CN113674257A - Method, device and equipment for measuring scoliosis angle and storage medium - Google Patents
Method, device and equipment for measuring scoliosis angle and storage medium Download PDFInfo
- Publication number
- CN113674257A CN113674257A CN202110985135.8A CN202110985135A CN113674257A CN 113674257 A CN113674257 A CN 113674257A CN 202110985135 A CN202110985135 A CN 202110985135A CN 113674257 A CN113674257 A CN 113674257A
- Authority
- CN
- China
- Prior art keywords
- image
- positioning
- spine
- vertebral body
- preset
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 206010039722 scoliosis Diseases 0.000 title claims abstract description 90
- 238000000034 method Methods 0.000 title claims abstract description 69
- 238000013136 deep learning model Methods 0.000 claims abstract description 64
- 238000012545 processing Methods 0.000 claims abstract description 61
- 238000005259 measurement Methods 0.000 claims abstract description 42
- 238000001514 detection method Methods 0.000 claims abstract description 40
- 238000005452 bending Methods 0.000 claims description 43
- 238000010606 normalization Methods 0.000 claims description 37
- 238000012549 training Methods 0.000 claims description 34
- 210000000115 thoracic cavity Anatomy 0.000 claims description 32
- 238000012163 sequencing technique Methods 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000005764 inhibitory process Effects 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 10
- 238000013473 artificial intelligence Methods 0.000 abstract description 6
- 238000004364 calculation method Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 10
- 238000003745 diagnosis Methods 0.000 description 6
- 239000011436 cob Substances 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 3
- 210000000038 chest Anatomy 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000002591 computed tomography Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000002595 magnetic resonance imaging Methods 0.000 description 2
- 210000001015 abdomen Anatomy 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 210000002216 heart Anatomy 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 210000003127 knee Anatomy 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 210000002784 stomach Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000002604 ultrasonography Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the field of artificial intelligence, is applied to the field of intelligent medical treatment, and discloses a method, a device, equipment and a storage medium for measuring a scoliosis angle, which are used for improving the efficiency of measuring the scoliosis angle. The method for measuring the scoliosis angle comprises the following steps: extracting pixel values and normalizing the image to be detected to obtain a preprocessed image; calling a plurality of preset deep learning models, and positioning the preprocessed image to obtain a spine positioning result; calling a preset target key point detection model, and analyzing a spine positioning result to obtain a plurality of centrum key point information; and calling a preset least square method to perform fitting processing on the information of the key points of the multiple vertebral bodies to obtain a fitting straight line segment, determining a bending center based on the fitting straight line segment, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measurement result. In addition, the invention also relates to a block chain technology, and the measurement result of the scoliosis angle can be stored in the block chain node.
Description
Technical Field
The invention relates to the field of region extraction, in particular to a method, a device, equipment and a storage medium for measuring a side bending angle of a spine.
Background
The Cobb angle, which is the angle of scoliosis, is used for clinical quantitative evaluation of the severity of scoliosis, after a patient shoots an X-ray film, an imaging physician manually determines a central vertebral body and upper and lower vertebrae of scoliosis, and measures an end intervertebral included angle through an angle measuring instrument.
In the prior art, a semiautomatic segmentation algorithm is used for an image to automatically measure and calculate the Cobb angle, but the method cannot realize the measurement of the scoliosis angle through the position positioning of a spinal column segment, only can be used as auxiliary work, and still needs an imaging physician to participate in judgment, so that the measurement efficiency of the scoliosis angle is low.
Disclosure of Invention
The invention provides a method, a device, equipment and a storage medium for measuring a scoliosis angle, which are used for positioning a spine of a preprocessed image by calling a plurality of preset deep learning models to obtain a spine positioning result, calling a preset target key point detection model to analyze the spine positioning result to obtain information of a plurality of centrum key points, and measuring and calculating the scoliosis angle based on the information of the plurality of centrum key points, thereby improving the measuring efficiency of the scoliosis angle.
The invention provides a method for measuring the side bending angle of a spine, which comprises the following steps: acquiring an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image; calling a plurality of preset deep learning models, and sequentially carrying out spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result; calling a preset target key point detection model, and analyzing the spine positioning result to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of spine vertebral body; and calling a preset least square method to perform fitting processing on the information of the plurality of centrum key points to obtain a fitting straight line section, determining a bending center based on the fitting straight line section, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measurement result.
Optionally, in a first implementation manner of the first aspect of the present invention, the invoking multiple preset deep learning models to sequentially perform spine positioning, spine vertebral body positioning, and positioning of the 12 th thoracic vertebra on the preprocessed image, and obtaining a spine positioning result includes: calling a preset first deep learning model, and positioning the preprocessed image based on the region where the spine is located to obtain a first positioning image, wherein the first positioning image comprises spine position information; calling a preset second deep learning model, and positioning the first positioning image based on the region where the vertebral body of the spine is located to obtain a second positioning image, wherein the second positioning image comprises position information of the vertebral body of the spine; and calling a preset third deep learning model, positioning the second positioning image based on the region where the 12 th thoracic vertebra is located to obtain a third positioning image, and determining the first positioning image, the second positioning image and the third positioning image as a spine positioning result, wherein the third positioning image comprises the position information of the 12 th thoracic vertebra.
Optionally, in a second implementation manner of the first aspect of the present invention, the calling a preset first deep learning model, and performing positioning on the preprocessed image based on a region where a spine is located to obtain a first positioning image, where the first positioning image includes spine position information, and the positioning includes: calling a preset first deep learning model, and carrying out coding processing on the preprocessed image to obtain a coding hidden layer space vector; carrying out full-connection processing on the coded hidden layer space vector to obtain an initial spine region coordinate, calling a preset non-maximum inhibition algorithm, and screening the initial spine region coordinate to obtain a target spine region coordinate; and substituting the target spine region coordinates into the preprocessed image to obtain a first positioning image, wherein the first positioning image comprises spine position information.
Optionally, in a third implementation manner of the first aspect of the present invention, the calling a preset target keypoint detection model, and analyzing the spine positioning result to obtain information of a plurality of vertebral body keypoints, where the information of the plurality of vertebral body keypoints includes position information of four vertices corresponding to each section of spine vertebral body: calling a preset target key point detection model, and identifying a plurality of vertebral body boundary points in the spine positioning result; and extracting coordinates of each vertebral body boundary point in the plurality of vertebral body boundary points to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of vertebral body of the spine.
Optionally, in a fourth implementation manner of the first aspect of the present invention, the fitting the plurality of centrum key point information by using a preset least square method to obtain a fitted straight line segment, determining a bending center based on the fitted straight line segment, and performing measurement and calculation of a scoliosis angle according to the bending center to obtain a scoliosis angle measurement result includes: determining a centrum center point of each centrum and a plurality of centrum end lines corresponding to each centrum based on the plurality of centrum key point information, and fitting the centrum center point of each centrum by using a preset least square method to obtain a fitting straight line segment; calculating the distance between the center point of each cone and the fitted straight line segment to obtain the center distance value of each cone, obtaining the cone length of each cone, and dividing the center distance value of each cone by the corresponding cone length to obtain a plurality of distance ratios; sequencing the distance ratios in a descending order to obtain distance ratio sequencing results, and determining the centrum center point corresponding to the distance ratio ranked as the first in the distance ratio sequencing results as a bending center; and traversing and calculating included angles formed by a plurality of cone end lines around the bending center to obtain a plurality of bending included angles, sequencing the bending included angles according to a sequence from large to small to obtain a bending included angle sequencing result, and determining the bending included angle with the first rank in the bending included angle sequencing result as a spinal column side bending angle measurement result.
Optionally, in a fifth implementation manner of the first aspect of the present invention, the acquiring an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image includes: acquiring an image to be detected, and reading a pixel value of the image to be detected to obtain an image pixel value; and carrying out normalization processing on the image pixel values to obtain an image array after the normalization processing, and storing the image array after the normalization processing as a gray-scale image to obtain a preprocessed image.
Optionally, in a sixth implementation manner of the first aspect of the present invention, before the obtaining an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image, the method further includes: obtaining model training data, preprocessing the model training data to obtain preprocessed training data, wherein the model training data is used for indicating a spine image set; and training a preset initial key point detection model according to the preprocessed training data to obtain a target key point detection model.
A second aspect of the present invention provides a device for measuring a scoliosis angle, comprising: the acquisition module is used for acquiring an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image; the positioning module is used for calling a plurality of preset deep learning models, and sequentially carrying out spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result; the analysis module is used for calling a preset target key point detection model and analyzing the spine positioning result to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of spine vertebral body; and the measuring and calculating module is used for calling a preset least square method to perform fitting processing on the information of the plurality of centrum key points to obtain a fitting straight line section, determining a bending center based on the fitting straight line section, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measuring result.
Optionally, in a first implementation manner of the second aspect of the present invention, the positioning module includes: the first positioning unit is used for calling a preset first deep learning model and positioning the preprocessed image based on the region where the spine is located to obtain a first positioning image, and the first positioning image comprises spine position information; the second positioning unit is used for calling a preset second depth learning model, positioning the first positioning image based on the region where the vertebral body of the spine is located to obtain a second positioning image, and the second positioning image comprises position information of the vertebral body of the spine; and the third positioning unit is used for calling a preset third deep learning model, positioning the second positioning image based on the region where the 12 th thoracic vertebra is located to obtain a third positioning image, and determining the first positioning image, the second positioning image and the third positioning image as a spine positioning result, wherein the third positioning image comprises the position information of the 12 th thoracic vertebra.
Optionally, in a second implementation manner of the second aspect of the present invention, the first positioning unit is specifically configured to: calling a preset first deep learning model, and carrying out coding processing on the preprocessed image to obtain a coding hidden layer space vector; carrying out full-connection processing on the coded hidden layer space vector to obtain an initial spine region coordinate, calling a preset non-maximum inhibition algorithm, and screening the initial spine region coordinate to obtain a target spine region coordinate; and substituting the target spine region coordinates into the preprocessed image to obtain a first positioning image, wherein the first positioning image comprises spine position information.
Optionally, in a third implementation manner of the second aspect of the present invention, the analysis module includes: the identification unit is used for calling a preset target key point detection model and identifying a plurality of vertebral body boundary points in the spine positioning result; and the extraction unit is used for extracting the coordinates of each vertebral body boundary point in the plurality of vertebral body boundary points to obtain a plurality of vertebral body key point information, and the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of vertebral body.
Optionally, in a fourth implementation manner of the second aspect of the present invention, the measurement and calculation module includes: the fitting unit is used for determining the centrum center point of each centrum and a plurality of centrum end lines corresponding to each centrum based on the plurality of centrum key point information, and fitting the centrum center point of each centrum by using a preset least square method to obtain a fitting straight line segment; the calculation unit is used for calculating the distance between the central point of each cone and the fitted straight line segment to obtain the central distance value of each cone, acquiring the cone length of each cone, and dividing the central distance value of each cone by the corresponding cone length to obtain a plurality of distance ratios; the sorting unit is used for sorting the distance ratios in a descending order to obtain distance ratio sorting results, and determining the centrum center point corresponding to the distance ratio which is ranked as the first in the distance ratio sorting results as a bending center; and the determining unit is used for traversing and calculating included angles formed by a plurality of vertebral body end lines around the bending center to obtain a plurality of bending included angles, sequencing the bending included angles according to a sequence from large to small to obtain a bending included angle sequencing result, and determining the bending included angle with the first rank in the bending included angle sequencing result as a spinal column side bending angle measurement result.
Optionally, in a fifth implementation manner of the second aspect of the present invention, the obtaining module includes: the reading unit is used for acquiring an image to be detected and reading a pixel value of the image to be detected to obtain an image pixel value; and the normalization unit is used for performing normalization processing on the image pixel values to obtain an image array after the normalization processing, and storing the image array after the normalization processing as a gray-scale image to obtain a preprocessed image.
Optionally, in a sixth implementation manner of the second aspect of the present invention, before the obtaining module, the apparatus for measuring a scoliosis angle further includes a training module, where the training module includes: the preprocessing unit is used for acquiring model training data, preprocessing the model training data to obtain preprocessed training data, and the model training data is used for indicating a spine image set; and the training unit is used for training a preset initial key point detection model according to the preprocessed training data to obtain a target key point detection model.
A third aspect of the present invention provides a device for measuring a scoliosis angle, comprising: a memory and at least one processor, the memory having stored therein a computer program; the at least one processor invokes the computer program in the memory to cause the apparatus for measuring scoliosis angle to perform the method for measuring scoliosis angle described above.
A fourth aspect of the present invention provides a computer-readable storage medium having stored therein a computer program which, when run on a computer, causes the computer to execute the above-described method of measuring a scoliosis angle.
According to the technical scheme provided by the invention, an image to be detected is obtained, and pixel value extraction and normalization processing are carried out on the image to be detected to obtain a preprocessed image; calling a plurality of preset deep learning models, and sequentially carrying out spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result; calling a preset target key point detection model, and analyzing the spine positioning result to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of spine vertebral body; and calling a preset least square method to perform fitting processing on the information of the plurality of centrum key points to obtain a fitting straight line section, determining a bending center based on the fitting straight line section, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measurement result. In the embodiment of the invention, the spine positioning is carried out on the preprocessed image by calling the preset multiple deep learning models to obtain the spine positioning result, the preset target key point detection model is called to analyze the spine positioning result to obtain the information of the multiple vertebral body key points, the measurement and calculation of the scoliosis angle are carried out based on the information of the multiple vertebral body key points, and the measurement efficiency of the scoliosis angle is improved.
Drawings
FIG. 1 is a schematic view of an embodiment of a method for measuring scoliosis angle in accordance with an embodiment of the present invention;
FIG. 2 is a schematic view of another embodiment of the method for measuring scoliosis angle according to the present invention;
FIG. 3 is a schematic view of an embodiment of a device for measuring scoliosis angle in accordance with an embodiment of the present invention;
FIG. 4 is a schematic view of another embodiment of the device for measuring scoliosis angle in the embodiment of the present invention;
fig. 5 is a schematic view of an embodiment of the device for measuring the scoliosis angle in the embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a method, a device, equipment and a storage medium for measuring a scoliosis angle, which are used for positioning a spine of a preprocessed image by calling a plurality of preset deep learning models to obtain a spine positioning result, calling a preset target key point detection model to analyze the spine positioning result to obtain information of a plurality of centrum key points, and measuring and calculating the scoliosis angle based on the information of the plurality of centrum key points, so that the measuring efficiency of the scoliosis angle is improved.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
For ease of understanding, a specific flow of an embodiment of the present invention is described below, with reference to fig. 1, an embodiment of a method for measuring a scoliosis angle in an embodiment of the present invention includes:
101. and acquiring an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image.
It is to be understood that the executing body of the present invention may be a device for measuring the spinal column side angle, and may also be a terminal or a server, which is not limited herein. The embodiment of the present invention is described by taking a server as an execution subject. The server may be an independent server, or may be a cloud server that provides basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), and a big data and artificial intelligence platform.
The embodiment of the application can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.
The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.
The server obtains an image to be detected, and pixel value extraction and normalization processing are carried out on the image to be detected to obtain a preprocessed image. The image to be detected is in a digital imaging and communications in medicine (DICOM) format, the image to be detected is a spine full-length X-ray image, a patient is detected through a preset medical instrument to obtain a spine full-length X-ray image, the medical instrument uploads the spine full-length X-ray image to a medical cloud, the spine full-length X-ray image can also be stored in a health archive corresponding to the patient, the server obtains the image to be detected from the medical cloud according to patient information, the image to be detected in the embodiment is authorized by a user, the server reads a pixel value of the image to be detected and performs normalization processing after obtaining the image to be detected, and a preprocessed image is obtained.
In a medical application scenario, an image to be detected is a medical image, and the type of an object included in the image to be detected is a focus, i.e., a part of an organism where a lesion occurs. Medical images refer to images of internal tissues, e.g., stomach, abdomen, heart, knee, brain, which are obtained in a non-invasive manner for medical treatment or medical research, such as images generated by medical instruments, e.g., CT (Computed Tomography), MRI (Magnetic Resonance Imaging), US (ultrasound), X-ray images, electroencephalograms, and photo lamps.
102. And calling a plurality of preset deep learning models, and sequentially carrying out spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result.
And the server calls a plurality of preset deep learning models, and carries out spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image in sequence to obtain a spine positioning result. In this embodiment, all of the multiple deep learning models used are target detection models (SSD), in this embodiment, three SSD models are collectively called, a first deep learning model is used to locate a region where a spine (thoracic vertebra and lumbar vertebra) is located on a preprocessed image, an output of the first deep learning model is an image including coordinates of a spine region, a second deep learning model is used to locate a region where a spine vertebral body is located and confirm a position of the spine vertebral body, an input of the second deep learning model is an output of the first deep learning model, a third deep learning model is used to locate a 12 th thoracic vertebra in the spine, an input of the third deep learning model is an output of the second deep learning model, and a finally obtained spine location result is an output result of the three deep learning models.
103. And calling a preset target key point detection model, analyzing the spine positioning result, and obtaining a plurality of centrum key point information, wherein the plurality of centrum key point information comprises position information of four vertexes corresponding to each spine.
And the server calls a preset target key point detection model, analyzes the spine positioning result and obtains a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprise position information of four vertexes corresponding to each section of spine vertebral body. The target key point detection model is used for identifying vertebral body key points in a spinal positioning result to obtain vertex positions (position information of four vertexes) corresponding to each section of spinal vertebral body, namely, information of a plurality of vertebral body key points.
104. And calling a preset least square method to perform fitting processing on the information of the key points of the multiple vertebral bodies to obtain a fitting straight line segment, determining a bending center based on the fitting straight line segment, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measurement result.
The server calls a preset least square method to perform fitting processing on the information of the key points of the multiple vertebral bodies to obtain a fitting straight line segment, a bending center is determined based on the fitting straight line segment, and the measurement and calculation of the lateral bending angle of the vertebral column are performed according to the bending center to obtain the measurement result of the lateral bending angle of the vertebral column. The implementation process of measuring and calculating the scoliosis angle mainly comprises the following steps: obtaining the central points of the end lines of the vertebral bodies and the central points of the end lines of the vertebral bodies according to the information of a plurality of key points of the vertebral bodies, fitting the central points of the end lines of the vertebral bodies to obtain a fitted straight line segment, obtaining a distance ratio by calculating the distance between the central point and the fitted straight line segment and dividing the distance by the width of the vertebral bodies to determine a bending center, traversing and calculating the included angle between each end line of the vertebral bodies above and each end line of the vertebral bodies below the bending center to finally obtain the measurement result of the lateral curvature angle of the vertebral column, substituting the measurement result of the lateral curvature angle of the vertebral column into a preset image report template to automatically output an image diagnosis report comprising the vertebral bodies and the bending direction of the bending center, a Cobb angle, the position of the end vertebrae and the like, and the embodiment can realize the full-automatic process from image input to diagnosis report output without manual participation, avoids human errors and has higher reliability, for example, in the medical field, the medical record information and diagnosis report required by a user can be inquired from a mass of electronic medical records based on an artificial intelligent model, and the medical record reference is provided for the user.
In the embodiment of the invention, the spine positioning is carried out on the preprocessed image by calling the preset multiple deep learning models to obtain the spine positioning result, the preset target key point detection model is called to analyze the spine positioning result to obtain the information of the multiple vertebral body key points, the measurement and calculation of the scoliosis angle are carried out based on the information of the multiple vertebral body key points, and the measurement efficiency of the scoliosis angle is improved. This scheme can be applied to in the wisdom medical field to promote the construction in wisdom city.
Referring to fig. 2, another embodiment of the method for measuring a scoliosis angle according to the embodiment of the present invention includes:
201. and acquiring an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image.
The server obtains an image to be detected, and pixel value extraction and normalization processing are carried out on the image to be detected to obtain a preprocessed image. Specifically, the server acquires an image to be detected, reads a pixel value of the image to be detected, and obtains an image pixel value; and the server performs normalization processing on the image pixel values to obtain an image array after the normalization processing, and stores the image array after the normalization processing as a gray-scale image to obtain a preprocessed image.
After acquiring an image to be detected in a DICOM format, a server extracts a pixel value of the image to be detected to obtain an image pixel value, wherein the image pixel value is generally between-4000 and 4000, an image array after normalization processing is obtained by performing normalization processing on the image pixel value, the image array after normalization processing is between 0 and 1, the image array after normalization processing is stored as a gray-scale image, and a preprocessed image in a bitmap (PNG) format of a lossless compression algorithm is finally obtained.
202. And calling a preset first deep learning model, and positioning the preprocessed image based on the region where the spine is located to obtain a first positioning image, wherein the first positioning image comprises spine position information.
The server calls a preset first deep learning model, and carries out positioning based on the region where the spine is located on the preprocessed image to obtain a first positioning image, wherein the first positioning image comprises spine position information. Specifically, the server calls a preset first deep learning model to perform coding processing on the preprocessed image to obtain a coding hidden layer space vector; the server performs full-connection processing on the coded hidden layer space vector to obtain an initial spine region coordinate, and calls a preset non-maximum inhibition algorithm to screen the initial spine region coordinate to obtain a target spine region coordinate; and substituting the coordinates of the target spine region into the preprocessed image by the server to obtain a first positioning image, wherein the first positioning image comprises spine position information. In the pre-training stage, 200 spine full-length X-ray pictures are required to be obtained and manually labeled to obtain spine coordinate values, the spine coordinate values are input into a first SSD network model as a target and are trained together with the pictures, a trained model (namely a first deep learning model) is obtained after iterative fitting, a preprocessed image is coded by calling the first deep learning model to obtain coded hidden layer space vectors, the coded hidden layer space vectors are fully connected through a full connection layer in the first deep learning model to obtain initial spine region coordinates, a Non Maximum Suppression (NMS) algorithm is an algorithm widely used in the fields of target detection, positioning and the like, and is mainly used for filtering the initial spine region coordinates to obtain target spine region coordinates, substituting the target spine region coordinates into the preprocessed image, and finally obtaining a first positioning image.
203. And calling a preset second deep learning model, and positioning the first positioning image based on the region where the vertebral body is located to obtain a second positioning image, wherein the second positioning image comprises position information of the vertebral body.
The data processing process of the second deep learning model is consistent with that of the first deep learning model, the input of the second deep learning is the output of the first deep learning (namely the first positioning image), and in the pre-training stage, manually marking 200X-ray pictures of vertebral bodies of the spine, marking out the position information of each vertebral column, inputting the position information of each vertebral column and the pictures of the vertebral bodies of the spine into a second SSD network model for training, obtaining a trained model (namely a second deep learning model) after iterative fitting, removing the interference of the cervical vertebral bodies on the detection of the thoracic vertebral bodies and the lumbar vertebral bodies by using the pictures of the vertebral column region for training, calling the second deep learning model, and positioning the first positioning image to finally obtain a second positioning image, wherein the second positioning image comprises position information (namely the vertebral body coordinates) of the vertebral body of the vertebral column.
204. And calling a preset third deep learning model, positioning the second positioning image based on the region where the 12 th thoracic vertebra is located to obtain a third positioning image, and determining the first positioning image, the second positioning image and the third positioning image as a spine positioning result, wherein the third positioning image comprises the position information of the 12 th thoracic vertebra.
The data processing process of the third deep learning model is consistent with the first deep learning model and the second deep learning model, the input of the third deep learning model is the output of the second deep learning model (namely the second positioning image), in the pre-training stage, manually labeling 200 full-length X-ray pictures of the spine, labeling the position of the 12 th thoracic vertebra, inputting the position information of the 12 th thoracic vertebra and the full-length X-ray pictures of the spine into a third SSD network model for training, obtaining a trained model (namely a third deep learning model) after iterative fitting, calling the third deep learning model, and positioning the second positioning image to finally obtain a third positioning image, wherein the third positioning image comprises the position information of the 12 th thoracic vertebra (namely the coordinates of the 12 th thoracic vertebra), and finally the first positioning image, the second positioning image and the third positioning image are determined as a spine positioning result.
205. And calling a preset target key point detection model, analyzing the spine positioning result, and obtaining a plurality of centrum key point information, wherein the plurality of centrum key point information comprises position information of four vertexes corresponding to each spine.
And the server calls a preset target key point detection model, analyzes the spine positioning result and obtains a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprise position information of four vertexes corresponding to each section of spine vertebral body. Specifically, the server calls a preset target key point detection model and identifies a plurality of vertebral body boundary points in the spine positioning result; the server extracts coordinates of each vertebral body boundary point in the plurality of vertebral body boundary points to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of vertebral body. In the step, a deep learning key point detection network HRNET trained in advance is used, spinal positioning results are input into a key point detection model, four vertex positions (namely key point information of a plurality of vertebral bodies) of each section of spinal vertebral body are obtained, after the key point information of the vertebral bodies is obtained, each section of spinal vertebral body can be numbered, the position of the 12 th thoracic vertebra is taken as a reference, the intersection and comparison between the position of each remaining section of spinal body and the position of the 12 th thoracic vertebra is calculated in a traversing mode, the vertebral body with the largest intersection and comparison is the 12 th thoracic vertebra, and then the thoracic part from the chest 1 to the chest 12 and the lumbar part from the waist 1 to the waist 5 are numbered sequentially according to the position coordinates of each spinal part from top to bottom, so that an image diagnosis report can be generated automatically in the later stage.
206. And calling a preset least square method to perform fitting processing on the information of the key points of the multiple vertebral bodies to obtain a fitting straight line segment, determining a bending center based on the fitting straight line segment, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measurement result.
The server calls a preset least square method to perform fitting processing on the information of the key points of the multiple vertebral bodies to obtain a fitting straight line segment, a bending center is determined based on the fitting straight line segment, and the measurement and calculation of the lateral bending angle of the vertebral column are performed according to the bending center to obtain the measurement result of the lateral bending angle of the vertebral column. Specifically, the server determines a centrum center point of each centrum and a plurality of centrum end lines corresponding to each centrum based on a plurality of centrum key point information, and a preset least square method is called to fit the centrum center point of each centrum to obtain a fitting straight line segment; the server calculates the distance between the center point of each cone and the fitted straight line segment to obtain the center distance value of each cone, obtains the cone length of each cone, and divides the center distance value of each cone by the corresponding cone length to obtain a plurality of distance ratios; the server sorts the distance ratios in the descending order to obtain distance ratio sorting results, and the central point of the vertebral body corresponding to the distance ratio ranked first in the distance ratio sorting results is determined as a bending center; the server calculates the included angle formed by the end lines of the vertebral bodies around the bending center in a traversing mode to obtain a plurality of bending included angles, the bending included angles are sequenced from large to small to obtain a sequencing result of the bending included angles, and the bending included angle with the first rank in the sequencing result of the bending included angles is determined as a measurement result of the side bending angle of the spinal column.
The server obtains end lines of the vertebral bodies and central points of the vertebral bodies according to four vertex positions of each section of the vertebral bodies, performs fitting by a least square method based on coordinates of the central points of the vertebral bodies to obtain a fitted straight line segment, calculates the distance between the central points of the vertebral bodies according to the straight line, divides the distance by the average width of the vertebral bodies to obtain distance ratios, wherein the central point of the vertebral body corresponding to the maximum distance ratio is a bending center (namely the central point of the vertebral body corresponding to the distance ratio with the first rank in the distance ratio sequencing result), and calculates an included angle between each end line of the vertebral body above the bending center and each end line of the vertebral body below the bending center in a traversing manner, the included angle with the maximum included angle is a Cobb angle (namely the bending included angle with the first rank in the bending included angle sequencing result), and finally obtains a measurement result of the lateral bending angle of the vertebral column and substitutes the measurement result of the lateral bending angle into a preset image report template to automatically output an image diagnosis report, the method comprises the bending of the central vertebral body, the bending direction, the Cobb angle, the position of the end vertebral body and the like, can realize the full-automatic process from image input to diagnosis report output, does not need manual participation, avoids human errors, and has higher reliability.
In the embodiment of the invention, the spine positioning is carried out on the preprocessed image by calling the preset multiple deep learning models to obtain the spine positioning result, the preset target key point detection model is called to analyze the spine positioning result to obtain the information of the multiple vertebral body key points, the measurement and calculation of the scoliosis angle are carried out based on the information of the multiple vertebral body key points, and the measurement efficiency of the scoliosis angle is improved. This scheme can be applied to in the wisdom medical field to promote the construction in wisdom city.
Referring to fig. 3, the method for measuring a scoliosis angle according to an embodiment of the present invention is described above, and an embodiment of the apparatus for measuring a scoliosis angle according to an embodiment of the present invention includes:
an obtaining module 301, configured to obtain an image to be detected, and perform pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image;
the positioning module 302 is configured to call a plurality of preset deep learning models, and sequentially perform spine positioning, spine vertebral body positioning, and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result;
the analysis module 303 is configured to call a preset target key point detection model, and analyze a spine positioning result to obtain a plurality of vertebral body key point information, where the plurality of vertebral body key point information includes position information of four vertexes corresponding to each vertebral body of the spine;
the measuring and calculating module 304 is configured to call a preset least square method to perform fitting processing on the information of the plurality of centrum key points to obtain a fitting straight line segment, determine a bending center based on the fitting straight line segment, and measure and calculate a scoliosis angle according to the bending center to obtain a scoliosis angle measurement result.
In the embodiment of the invention, the spine positioning is carried out on the preprocessed image by calling the preset multiple deep learning models to obtain the spine positioning result, the preset target key point detection model is called to analyze the spine positioning result to obtain the information of the multiple vertebral body key points, the measurement and calculation of the scoliosis angle are carried out based on the information of the multiple vertebral body key points, and the measurement efficiency of the scoliosis angle is improved. This scheme can be applied to in the wisdom medical field to promote the construction in wisdom city.
Referring to fig. 4, another embodiment of the apparatus for measuring a scoliosis angle according to the present invention includes:
an obtaining module 301, configured to obtain an image to be detected, and perform pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image;
the positioning module 302 is configured to call a plurality of preset deep learning models, and sequentially perform spine positioning, spine vertebral body positioning, and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result;
wherein, the positioning module 302 comprises:
the first positioning unit 3021 is configured to invoke a preset first deep learning model, and perform positioning based on a region where a spine is located on the preprocessed image to obtain a first positioning image, where the first positioning image includes spine position information;
the second positioning unit 3022 is configured to call a preset second deep learning model, perform positioning based on a region where the vertebral body is located on the first positioning image, and obtain a second positioning image, where the second positioning image includes position information of the vertebral body;
a third positioning unit 3023, configured to invoke a preset third depth learning model, perform positioning based on a region where the 12 th thoracic vertebra is located on the second positioning image, to obtain a third positioning image, and determine the first positioning image, the second positioning image, and the third positioning image as a spine positioning result, where the third positioning image includes position information of the 12 th thoracic vertebra;
the analysis module 303 is configured to call a preset target key point detection model, and analyze a spine positioning result to obtain a plurality of vertebral body key point information, where the plurality of vertebral body key point information includes position information of four vertexes corresponding to each vertebral body of the spine;
the measuring and calculating module 304 is configured to call a preset least square method to perform fitting processing on the information of the plurality of centrum key points to obtain a fitting straight line segment, determine a bending center based on the fitting straight line segment, and measure and calculate a scoliosis angle according to the bending center to obtain a scoliosis angle measurement result.
Optionally, the obtaining module 301 includes:
the reading unit 3011 is configured to obtain an image to be detected, and read a pixel value of the image to be detected to obtain an image pixel value;
and the normalization unit 3012 is configured to perform normalization processing on the image pixel values to obtain an image array after the normalization processing, and store the image array after the normalization processing as a grayscale map to obtain a preprocessed image.
Optionally, the first positioning unit module 3021 may be further specifically configured to:
calling a preset first deep learning model, and carrying out coding processing on the preprocessed image to obtain a coding hidden layer space vector; carrying out full-connection processing on the coded hidden layer space vector to obtain an initial spine region coordinate, calling a preset non-maximum inhibition algorithm, and screening the initial spine region coordinate to obtain a target spine region coordinate; and substituting the coordinates of the target spine region into the preprocessed image to obtain a first positioning image, wherein the first positioning image comprises spine position information.
Optionally, the analysis module 303 includes:
the identification unit 3031 is configured to invoke a preset target key point detection model and identify a plurality of vertebral body boundary points in a spine positioning result;
the extracting unit 3032 is configured to extract coordinates of each vertebral body boundary point of the plurality of vertebral body boundary points to obtain a plurality of vertebral body key point information, where the plurality of vertebral body key point information includes position information of four vertexes corresponding to each vertebral body segment.
Optionally, the calculating module 304 includes:
the fitting unit 3041 is configured to determine a central point of each vertebral body and a plurality of end lines of each vertebral body corresponding to each vertebral body based on the information of the key points of the plurality of vertebral bodies, and fit the central point of each vertebral body by using a preset least square method to obtain a fitted straight line segment;
a calculating unit 3042, configured to calculate a distance between a center point of each vertebral body and the fitted straight line segment, to obtain a center distance value of each vertebral body, obtain a length of each vertebral body, and divide the center distance value of each vertebral body by the length of the corresponding vertebral body, to obtain a plurality of distance ratios;
the sorting unit 3043 is configured to sort the distance ratios in a descending order to obtain a distance ratio sorting result, and determine a centrum center point corresponding to a distance ratio ranked first in the distance ratio sorting result as a bending center;
the determining unit 3044 is configured to traverse and calculate included angles formed by a plurality of vertebral body end lines around a bending center to obtain a plurality of bending included angles, sort the plurality of bending included angles according to a descending order to obtain a bending included angle sorting result, and determine a bending included angle ranked first in the bending included angle sorting result as a spinal column side bending angle measurement result.
Optionally, before the obtaining module 301, the device for measuring the scoliosis angle further includes a training module 305, including:
the preprocessing unit 3051 is configured to obtain model training data, preprocess the model training data to obtain preprocessed training data, where the model training data is used to indicate a spine image set;
and the training unit 3052, configured to train a preset initial keypoint detection model according to the preprocessed training data, so as to obtain a target keypoint detection model.
In the embodiment of the invention, the spine positioning is carried out on the preprocessed image by calling the preset multiple deep learning models to obtain the spine positioning result, the preset target key point detection model is called to analyze the spine positioning result to obtain the information of the multiple vertebral body key points, the measurement and calculation of the scoliosis angle are carried out based on the information of the multiple vertebral body key points, and the measurement efficiency of the scoliosis angle is improved. This scheme can be applied to in the wisdom medical field to promote the construction in wisdom city.
Fig. 3 and 4 above describe the apparatus for measuring the scoliosis angle in the embodiment of the present invention in detail from the perspective of the modular functional entity, and the apparatus for measuring the scoliosis angle in the embodiment of the present invention is described in detail from the perspective of the hardware processing.
Fig. 5 is a schematic structural diagram of a device for measuring a scoliosis angle, according to an embodiment of the present invention, the device 500 for measuring a scoliosis angle may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 510 (e.g., one or more processors) and a memory 520, one or more storage media 530 (e.g., one or more mass storage devices) storing applications 533 or data 532. Memory 520 and storage media 530 may be, among other things, transient or persistent storage. The program stored on the storage medium 530 may include one or more modules (not shown), each of which may include a series of computer program operations in the apparatus 500 for measuring scoliosis angle. Still further, the processor 510 may be configured to communicate with the storage medium 530, and execute a series of computer program operations in the storage medium 530 on the scoliosis angle measuring apparatus 500.
The scoliosis angle measurement apparatus 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input-output interfaces 560, and/or one or more operating systems 531, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, and the like. Those skilled in the art will appreciate that the configuration of the measurement device for the angle of scoliosis shown in fig. 5 does not constitute a limitation of the measurement device for the angle of scoliosis, and may include more or fewer components than those shown, or some components in combination, or a different arrangement of components.
The application also provides a spinal column side bend angle's measuring equipment, includes: a memory having a computer program stored therein and at least one processor, the memory and the at least one processor interconnected by a line; the at least one processor invokes the computer program in the memory to cause the apparatus for measuring scoliosis angle to perform the steps of the method for measuring scoliosis angle described above.
The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, which may also be a volatile computer-readable storage medium, having stored thereon a computer program which, when run on a computer, causes the computer to perform the steps of the method for measuring scoliosis angle.
Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes several computer programs to enable a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for measuring a scoliosis angle, the method comprising:
acquiring an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image;
calling a plurality of preset deep learning models, and sequentially carrying out spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result;
calling a preset target key point detection model, and analyzing the spine positioning result to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of spine vertebral body;
and calling a preset least square method to perform fitting processing on the information of the plurality of centrum key points to obtain a fitting straight line section, determining a bending center based on the fitting straight line section, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measurement result.
2. The method for measuring the lateral curvature angle of the spine according to claim 1, wherein the calling of a plurality of preset deep learning models sequentially performs spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image, and obtaining a spine positioning result comprises:
calling a preset first deep learning model, and positioning the preprocessed image based on the region where the spine is located to obtain a first positioning image, wherein the first positioning image comprises spine position information;
calling a preset second deep learning model, and positioning the first positioning image based on the region where the vertebral body of the spine is located to obtain a second positioning image, wherein the second positioning image comprises position information of the vertebral body of the spine;
and calling a preset third deep learning model, positioning the second positioning image based on the region where the 12 th thoracic vertebra is located to obtain a third positioning image, and determining the first positioning image, the second positioning image and the third positioning image as a spine positioning result, wherein the third positioning image comprises the position information of the 12 th thoracic vertebra.
3. The method for measuring the scoliosis angle according to claim 2, wherein the step of calling a preset first deep learning model to perform positioning based on a region where a spine is located on the preprocessed image to obtain a first positioning image, wherein the first positioning image comprises spine position information, and the method comprises the steps of:
calling a preset first deep learning model, and carrying out coding processing on the preprocessed image to obtain a coding hidden layer space vector;
carrying out full-connection processing on the coded hidden layer space vector to obtain an initial spine region coordinate, calling a preset non-maximum inhibition algorithm, and screening the initial spine region coordinate to obtain a target spine region coordinate;
and substituting the target spine region coordinates into the preprocessed image to obtain a first positioning image, wherein the first positioning image comprises spine position information.
4. The method for measuring scoliosis angle according to claim 1, wherein the calling of a preset target key point detection model analyzes the spine positioning result to obtain a plurality of vertebral body key point information, and the plurality of vertebral body key point information including position information of four vertexes corresponding to each section of spine vertebral body comprises:
calling a preset target key point detection model, and identifying a plurality of vertebral body boundary points in the spine positioning result;
and extracting coordinates of each vertebral body boundary point in the plurality of vertebral body boundary points to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of vertebral body of the spine.
5. The method for measuring the scoliosis angle according to claim 1, wherein the fitting of the plurality of centrum key point information by using a preset least square method to obtain a fitted straight line segment, the determination of a bending center based on the fitted straight line segment, and the measurement of the scoliosis angle according to the bending center comprise:
determining a centrum center point of each centrum and a plurality of centrum end lines corresponding to each centrum based on the plurality of centrum key point information, and fitting the centrum center point of each centrum by using a preset least square method to obtain a fitting straight line segment;
calculating the distance between the center point of each cone and the fitted straight line segment to obtain the center distance value of each cone, obtaining the cone length of each cone, and dividing the center distance value of each cone by the corresponding cone length to obtain a plurality of distance ratios;
sequencing the distance ratios in a descending order to obtain distance ratio sequencing results, and determining the centrum center point corresponding to the distance ratio ranked as the first in the distance ratio sequencing results as a bending center;
and traversing and calculating included angles formed by a plurality of cone end lines around the bending center to obtain a plurality of bending included angles, sequencing the bending included angles according to a sequence from large to small to obtain a bending included angle sequencing result, and determining the bending included angle with the first rank in the bending included angle sequencing result as a spinal column side bending angle measurement result.
6. The method for measuring scoliosis angle according to claim 1, wherein the obtaining of the image to be measured, the pixel value extraction and normalization of the image to be measured, and the obtaining of the preprocessed image comprise:
acquiring an image to be detected, and reading a pixel value of the image to be detected to obtain an image pixel value;
and carrying out normalization processing on the image pixel values to obtain an image array after the normalization processing, and storing the image array after the normalization processing as a gray-scale image to obtain a preprocessed image.
7. The method for measuring the lateral curvature angle of the spine according to any one of claims 1-6, wherein before the obtaining the image to be measured, and performing pixel value extraction and normalization processing on the image to be measured to obtain a pre-processed image, the method further comprises:
obtaining model training data, preprocessing the model training data to obtain preprocessed training data, wherein the model training data is used for indicating a spine image set;
and training a preset initial key point detection model according to the preprocessed training data to obtain a target key point detection model.
8. A device for measuring a scoliosis angle, characterized in that it comprises:
the acquisition module is used for acquiring an image to be detected, and performing pixel value extraction and normalization processing on the image to be detected to obtain a preprocessed image;
the positioning module is used for calling a plurality of preset deep learning models, and sequentially carrying out spine positioning, spine vertebral body positioning and 12 th thoracic vertebra positioning on the preprocessed image to obtain a spine positioning result;
the analysis module is used for calling a preset target key point detection model and analyzing the spine positioning result to obtain a plurality of vertebral body key point information, wherein the plurality of vertebral body key point information comprises position information of four vertexes corresponding to each section of spine vertebral body;
and the measuring and calculating module is used for calling a preset least square method to perform fitting processing on the information of the plurality of centrum key points to obtain a fitting straight line section, determining a bending center based on the fitting straight line section, and measuring and calculating the scoliosis angle according to the bending center to obtain a scoliosis angle measuring result.
9. A measurement apparatus of a scoliosis angle characterized by comprising:
a memory and at least one processor, the memory having stored therein a computer program;
the at least one processor invokes the computer program in the memory to cause the scoliosis angle measuring apparatus to perform the scoliosis angle measuring method according to any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out a method of measuring a scoliosis angle according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110985135.8A CN113674257B (en) | 2021-08-26 | 2021-08-26 | Measuring method, device, equipment and storage medium for scoliosis angle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110985135.8A CN113674257B (en) | 2021-08-26 | 2021-08-26 | Measuring method, device, equipment and storage medium for scoliosis angle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113674257A true CN113674257A (en) | 2021-11-19 |
| CN113674257B CN113674257B (en) | 2024-06-04 |
Family
ID=78546430
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110985135.8A Active CN113674257B (en) | 2021-08-26 | 2021-08-26 | Measuring method, device, equipment and storage medium for scoliosis angle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113674257B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114078120A (en) * | 2021-11-22 | 2022-02-22 | 北京欧应信息技术有限公司 | Method, apparatus and medium for detecting scoliosis |
| CN114372970A (en) * | 2022-01-04 | 2022-04-19 | 杭州三坛医疗科技有限公司 | Operation reference information generation method and device |
| CN114494271A (en) * | 2022-02-18 | 2022-05-13 | 芙索特(上海)医疗科技有限公司 | Method for calculating ATR angle of human back |
| CN116869481A (en) * | 2023-07-12 | 2023-10-13 | 北京鹰之眼智能健康科技有限公司 | Spine overall structure state detection method based on infrared image and electronic equipment |
| CN116912245A (en) * | 2023-09-12 | 2023-10-20 | 吉林大学 | Scoliosis measuring method, related device and computer storage medium |
| CN117257455A (en) * | 2023-11-21 | 2023-12-22 | 中国人民解放军总医院第一医学中心 | Lumbar operation fixing rod pre-bending method and device |
| CN117765062A (en) * | 2024-02-22 | 2024-03-26 | 天津市天津医院 | Image processing method and system for detecting scoliosis of teenagers |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060120583A1 (en) * | 2004-11-10 | 2006-06-08 | Agfa-Gevaert | Method of performing measurements on digital images |
| CN110415291A (en) * | 2019-08-07 | 2019-11-05 | 清华大学 | Image processing method and relevant device |
-
2021
- 2021-08-26 CN CN202110985135.8A patent/CN113674257B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060120583A1 (en) * | 2004-11-10 | 2006-06-08 | Agfa-Gevaert | Method of performing measurements on digital images |
| CN110415291A (en) * | 2019-08-07 | 2019-11-05 | 清华大学 | Image processing method and relevant device |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114078120A (en) * | 2021-11-22 | 2022-02-22 | 北京欧应信息技术有限公司 | Method, apparatus and medium for detecting scoliosis |
| CN114372970A (en) * | 2022-01-04 | 2022-04-19 | 杭州三坛医疗科技有限公司 | Operation reference information generation method and device |
| CN114372970B (en) * | 2022-01-04 | 2024-02-06 | 杭州三坛医疗科技有限公司 | Surgical reference information generation method and device |
| CN114494271A (en) * | 2022-02-18 | 2022-05-13 | 芙索特(上海)医疗科技有限公司 | Method for calculating ATR angle of human back |
| CN116869481A (en) * | 2023-07-12 | 2023-10-13 | 北京鹰之眼智能健康科技有限公司 | Spine overall structure state detection method based on infrared image and electronic equipment |
| CN116869481B (en) * | 2023-07-12 | 2024-02-20 | 北京鹰之眼智能健康科技有限公司 | Spine overall structure state detection method based on infrared image and electronic equipment |
| CN116912245A (en) * | 2023-09-12 | 2023-10-20 | 吉林大学 | Scoliosis measuring method, related device and computer storage medium |
| CN117257455A (en) * | 2023-11-21 | 2023-12-22 | 中国人民解放军总医院第一医学中心 | Lumbar operation fixing rod pre-bending method and device |
| CN117257455B (en) * | 2023-11-21 | 2024-02-20 | 中国人民解放军总医院第一医学中心 | Lumbar operation fixing rod pre-bending method and device |
| CN117765062A (en) * | 2024-02-22 | 2024-03-26 | 天津市天津医院 | Image processing method and system for detecting scoliosis of teenagers |
| CN117765062B (en) * | 2024-02-22 | 2024-04-26 | 天津市天津医院 | Image processing method and system for detecting scoliosis of teenagers |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113674257B (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113674257B (en) | Measuring method, device, equipment and storage medium for scoliosis angle | |
| CN112734757B (en) | Spine X-ray image cobb angle measuring method | |
| Thong et al. | Three-dimensional morphology study of surgical adolescent idiopathic scoliosis patient from encoded geometric models | |
| CN111417980A (en) | Three-dimensional medical image analysis method and system for identification of vertebral fractures | |
| CN112184617B (en) | Spine MRI image key point detection method based on deep learning | |
| Korez et al. | A deep learning tool for fully automated measurements of sagittal spinopelvic balance from X-ray images: performance evaluation | |
| CN111047572A (en) | Automatic spine positioning method in medical image based on Mask RCNN | |
| US20210174503A1 (en) | Method, system and storage medium with a program for the automatic analysis of medical image data | |
| CN109124662B (en) | Rib center line detection device and method | |
| US9241634B2 (en) | Analytic morphomics: high speed medical image automated analysis method | |
| CHOI et al. | CNN-based spine and Cobb angle estimator using moire images | |
| CN115187606B (en) | Juvenile idiopathic scoliosis PUMC typing method | |
| WO2010095508A1 (en) | Midline determining device and program | |
| CN113284090A (en) | Scoliosis detection method and medical platform | |
| Alsiddiky et al. | Vertebral tumor detection and segmentation using analytical transform assisted statistical characteristic decomposition model | |
| EP3858245A1 (en) | Automatic recognition method for measurement point in cephalo image | |
| CN116645391A (en) | Student spine health management system and method | |
| US20230169644A1 (en) | Computer vision system and method for assessing orthopedic spine condition | |
| Ebrahimi et al. | Lumbar spine posterior corner detection in X-rays using Haar-based features | |
| Gabriel et al. | Development and clinical application of Vertebral Metrics: Using a stereo vision system to assess the spine | |
| CN114628010A (en) | Intelligent medical film reading method, device, equipment and storage medium | |
| KR101455198B1 (en) | Method, system and recordable medium for calculating quantitative assessment index of chest-wall deformities using ct image analysis | |
| Venkatapathiah et al. | Deep learning based object detection in nailfold capillary images | |
| EP2890299A1 (en) | Analytic morphomics: high speed medical image automated analysis method | |
| KR102598233B1 (en) | System and method for providing medical information based on deep learning targeting medical images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |