CN115690207A - Automatic positioning method and device based on head clinical image - Google Patents
Automatic positioning method and device based on head clinical image Download PDFInfo
- Publication number
- CN115690207A CN115690207A CN202211274873.2A CN202211274873A CN115690207A CN 115690207 A CN115690207 A CN 115690207A CN 202211274873 A CN202211274873 A CN 202211274873A CN 115690207 A CN115690207 A CN 115690207A
- Authority
- CN
- China
- Prior art keywords
- head
- cvh
- patient
- registration
- mapping relation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to an automatic positioning method and device based on head clinical images, which carries out coordinate transformation on a head anatomical structure of a CVH (composite video H) according to a space direction mapping relation and a space size mapping relation to obtain a head anatomical structure of the CVH; and the CVH registration head anatomical structure and the position identification and the organ name of each organ area in the CVH registration head anatomical structure form a new CVH anatomical knowledge base. Performing multi-plane reconstruction on the clinical image of the head of the patient to obtain the head anatomical structure of the patient; the head anatomy structure of the patient and the CVH registration head anatomy structure have a real-time mapping relation, so that the position identification and the organ name of each organ area in the head anatomy structure of the patient and the CVH registration head anatomy structure have a one-to-one correspondence relation. And an interested region is selected from the anatomical knowledge map, and the corresponding position of the interested region in the head anatomical structure of the patient is automatically positioned, so that the identification difficulty of the clinical image of the head of the patient is reduced.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to an automatic positioning method and device based on head clinical images.
Background
Clinical medical tomographic images such as CT and MRI have an important role in the diagnosis and treatment of diseases. However, clinical medical images often require expertise and experience of high-tech physicians, especially imaging physicians, to correctly read due to the abstraction of the imaging modality and the complexity of the human anatomy. Especially, the clinical medical tomograph of the head with a complex structure is too abstract for low-grade medical students, interdisciplinary scientific researchers, medical engineering personnel and common patients to accurately identify.
For interdisciplinary professionals, young physicians and medical students, the professionals in the non-skilled medical field have difficulty in effectively positioning and identifying some complex lesion areas and fine anatomical structures in medical images. For patients and family members, in the process of disease diagnosis and treatment, the active participation awareness of the patients and family members is continuously enhanced, medical units also hand image data to individuals by means of carving discs and the like, but the existing browsers still need the guidance of professional doctors to help the patients to establish the intuitive understanding of focuses, and the patients and the family members cannot automatically understand all organ regions in the images.
Therefore, in the intelligent information age with universality emphasis, how to rapidly position the interested region in the clinical head clinical image and present the interested region to non-qualified depth imaging professionals so as to reduce the identification difficulty of the clinical head clinical image has unique value.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides an automatic positioning method and device based on a head clinical image, so as to reduce the identification difficulty of the head clinical image of a patient and improve the universality.
In a first aspect, the invention provides an automatic positioning method based on head clinical images.
In a first implementation manner, an automatic positioning method based on head clinical images includes:
acquiring a clinical image of the head of a patient;
registering the clinical image of the head of the patient and the CVH to obtain a space direction mapping relation and a space size mapping relation;
performing coordinate transformation on the head anatomical structure of the CVH according to the space direction mapping relation and the space size mapping relation to obtain a CVH registration head anatomical structure; the CVH registration head anatomical structure and the position identification and the organ name of each organ area in the CVH registration head anatomical structure form a new CVH anatomical knowledge base;
performing multi-plane reconstruction on the clinical image of the head of the patient to obtain the head anatomical structure of the patient; a real-time mapping relation exists between the head anatomy structure of the patient and the CVH registration head anatomy structure;
and selecting a region of interest from the anatomical knowledge map, and automatically positioning the corresponding position of the region of interest in the anatomical structure of the head of the patient.
In combination with the first implementable manner, in a second implementable manner, the clinical image of the head of the patient and the CVH are registered, and a spatial direction mapping relation and a spatial dimension mapping relation are obtained, wherein the method comprises the following steps:
carrying out first registration on the clinical image of the head of the patient and the CVH to obtain a space direction mapping relation and a first registration result;
and performing secondary registration on the clinical image of the head of the patient and the CVH according to the first registration result to obtain a spatial dimension mapping relation.
With reference to the second implementable manner, in a third implementable manner, performing first registration on the clinical image of the head of the patient and the CVH to obtain a spatial direction mapping relationship and a first registration result, including:
carrying out three-dimensional contour reconstruction on the clinical image of the head of the patient to obtain a three-dimensional contour structure of the patient;
acquiring a first characteristic point according to the three-dimensional contour structure of the patient;
acquiring a second characteristic point of the CVH three-dimensional contour structure;
registering the first characteristic point and the second characteristic point to obtain a space direction mapping relation and a first registration result; the first registration result is a primary registration structure of the three-dimensional contour of the patient, and the primary registration structure of the three-dimensional contour of the patient is consistent with the spatial direction of the three-dimensional contour structure of the CVH.
With reference to the third implementable manner, in a fourth implementable manner, the acquiring a first feature point according to a three-dimensional contour structure of a patient includes:
and respectively selecting the eye-nose triangular areas from the three-dimensional contour structure of the patient and generating first characteristic points of the eye-nose triangular areas.
With reference to the third implementable manner, in a fifth implementable manner, the second registration is performed on the clinical image of the head of the patient and the CVH according to the first registration result, and a spatial dimension mapping relationship is obtained, including:
acquiring a first annular contour according to the three-dimensional contour structure of the patient;
acquiring a second annular contour of the CVH three-dimensional contour structure;
in the primary registration structure of the three-dimensional contour of the patient, registration is carried out according to the first annular contour and the second annular contour, and a space size mapping relation is obtained.
With reference to the third implementable manner, in a sixth implementable manner, the coordinate transformation is performed on the head anatomy structure of the CVH according to the spatial direction mapping relationship and the spatial dimension mapping relationship, so as to obtain a CVH registration head anatomy structure, including:
acquiring a homogeneous transformation matrix according to the space direction mapping relation and the space size mapping relation; the homogeneous transformation matrix is used for representing the coordinate transformation relation between the three-dimensional contour structure of the patient and the head anatomical structure of the CVH;
and performing coordinate transformation on the head anatomical structure of the CVH according to the homogeneous transformation matrix to obtain the head anatomical structure of the CVH.
With reference to the sixth implementable manner, in a seventh implementable manner, the homogeneous transformation matrix is obtained by the following formula:
With reference to the first implementable manner, in an eighth implementable manner, the method includes:
and after the corresponding position of the region of interest in the head anatomical structure of the patient is automatically positioned, marking colors on the corresponding position region, and displaying the plane of the corresponding position region.
In a first aspect, the invention provides an automatic positioning device for head clinical images.
In a ninth implementation manner, an automatic positioning device for head clinical images includes:
an acquisition module configured to acquire clinical images of a patient's head;
the registration module is configured to register the clinical image of the head of the patient and the CVH to obtain a spatial direction mapping relation and a spatial dimension mapping relation;
the coordinate transformation module is configured to perform coordinate transformation on the head anatomical structure of the CVH according to the space direction mapping relation and the space size mapping relation to obtain a CVH registration head anatomical structure; the CVH registration head anatomical structure and the position identification and the organ name of each organ area in the CVH registration head anatomical structure form a new CVH anatomical knowledge base;
a multi-plane reconstruction module configured to perform multi-plane reconstruction on clinical images of the head of the patient to obtain an anatomical structure of the head of the patient; a real-time mapping relation exists between the head anatomy structure of the patient and the CVH registration head anatomy structure;
an automatic localization module configured to select a region of interest from the anatomical knowledge map and automatically localize a corresponding position of the region of interest in the patient's head anatomy.
According to the technical scheme, the beneficial technical effects of the invention are as follows:
performing coordinate transformation on the head anatomy structure of the CVH according to the space direction mapping relation and the space size mapping relation to obtain a CVH registration head anatomy structure; the CVH registration head anatomical structure, position marks of all organ areas in the CVH registration head anatomical structure and organ names form a new CVH anatomical knowledge map, and then multi-plane reconstruction is carried out on clinical images of the head of a patient to obtain the head anatomical structure of the patient; the head anatomy structure of the patient and the CVH registration head anatomy structure have a real-time mapping relation, so that the position identification and the organ name of each organ area in the head anatomy structure of the patient and the CVH registration head anatomy structure have a one-to-one correspondence relation. And selecting a region of interest from the anatomical knowledge map, and automatically positioning the corresponding position of the region of interest in the head anatomy of the patient. Therefore, the professional doctors in the non-qualification field can also identify each organ area of the clinical image of the head of the patient, the identification difficulty of the clinical image of the head of the patient is reduced, cross-subject professionals, young doctors, medical students, patients and family members can understand the focus condition of the clinical image of the head of the patient, and the universality is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. Throughout the drawings, like elements or portions are generally identified by like reference numerals. In the drawings, elements or portions are not necessarily drawn to scale.
Fig. 1 is a schematic diagram of an automatic positioning method based on head clinical images according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an automatic positioning method for head clinical images according to an embodiment of the present invention.
Detailed Description
Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and therefore are only examples, and the protection scope of the present invention is not limited thereby.
It is to be noted that, unless otherwise specified, technical or scientific terms used herein shall have the ordinary meaning as understood by those skilled in the art to which the invention pertains.
Referring to fig. 1, the present embodiment provides an automatic positioning method based on head clinical images, including:
s01, acquiring a clinical image of the head of a patient;
s02, registering the clinical image of the head of the patient and the CVH to obtain a spatial direction mapping relation and a spatial dimension mapping relation;
s03, performing coordinate transformation on the head anatomical structure of the CVH according to the space direction mapping relation and the space size mapping relation to obtain a CVH registration head anatomical structure; the CVH registration head anatomical structure and the position identification and the organ name of each organ area in the CVH registration head anatomical structure form a new CVH anatomical knowledge base;
s04, performing multi-plane reconstruction on the clinical image of the head of the patient to obtain the head anatomical structure of the patient; the CVH registration head anatomical structure, and position marks and organ names of all organ areas in the CVH registration head anatomical structure form a new CVH anatomical knowledge map;
and S05, selecting an interested region from the anatomical knowledge map, and automatically positioning the corresponding position of the interested region in the head anatomical structure of the patient.
Performing coordinate transformation on the head anatomy structure of the CVH according to the space direction mapping relation and the space size mapping relation to obtain a CVH registration head anatomy structure; the CVH registration head anatomical structure, position marks of all organ areas in the CVH registration head anatomical structure and organ names form a new CVH anatomical knowledge map, and then multi-plane reconstruction is carried out on clinical images of the head of a patient to obtain the head anatomical structure of the patient; the head anatomy structure of the patient and the CVH registration head anatomy structure have a real-time mapping relation, so that the position identification and the organ name of each organ area in the head anatomy structure of the patient and the CVH registration head anatomy structure have a one-to-one correspondence relation. And selecting an interested region from the anatomical knowledge map, and automatically positioning the corresponding position of the interested region in the head anatomical structure of the patient according to the real-time mapping relation between the head anatomical structure of the patient and the CVH registration head anatomical structure. Therefore, the professional doctors in the non-qualification field can also identify each organ area of the clinical image of the head of the patient, the identification difficulty of the clinical image of the head of the patient is reduced, cross-subject professionals, young doctors, medical students, patients and family members can understand the focus condition of the clinical image of the head of the patient, and the universality is improved.
Alternatively, clinical images of the patient's head are obtained by CT (Computed Tomography) or MRI (Nuclear Magnetic Resonance Imaging).
In some embodiments, a CVH (Chinese visual Human) includes a Chinese visual Human anatomy cross-sectional dataset and corresponding CT, MRI datasets, which include clinical images and anatomical structures of various parts of the body. The CVH includes an anatomical knowledge map that includes the anatomical structure of each site, as well as knowledge relating to the location of each organ region in the structure, the name of the organ, and the like. The three-dimensional contour structure of the head in the CVH is obtained by three-dimensional contour reconstruction of a clinical image of the head of the CVH.
Optionally, the CVH registered head anatomy replaces the CVH head anatomy in the old CVH anatomical knowledge base, forming a new CVH anatomical knowledge base. The patient's head anatomy is mapped one-to-one with the CVH registration head anatomy, where the CVH registration head anatomy has the location identification and organ name of each organ region. Therefore, after each region in the head anatomy structure of the patient is corresponding to each region in the CVH registration head anatomy structure, the position identification and the organ name of each organ region in the CVH registration head anatomy structure also correspond to each region in the head anatomy structure of the patient, and automatic positioning is realized according to the corresponding relation between the position identification and the organ name of each organ region in the CVH registration head anatomy structure and each region in the head anatomy structure of the patient. For example, a mapping relationship exists between a and B, if a is selected, B corresponds to a, and the position identifier and the organ name related to B also correspond to a.
Optionally, the registration of the clinical image of the head of the patient and the CVH to obtain a spatial direction mapping relationship and a spatial dimension mapping relationship includes: carrying out first registration on the clinical image of the head of the patient and the CVH to obtain a space direction mapping relation and a first registration result; and performing secondary registration on the clinical image of the head of the patient and the CVH according to the first registration result to obtain a space size mapping relation.
Optionally, performing first registration on the clinical image of the head of the patient and the CVH by using an Iterative Closest Point (ICP) algorithm to obtain a transformation matrix of a spatial direction mapping relationship and a first registration result; and performing secondary registration on the clinical image of the head of the patient and the CVH by adopting an ICP (inductively coupled plasma) algorithm according to the first registration result to obtain a transformation matrix of the space size mapping relation. In some embodiments, in the common heterogeneous registration, due to a certain morphological difference between different human heads, the protruding parts of the nose, ears and the like and the similar gradient parts of the forehead, the occiput and the like are easy to cause wrong "best" matching in the registration process, i.e. the problem of local extremum, so that the deviation is still large, and the registration result is not ideal. The method and the device have the advantages that the result of the first registration is used as the basis of the second registration to carry out the second progressive registration, so that the wrong 'best' matching condition caused by stopping the iteration after the ICP algorithm falls into a local extreme value in the iteration process is avoided, the deviation is effectively reduced, the registration degree between the clinical image of the head of the patient and the CVH is improved, and the accuracy of automatic positioning is further improved.
Optionally, the first registration of the clinical image of the head of the patient and the CVH is performed to obtain a spatial direction mapping relationship and a first registration result, including: carrying out three-dimensional contour reconstruction on the clinical image of the head of the patient to obtain a three-dimensional contour structure of the patient; acquiring a first characteristic point according to the three-dimensional contour structure of the patient; acquiring a second characteristic point of the CVH three-dimensional contour structure; registering the first characteristic point and the second characteristic point to obtain a space direction mapping relation and a first registration result; the first registration result is CVH primary registration three-dimensional contour structure, and the spatial direction of the patient three-dimensional contour structure is consistent with that of the CVH primary registration three-dimensional contour structure.
Optionally, the first feature Point and the second feature Point are registered by an Iterative Closest Point algorithm (ICP). In some embodiments, the iterative closest point algorithm is a method for registration of surfaces by iteratively calculating the sum of squared residuals of corresponding points between the surfaces based on a quaternion method. For example, there are two different sets of coordinate points in the world coordinate system P = { Pi, i =0,1,2, …, k } (red patches) and U = { Ui, i =0,1,2, …, n } (blue patches). And setting a point set P as a target point set and a point set U as a source point set, and assuming that the P and the U can roughly correspond in space (generally setting k to be more than or equal to n), obtaining a new point set U ' by continuously rotating and translating the point set U, so that the distance between the homologous points of the point set U ' and the point set P is minimum (enabling the U ' and the P to be overlapped as much as possible). U 'can be obtained by the rigid body geometry transformation formula U' = RU + T. Wherein, R represents a three-dimensional rotation matrix of the transformation point set U, and T represents a translation vector of the transformation point set U. The core of the process is to adopt a minimum root mean square method, iteratively calculate the sum of the square residuals of corresponding points between the point sets U 'and P by continuously correcting R and T, find the minimum root mean square error between U' and Q, and if the error is smaller than a preset limit value, end iteration to obtain the optimal registration solution.
Optionally, the obtaining the first feature point according to the three-dimensional contour structure of the patient includes: and respectively selecting the eye-nose triangular areas from the three-dimensional contour structure of the patient and generating first characteristic points of the eye-nose triangular areas. In some embodiments, the eye-nose trigone areas are obvious and have large differences and are used as facial feature points for registration, so that the problem of local extreme value of common heterogeneous registration is avoided, and the registration is more accurate.
Optionally, performing a second registration on the clinical image of the head of the patient and the CVH according to the first registration result to obtain a spatial dimension mapping relationship, including: acquiring a first annular contour according to the three-dimensional contour structure of the patient; acquiring a second annular contour of the CVH three-dimensional contour structure; on the basis of the primary registration of the three-dimensional contour structure of the patient and the CVH, registration is carried out according to the first annular contour and the second annular contour, and a space size mapping relation is obtained.
In some embodiments, obtaining the annular contour comprises reading original data by using a C + + open-source FO-DICOM library, converting the original data into an 8-bit bitmap format, and then performing binarization processing by using an Ostu adaptive threshold segmentation method; filling small intracranial holes by using an opening operation, and filling an area in the whole cranium by using a FloodFill algorithm; and finally, performing forward and backward subtraction operation by using the corrosion template 3*3 to extract the outline of one frame of data, and continuously processing the next frame of data until batch processing is completed, thereby finally realizing extraction of the annular outline.
In some embodiments, a space direction mapping relation is determined by first registration, the space direction between the three-dimensional outline structure of the patient and the three-dimensional outline structure of the CVH is transformed to be consistent through the first registration, after the space direction between the three-dimensional outline structure of the patient and the three-dimensional outline structure of the CVH is consistent, annular feature outlines with the same data are respectively selected from the three-dimensional outline structure of the patient and the three-dimensional outline structure of the CVH through manual operation, and the registration of space sizes is further performed according to the annular feature outlines to adapt to the size difference of different human heads, so that the three-dimensional outline structure of the CVH is close to the three-dimensional outline structure of the patient after two times of registration, and the real-time mapping between the three-dimensional outline structures of the patient and the CVH is favorably realized. Meanwhile, the annular feature contour with the same data is manually selected from the three-dimensional contour structure of the patient and the three-dimensional contour structure of the CVH, so that the risk of local extreme values is reduced.
Optionally, performing coordinate transformation on the head anatomy structure of the CVH according to the spatial direction mapping relationship and the spatial dimension mapping relationship to obtain a CVH registration head anatomy structure, including: acquiring a homogeneous transformation matrix according to the space direction mapping relation and the space size mapping relation; the homogeneous transformation matrix is used for representing the coordinate transformation relation between the three-dimensional contour structure of the patient and the head anatomical structure of the CVH; and performing coordinate transformation on the head anatomy structure of the CVH according to the homogeneous transformation matrix to obtain the head anatomy structure of the CVH.
Optionally, the head anatomy structure of the CVH is transformed through a spatial direction mapping relationship and a spatial dimension mapping relationship, so that the head anatomy structure of the CVH is close to the three-dimensional contour structure of the head of the patient, and is consistent with the spatial direction and the spatial dimension of the CVH. The data between the CVH registration head anatomy obtained by the transformation and the three-dimensional contour structure of the patient head has a one-to-one mapping relation. And then performing multi-plane reconstruction on the three-dimensional contour structure of the head of the patient, and performing scene superposition to obtain the final multi-plane anatomical structure of the head of the patient. The head anatomy of the patient corresponds to the head anatomy of the CVH registration one-to-one; and the head anatomy structure of the CVH and the anatomy knowledge map have a one-to-one correspondence relationship, and finally the head anatomy structure of the patient, the head anatomy structure of the CVH, and the anatomy knowledge map have a mapping relationship.
In some embodiments, a clinical image of the head of a patient (e.g., patient image data such as CT/MRI) is used as a target image, and a CVH data set is used as a floating image to perform registration movement on the target image, so as to obtain a homogeneous transformation matrix; the CVH general anatomical structure segmentation image layer is synchronously transformed along with the movement of the floating image, is registered with the head clinical image of the patient, and is subjected to three-dimensional reconstruction to obtain real-time mapping under any visual angle, so that the aim of focus/organ navigation is fulfilled, and the registration result allows quick evaluation and correction.
Optionally, the homogeneous transformation matrix is obtained by the following formula:
Optionally, during the multi-plane reconstruction of the clinical image of the head of the patient, the multi-plane reconstruction is performed by referring to a multi-plane reconstruction method in the prior art, so as to obtain a patient anatomical structure of multiple planes.
Optionally, the automatic positioning method based on the clinical image of the head includes: and after the corresponding position of the region of interest in the head anatomical structure of the patient is automatically positioned, marking colors on the corresponding position region, and displaying the plane of the corresponding position region.
Optionally, the present embodiment provides an automatic positioning method based on clinical head images, including: acquiring a clinical image of the head of a patient; carrying out secondary descending registration on the clinical image of the head of the patient and CVH (Chinese visual Human body) to obtain a CVH registration head anatomical structure; performing multi-plane reconstruction on the clinical image of the head of the patient to obtain the head anatomical structure of the patient; real-time mapping between the patient's head anatomy and an anatomical knowledge-map in the CVH; and selecting a region of interest from the anatomical knowledge map, and automatically positioning the corresponding position of the region of interest in the head anatomy of the patient.
As shown in fig. 2, the present invention provides an automatic positioning device for head clinical image, comprising: the system comprises an acquisition module 101, a registration module 102, a coordinate transformation module 103, a multi-plane reconstruction module 104 and an automatic positioning module 105. The acquisition module 101 is configured to acquire clinical images of the head of a patient; the registration module 102 is configured to register the clinical image of the head of the patient and the CVH, and obtain a spatial direction mapping relation and a spatial dimension mapping relation; the coordinate transformation module 103 is configured to perform coordinate transformation on the head anatomy structure of the CVH according to the spatial direction mapping relationship and the spatial size mapping relationship, so as to obtain a CVH registration head anatomy structure; the CVH registration head anatomical structure and the position identification and the organ name of each organ area in the CVH registration head anatomical structure form a new CVH anatomical knowledge base; the multi-plane reconstruction module 104 is configured to perform multi-plane reconstruction on clinical images of the head of the patient to obtain an anatomical structure of the head of the patient; a real-time mapping relation exists between the head anatomy structure of the patient and the CVH registration head anatomy structure; the automatic localization module 105 is configured to select a region of interest from the anatomical knowledge map and automatically localize a corresponding location of the region of interest in the patient's head anatomy.
In some embodiments, the method for automatically positioning head clinical images comprises: and S21, selecting clinical image data by a user, and displaying a three-dimensional contour structure model of the CVH standard atlas and an automatic three-dimensional contour structure model of the clinical image data selected by the user. And S22, respectively adjusting the two contour models to visual angles (without setting the same size and direction) convenient for selecting the eye-nose triangular area, selecting the eye-nose triangular area in a drawing rectangular mode, and respectively generating blue and red characteristic points on the CVH (left) and CT/MRI models according to the drawing areas. S23, carrying out primary registration to obtain a matrix equation of a space direction mapping relation, and displaying a coarse registration result; wherein the blue CVH feature points are close to the red CT/MRI feature points. And S24, automatically extracting the annular contour, carrying out secondary registration, obtaining a final homogeneous matrix equation, guiding the CVH anatomical knowledge data to approach to the CT/MRI data, visually presenting a registration result in a same-scene superposition mode, and displaying the superposition effect of the two groups of images.
Performing multi-planar Reconstruction (MPR) of CT/MRI, displaying the result of multi-planar Reconstruction, listing a list of CVH gross organs for screening and selection. For example, the "brainstem" is selected, namely, the brainstem region is marked with yellow correspondingly in the MPR reconstruction region of the CT/MRI, so that the user can conveniently read the brainstem region.
In some embodiments, the clinical image set (CT and MRI) head region is collected and analyzed for rigid structure corresponding features of the CVH gross anatomy, completing three-dimensional reconstruction of the outer contour rigid features and feature point picking. Secondly, determining a space direction by adopting a secondary progressive registration method, namely, rigidly transforming facial features of an eye-nose triangular area to finish coarse registration for the first time; further, the annular contour features are adopted for similarity registration to determine the space size, the problem of local extreme values and deformation easily generated by heterogeneous registration is solved, and a standardized operation process including a secondary registration mode is summarized. And finally, resampling the head in-production image, performing spatial transformation to obtain the head anatomical structure of the patient, performing coordinate field transformation on a volume data field formed by a gross anatomical structure layer set of the CVH according to a registration result to realize the purpose of anatomical structure mapping to clinical image data, and designing and optimizing an MPR multi-plane reconstruction interactive UI (user interface) to achieve a productive application form.
In some embodiments, the registration of the existing anatomical structure knowledge-map with the medical image requires certain requirements on the integrity and quality of the image structure, and is different from the clinical image scanning specifications. In order to prevent excessive medical treatment and control medical treatment cost, clinical real data are often local scanning rather than thin-layer high-field scanning required by scientific research, the requirements on 'complete structure' and quality required by the current mainstream registration algorithm are difficult to achieve, and the solution mode is usually at the cost of huge calculation consumption. The automatic positioning method based on the head clinical image has the advantages of being small in calculation amount, low in requirement for image code scanning and light in weight.
In some embodiments, the automatic positioning method based on the head clinical image provided by the application can map the anatomical knowledge mapping in the CVH to the private CT/MRI data set in real time without professional explanation and special 3D organ printing and labeling, visually present the current organ region and position the region of interest in an intelligent navigation mode, achieve the purpose of unmanned intelligent assistance of explanation, have the characteristic of man-machine interaction, serve active participation requirements of patients and family members in disease diagnosis and working requirements of low-grade medical students, cross-discipline researchers and medical engineering personnel, and support the large direction of taking the patients as the center.
In some embodiments, the open source assembly of the present application comprises: a Visual Studio 2022 community version development platform, a core development platform; ) VTK (Visualization Toolkit) for visual reconstruction of medical tomograms; openCV, for morphological operations of medical images; FO-DICOM is used for reading and writing medical DICOM images.
Optionally, an automatic positioning method based on clinical head images further includes: and acquiring test data, and performing time-consuming test and functional test on the automatic positioning method based on the head clinical image according to the test data to obtain a test result.
In some embodiments, table 3 is an example table of test data. In table 3, the data one, two, and three have different positioning and different head scanning intervals and come from three different patients with different face types, so that the application range of the method can be evaluated to a certain extent, each group of data is repeated twice, and objective repeatability evaluation can be obtained. Different from the current mainstream registration algorithm, the complete structure of the image data is required, the diversification characteristic of the clinical real image scanning part is reflected, and 3 sets of representative tomographic image data sets are screened in the test. Wherein, data one is CT image, data two to three are MRI images with poor spatial resolution, the spatial resolution is 512 × 121, 288 × 384 × 18, 160 × 126 × 160 respectively, and the pixel pitch is 0.43 × 0.70, 0.625 × 0.625, 3.15, 1.625 × 1.625 respectively. Considering the situation that the focus region is emphasized by clinical image scanning to perform local imaging due to prevention of over-treatment and medical cost control, the scanning interval of the first data only covers the forehead to the nasal ala, and the ranges of the second data and the third data are further compressed to the eyebrow arch to the nasal ala, so that the robustness of the project scheme in registration can be better tested.
Table 1 example table of test data
In some embodiments, the method for automatic head-based clinical image positioning is time-consuming based on test data, and comprises: and performing three-dimensional reconstruction on the CT/MRI and the CVH data set after registration transformation in the same scene, and evaluating the registration accuracy according to the fitting degree of the CT/MRI and the CVH data set. Because CVH is a 24-bit true color bitmap data set and is completely different from CT and MRI imaging characteristics, the coloring range and the transparency scheme of two sets of volume rendering can be obviously separated. In the testing process, part of typical anatomical structures are selected to be mapped on CT/MRI images, and registration accuracy evaluation is carried out according to the proportion of covered parts. The typical anatomical structure is selected from brain stem and optic nerve, the position relation of the brain stem and the optic nerve in the head is relatively representative, and the shape and the size are moderate.
In some embodiments, table 2 is the test results. As shown in table 2, five times of repetitive tests are performed in the configuration of i5-4210,2.60GH in 2,8GB and Windows 7 ordinary personal computer, all the tests are normally performed, the time consumption of the method is not more than 40 seconds, and the speed is high.
TABLE 2
In some embodiments, the automatic positioning method based on the head clinical image is functionally tested according to the test data, and images of the data I, the data II and the data III are respectively registered by the method of the application, so as to obtain three registration results. The spatial differences between the unregistered clinical image and the original spatial pose of the CVH head can be clearly seen in the same scene overlay. And (4) after the registration, the fixed observation screenshots of the two visual angles are observed, and the registration result is subjectively considered to be very ideal through the observation of different visual angles. The method is also a reference basis for judging whether a common operator performs more than three times of correction registration. On the basis, the invited five-year seniority imaging physician evaluates the observation result by a five-score table, and respectively carries out three-time registration on three groups of data to obtain scores of 9-time registration effect, wherein the score distribution condition is shown in the following table 3:
TABLE 3
As shown in table 3, of the 9 registrations, 7 registrations are successful once according to the procedure, the effect is satisfactory, and 2 registrations are satisfactory if similar registration correction needs to be performed again, and the purpose of registration and the following situations are not basically achieved. And from the perspective of professional imaging physicians, the registration result is generally considered to be suitable for the popular universality requirement provided by the project.
In some embodiments, the head clinical image-based automatic positioning method is functionally tested according to test data, and comprises the following steps: and selecting brainstem and optic nerve for calibration measurement, and performing function test according to the measurement result. And respectively carrying out secondary registration on the first data, the second data and the third data, and randomly extracting the fault picture at the same position of one of the primary registration for calibration indication. In some embodiments, the same-position tomograms which are registered once are arbitrarily extracted for calibration, and a brain stem map and an optic nerve map of a complete anatomical structure of a plurality of frames and a brain stem map and an optic nerve map of an anatomical structure cut along a sampling plane are obtained. On the basis, the invited five-year seniority imaging physician evaluates the observation result by a five-score table, and respectively carries out three-time registration on three groups of data to obtain scores of 9-time registration effect, wherein the score distribution condition is shown in the following table 4:
TABLE 4
The mapping accuracy of the brainstem and the optic nerve in the table shows that the matching degree of 3 times and 5 times in 9 times of registration is A, the matching degree of 6 times and 4 times of registration is B, for heterogeneous data, the perfect matching degree cannot be achieved under rigid and similar registration, but most of matching reaches the purpose that the project can rapidly read clinical tomograms for common patients, interdisciplinary researchers, medical engineering personnel, low-grade medical students and the like, and the table has great advantages in light weight and processing speed. And from the perspective of professional imaging physicians, the registration result is generally considered to be suitable for the popular universality requirement provided by the project.
In some embodiments, the present application provides an automatic positioning method based on a clinical image of a head, which performs fast region-similarity registration and spatial mapping on head region tomographic image data (e.g., CT, MRI) from any clinical patient and a chinese visual human body data set (CVH), marks specific anatomical names of regions of the clinical image data according to an existing general anatomical structure knowledge atlas of the CVH, enables an original abstract clinical tomographic image to be contrasted with the CVH high-definition atlas, and visually presents any anatomical region in the form of multi-planar reconstruction (MPR), volume Reconstruction (VR), and the like through a plurality of retrieval methods, thereby embodying certain technical innovation. The method has the characteristics of simple operation, light operation weight, environmental friendliness and no installation, can visually present the region of interest of the current image to the user in an intelligent navigation mode, realizes the purpose of unmanned intelligent assistance 'explanation', serves the active participation requirements of patients and family members, meets the working requirements of low-grade medical students, cross-subject researchers and medical engineering personnel, and embodies certain application innovation. According to the test result, the method can realize the function of heterologous multi-modal registration, can meet the functional target provided by the text, has better adaptability and repeatability on clinical image data, and is suitable for the general demand of the public; the processing time on a common personal computer is less than 40 seconds, so that the light-weight LED lamp has a good light-weight advantage; and the system can be operated in any version of Windows desktop system, and the NET framework version is more than 4.0, so that the development requirement of the lightweight intelligent positioning system based on head CT/MRI is met.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit and scope of the present invention, and they should be construed as being included in the following claims and description.
Claims (9)
1. An automatic positioning method based on head clinical image is characterized by comprising the following steps:
acquiring a clinical image of the head of a patient;
registering the clinical image of the head of the patient and the CVH to obtain a space direction mapping relation and a space size mapping relation;
performing coordinate transformation on the head anatomical structure of the CVH according to the space direction mapping relation and the space size mapping relation to obtain a CVH registration head anatomical structure; the CVH registration head anatomical structure and the position identification and the organ name of each organ area in the CVH registration head anatomical structure form a new CVH anatomical knowledge base;
performing multi-plane reconstruction on the clinical image of the head of the patient to obtain an anatomical structure of the head of the patient; a real-time mapping relationship exists between the patient's head anatomy and the CVH registered head anatomy;
and selecting a region of interest from the anatomical knowledge map, and automatically positioning the corresponding position of the region of interest in the head anatomy of the patient.
2. The method of claim 1, wherein registering the clinical image of the patient's head with the CVH to obtain a spatial orientation map and a spatial dimension map comprises:
carrying out first registration on the clinical image of the head of the patient and the CVH to obtain a space direction mapping relation and a first registration result;
and performing secondary registration on the clinical image of the head of the patient and the CVH according to the first registration result to obtain a space size mapping relation.
3. The method as claimed in claim 2, wherein the first registration of the clinical image of the patient's head with the CVH to obtain the spatial orientation mapping and the first registration result comprises:
carrying out three-dimensional contour reconstruction on the clinical image of the head of the patient to obtain a three-dimensional contour structure of the patient;
acquiring a first characteristic point according to the three-dimensional contour structure of the patient;
acquiring a second characteristic point of the CVH three-dimensional contour structure;
registering the first characteristic point and the second characteristic point to obtain a spatial direction mapping relation and a first registration result; and the first registration result is a CVH primary registration three-dimensional contour structure, and the spatial direction of the patient three-dimensional contour structure is consistent with that of the CVH primary registration three-dimensional contour structure.
4. The method of claim 3, wherein obtaining a first feature point from the three-dimensional contour structure of the patient comprises:
and respectively selecting the eye-nose triangular areas from the three-dimensional contour structure of the patient and generating first characteristic points of the eye-nose triangular areas.
5. The method as claimed in claim 3, wherein performing a second registration on the clinical image of the head of the patient and the CVH according to the first registration result to obtain a spatial dimension mapping relationship, comprises:
acquiring a first annular contour according to the three-dimensional contour structure of the patient;
acquiring a second annular contour of the CVH three-dimensional contour structure;
and in the primary registration three-dimensional contour structure of the patient and the CVH, registering according to the first annular contour and the second annular contour to obtain a space size mapping relation.
6. The method of claim 3, wherein performing a coordinate transformation on the head anatomy of the CVH according to the spatial direction mapping relationship and the spatial dimension mapping relationship to obtain a CVH registration head anatomy, comprises:
acquiring a homogeneous transformation matrix according to the space direction mapping relation and the space size mapping relation; the homogeneous transformation matrix is used for representing the coordinate transformation relation between the three-dimensional contour structure of the patient and the head anatomical structure of the CVH;
and performing coordinate transformation on the head anatomy structure of the CVH according to the homogeneous transformation matrix to obtain the head anatomy structure of the CVH.
7. The method of claim 6, wherein the homogeneous transformation matrix is obtained by the following formula:
8. The method of claim 1, comprising:
and after automatically positioning the corresponding position of the region of interest in the head anatomical structure of the patient, marking colors on the corresponding position region, and displaying the plane of the corresponding position region.
9. An automatic positioning device for head clinical image, comprising:
an acquisition module configured to acquire clinical images of a patient's head;
a registration module configured to register the clinical image of the head of the patient and the CVH to obtain a spatial direction mapping relation and a spatial dimension mapping relation;
the coordinate transformation module is configured to perform coordinate transformation on the head anatomical structure of the CVH according to the space direction mapping relation and the space size mapping relation to obtain a CVH registration head anatomical structure; the CVH registration head anatomical structure and the position identification and the organ name of each organ area in the CVH registration head anatomical structure form a new CVH anatomical knowledge base;
a multi-plane reconstruction module configured to perform multi-plane reconstruction on the clinical image of the head of the patient to obtain an anatomical structure of the head of the patient; a real-time mapping relationship exists between the patient's head anatomy and the CVH registered head anatomy;
an automatic localization module configured to select a region of interest from the anatomical knowledge-map and automatically localize a corresponding position of the region of interest in the patient's head anatomy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274873.2A CN115690207A (en) | 2022-10-18 | 2022-10-18 | Automatic positioning method and device based on head clinical image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274873.2A CN115690207A (en) | 2022-10-18 | 2022-10-18 | Automatic positioning method and device based on head clinical image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115690207A true CN115690207A (en) | 2023-02-03 |
Family
ID=85065992
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211274873.2A Pending CN115690207A (en) | 2022-10-18 | 2022-10-18 | Automatic positioning method and device based on head clinical image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115690207A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117323002A (en) * | 2023-11-30 | 2024-01-02 | 北京万特福医疗器械有限公司 | Neural endoscopic surgery visualization system based on mixed reality technology |
-
2022
- 2022-10-18 CN CN202211274873.2A patent/CN115690207A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117323002A (en) * | 2023-11-30 | 2024-01-02 | 北京万特福医疗器械有限公司 | Neural endoscopic surgery visualization system based on mixed reality technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Automatic segmentation of individual tooth in dental CBCT images from tooth surface map by a multi-task FCN | |
| US8698795B2 (en) | Interactive image segmentation | |
| US7817836B2 (en) | Methods for volumetric contouring with expert guidance | |
| CN110338840B (en) | Three-dimensional imaging data display processing method and three-dimensional ultrasonic imaging method and system | |
| Blackall et al. | Alignment of sparse freehand 3-D ultrasound with preoperative images of the liver using models of respiratory motion and deformation | |
| CN107909622B (en) | Model generation method, medical imaging scanning planning method and medical imaging system | |
| US20100067761A1 (en) | Automatic interpretation of 3-d medicine images of the brain and methods for producing intermediate results | |
| CN107456278A (en) | A kind of ESS air navigation aid and system | |
| Kok et al. | Articulated planar reformation for change visualization in small animal imaging | |
| US20070177166A1 (en) | Image processing apparatus, an imaging system, a computer program and a method for scaling an object in an image | |
| CN108109170B (en) | Medical image scanning method and medical imaging equipment | |
| CN107221029A (en) | A kind of three-dimensional image reconstruction method | |
| CN107256575A (en) | A kind of three-dimensional tongue based on binocular stereo vision is as method for reconstructing | |
| CN110993067A (en) | Medical image labeling system | |
| CN115690207A (en) | Automatic positioning method and device based on head clinical image | |
| CN116993790B (en) | Planting navigation registration method, system and storage medium | |
| Simmons-Ehrhardt et al. | Open-source tools for dense facial tissue depth mapping of computed tomography models | |
| CN110148208B (en) | Nasopharyngeal radiotherapy teaching model construction method based on Chinese digital person | |
| JP5783627B2 (en) | Human body modeling system | |
| CN116152235A (en) | Cross-modal synthesis method for medical image from CT (computed tomography) to PET (positron emission tomography) of lung cancer | |
| Dzyubachyk et al. | Comparative exploration of whole-body MR through locally rigid transforms | |
| Lee et al. | Facial identification of the dead | |
| CN111127636A (en) | Intelligent desktop-level three-dimensional diagnosis system for complex intra-articular fracture | |
| US20220130128A1 (en) | System and method for normalizing volumetric imaging data of a patient | |
| Richmond | Evaluation of Craniofacial Superimposition as a Technique for Measuring Mountain Gorilla Facial Soft Tissue Depth and Implications for Hominid Facial Approximation |
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 |