR20130136519A (en) Diagnosis assitance system utilizing panoramic radiographs, and diagnosis assistance program utilizing panoramic radiographs
CN103249358A (en) Medical image processing device
CN104093354A (en) Method and apparatus for assessment of medical images
CN112767374A (en) Alzheimer disease focus region semantic segmentation algorithm based on MRI
CN111227834B (en) Automatic rapid visualization method for resting brain function connection
CN113822917A (en) Accurate registration method for liver cancer imaging omics images
CN110782983A (en) Remote dynamic pathological diagnosis method
CN112052882B (en) Classification model construction, classification and visualization method for magnetic resonance brain structure image
CN114334097A (en) Automatic assessment method based on lesion progress on medical image and related product
CN114300071A (en) Real data-based informatization system and method
Abd Hamid et al. Incorporating attention mechanism in enhancing classification of alzheimer’s disease
CN113822872A (en) Image feature information extraction method for hepatoma imaging omics
Veltri et al. Clinical data annotation for parotid neoplasia management
CN114399462B (en) Medical image anomaly detection method and device based on depth characterization network
CN117351489B (en) Head and neck tumor target area delineating system for whole-body PET/CT scanning
CN112766333B (en) Medical image processing model training method, medical image processing method and device

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
WW01 Invention patent application withdrawn after publication

Application publication date: 20211221

WW01 Invention patent application withdrawn after publication