1. A method for reading a vertebral body photograph is characterized by comprising the following steps:

s1: acquiring preset quantity of shooting image information, uploading the acquired shooting image information to a server and establishing a shooting database;

s2: matching the photo to be read with a photo database to obtain preset vertebral body classification information, and identifying a preset position of a vertebral body according to the preset vertebral body classification information;

s3: positioning and analyzing the vertebral body at the preset position through preset positioning, and obtaining the compression rate corresponding to the vertebral body at the preset position;

s4: and obtaining a corresponding preset fracture grade through a preset grading interval in which the compression rate falls, and generating a diagnosis report.

2. The method for centrum photograph reading according to claim 1, wherein the step S3 includes:

s31: and positioning the upper and lower positioning points of the front edge of the vertebral body, the rear edge of the vertebral body and the middle of the vertebral body at the preset position.

3. The method for centrum photograph reading according to claim 2, wherein the step S3 includes:

s32: and (4) obtaining the maximum value and the minimum value of the height difference values of the upper and lower positioning points of the three groups of the front edge of the vertebral body, the rear edge of the vertebral body and the middle of the vertebral body by comparing the positioning points in the step S31.

4. The method for centrum photograph reading according to claim 3, wherein the step S3 further comprises:

s33: and obtaining the compression rate corresponding to the vertebral body at the preset position through a preset algorithm.

5. The method for reading vertebral body radiography according to claim 4, wherein the preset algorithm is as follows: (the maximum value of the altitude difference values of the three groups of upper and lower positioning points-the minimum value of the altitude difference values of the three groups of upper and lower positioning points)/the maximum value of the altitude difference values of the three groups of upper and lower positioning points.

6. The method for centrum photograph reading according to claim 1, wherein the step S4 includes:

s41: and judging whether the compression rate falls into a first preset classification interval, a second preset classification interval and a third preset classification interval.

7. The method for centrum photograph reading according to claim 6, wherein the step S41 comprises:

s411: and when the compression rate falls into a first preset grading interval, diagnosing the vertebral body at the preset position according to the first-degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

8. The method for centrum photograph reading according to claim 6, wherein the step S41 comprises:

s412: and when the compression rate falls into a second preset grading interval, diagnosing the vertebral body at the preset position according to the second degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

9. The method for centrum photograph reading according to claim 6, wherein the step S41 comprises:

s413: and when the compression rate falls into a third preset grading interval, diagnosing the vertebral body at the preset position according to the three-degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

10. The method for centrum photograph reading according to claim 4, characterized in that the step S4 is followed by the further steps of:

s5: and uploading the vertebral body diagnosis report with the fracture grade more than or equal to one degree to a server for storage, and summarizing to form a clinical vertebral body fracture diagnosis report.

Background

The reading of the vertebral fracture is heavy and heavy because the reading of the vertebral fracture generally adopts a lateral slice of thoracic X-ray and a lateral slice of lumbar X-ray, the reading contents at least comprise 13 vertebral bodies which are most prone to generating osteoporotic spine compression fracture, namely a thoracic T4-12 vertebral body and a lumbar L1-4 vertebral body, and the condition of nearly 20 vertebral bodies needs to be interpreted if all thoracic and lumbar vertebrae need to be considered. Although the morphology of the vertebral bodies of the thoracic and lumbar vertebrae is slightly different, the general epidemiological investigation of osteoporosis at least comprises 2000 and 3000 radiographs.

The requirements on the working strength and consistency of the film reader are several, visual fatigue and other nonresistible factors are easily caused, and the final film reading result is deviated.

Meanwhile, the methods for reading the fracture of the vertebral body are numerous, and no completely consistent accepted method for reading the fracture of the vertebral body exists in clinic or scientific research.

The ABQ interpretation method and the semi-quantitative interpretation method are the existing vertebral fracture interpretation methods which are used more and accepted more, but for scientific researches, particularly epidemiological investigation of large sample amount, the two interpretation methods are relatively labor-consuming and time-consuming, and the technical level and consistency requirements of long-time interpretation on the interpretation person are high.

In the case of large operating pressures, deflections are likely to occur. The clinical commonly used qualitative interpretation method has high interpretation speed and low working pressure, but is hardly used in scientific research because the method cannot quantitatively analyze data such as serious classification of vertebral body fracture.

Disclosure of Invention

In view of the current situation of the prior art, the technical problem to be solved by the present invention is to provide a method for reading vertebral body fracture, so as to solve the problems that the current methods for reading vertebral body fracture are not uniform, the task of reading a large number of fracture is heavy, and the reading result may be deviated due to multiple persons reading the fracture.

The invention relates to a method for reading a vertebral body photograph, which specifically comprises the following technical scheme:

a method for reading a vertebral body photograph comprises the following steps:

s1: acquiring preset quantity of shooting image information, uploading the acquired shooting image information to a server and establishing a shooting database;

s2: matching the photo to be read with a photo database to obtain preset vertebral body classification information, and identifying a preset position of a vertebral body according to the preset vertebral body classification information;

s3: positioning and analyzing the vertebral body at the preset position through preset positioning, and obtaining the compression rate corresponding to the vertebral body at the preset position;

s4: and obtaining a corresponding preset fracture grade through a preset grading interval in which the compression rate falls, and generating a diagnosis report.

Further, the step S3 includes:

s31: and positioning the upper and lower positioning points of the front edge of the vertebral body, the rear edge of the vertebral body and the middle of the vertebral body at the preset position.

Further, the step S3 includes:

s32: and (4) obtaining the maximum value and the minimum value of the height difference values of the upper and lower positioning points of the three groups of the front edge of the vertebral body, the rear edge of the vertebral body and the middle of the vertebral body by comparing the positioning points in the step S31.

Further, the step S3 further includes:

s33: and obtaining the compression rate corresponding to the vertebral body at the preset position through a preset algorithm.

Further, the preset algorithm is as follows: (the maximum value of the altitude difference values of the three groups of upper and lower positioning points-the minimum value of the altitude difference values of the three groups of upper and lower positioning points)/the maximum value of the altitude difference values of the three groups of upper and lower positioning points.

Further, the step S4 includes: s41: and judging whether the compression rate falls into a first preset classification interval, a second preset classification interval and a third preset classification interval.

Further, the step S41 includes:

s411: and when the compression rate falls into a first preset grading interval, diagnosing the vertebral body at the preset position according to the first-degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

Further, the step S41 includes:

s412: and when the compression rate falls into a second preset grading interval, diagnosing the vertebral body at the preset position according to the second degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

Further, the step S41 includes:

s413: and when the compression rate falls into a third preset grading interval, diagnosing the vertebral body at the preset position according to the three-degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

Further, the step S4 is followed by the step of:

s5: and uploading the vertebral body diagnosis report with the fracture grade more than or equal to one degree to a server for storage, and summarizing to form a clinical vertebral body fracture diagnosis report.

The technical scheme adopted by the invention has the following beneficial effects:

the method for reading the vertebral body shooting provided by the invention solves the problems that the method for reading the vertebral body fracture in the current clinical or scientific research work is not uniform, and the reading result is different due to different reading personnel or adopted methods when the same shooting work or even the same shooting work is performed, so that the effectiveness of clinical diagnosis and scientific research data is influenced.

When a large number of film reading tasks are heavy, the film reading personnel are seriously burdened during manual film reading, and the deviation of the film reading result can be caused under the high-intensity working condition.

Drawings

Fig. 1 is a first flowchart of a method for reading a vertebral body radiography according to an embodiment of the present invention;

FIG. 2 is a second flowchart of a method for reading a vertebral body radiograph according to an embodiment of the present invention;

fig. 3 is a third flowchart of a method for reading a vertebral body photograph according to an embodiment of the present invention;

fig. 4 is a fourth flowchart of a method for reading a vertebral body radiography according to the embodiment of the present invention.

Detailed Description

The following are specific embodiments of the present invention and are further described with reference to the drawings, but the present invention is not limited to these embodiments.

This example

The embodiment provides a method for reading a vertebral body photograph, as shown in fig. 1, the method comprises the following steps:

s1: acquiring preset quantity of shooting image information, uploading the acquired shooting image information to a server and establishing a shooting database;

s2: matching the photo to be read with a photo database to obtain preset vertebral body classification information, and identifying a preset position of a vertebral body according to the preset vertebral body classification information;

s3: positioning and analyzing the vertebral body at the preset position through preset positioning, and obtaining the compression rate corresponding to the vertebral body at the preset position;

s4: and obtaining a corresponding preset fracture grade through a preset grading interval in which the compression rate falls, and generating a diagnosis report.

Specifically, in this embodiment, 14075 photographs in the DICOM format obtained by investigation are uploaded to the server before the reading of the photographs, and a total photograph database is established. And importing the photos to be read into a global photo database and matching preset vertebral body classification information corresponding to the photos, wherein the preset vertebral body classification information comprises classification information of thoracic vertebrae and lumbar vertebrae.

The preset position of the vertebral body in the picture to be read can be identified according to the classification information, and the preset position is used for positioning the position of each vertebral body, namely the thoracic vertebra T4-12 and the lumbar vertebra L1-4 according to the form of the whole X-ray picture.

Starting single reading work, automatically identifying a plurality of positioning points of the appearance of the shape of the vertebral body, obtaining the compression ratio corresponding to the vertebral body at the preset position through the plurality of positioning points, obtaining the corresponding preset fracture grade through the preset grading interval in which the compression ratio falls, and obtaining the percentage value according to the compression ratio after the compression ratio of the vertebral body is obtained. The diagnosis is carried out by grading the fracture grades of 0, 1, 2 and 3 degrees of the vertebral body fracture by referring to a semi-quantitative diagnosis method, and meanwhile, a diagnosis report containing corresponding diagnosis information is generated.

Wherein, step S3 includes:

s31: and positioning the upper and lower positioning points of the front edge of the vertebral body, the rear edge of the vertebral body and the middle of the vertebral body at the preset position.

Wherein, step S3 includes:

s32: and (4) obtaining the maximum value and the minimum value of the height difference values of the upper and lower positioning points of the three groups of the front edge of the vertebral body, the rear edge of the vertebral body and the middle of the vertebral body by comparing the positioning points in the step S31.

Wherein, step S3 further includes:

s33: and obtaining the compression rate corresponding to the vertebral body at the preset position through a preset algorithm.

Wherein, the preset algorithm is as follows:

(the maximum value of the altitude difference values of the three groups of upper and lower positioning points-the minimum value of the altitude difference values of the three groups of upper and lower positioning points)/the maximum value of the altitude difference values of the three groups of upper and lower positioning points.

Specifically, in this embodiment, 6 positioning points located on the appearance of the cone shape are preset and positioned, the boundary of the cone can be divided by using the density difference of the cone, the intervertebral disc, the soft tissue and the like on the X-ray film, the boundary is a quasi-rectangular boundary, and the 6 positioning points are 4 vertexes of a rectangle and 2 middle points of the upper edge and the lower edge, and are the highest point and the lowest point of the front edge of the cone, the highest point and the lowest point of the rear edge of the cone and the highest point and the lowest point of the middle of the cone.

After the positioning points are automatically confirmed, the film reader can still perform manual fine adjustment, after the positioning is completed, the film reader can manually adjust 6 positioning points again according to own experience or visual film reading suggestions, so that the precise vertebral body shape can be more precisely positioned, the manual adjustment process can be omitted, the time can be saved, and the film reading efficiency and the consistency of the whole film reading can be improved.

After the positioning step is completed, automatically calculating the height value among each group of positioning points, namely the height value of the front edge of the vertebral body, the height value of the rear edge of the vertebral body and the height value of the middle of the vertebral body, and then calculating the maximum value and the minimum value in the 3 groups of data, wherein the algorithm is as follows: (max-min)/max; the value is the compression ratio of the vertebral body.

Meanwhile, if the minimum value is the height of the front edge of the cone, the minimum value is the height of the rear edge of the cone, the minimum value is compression (crush), and the minimum value is the height of the middle of the cone, the minimum value is biconcave.

Wherein, step S4 includes:

s41: and judging whether the compression rate falls into a first preset classification interval, a second preset classification interval and a third preset classification interval.

Referring to fig. 2, therein;

step S41 includes: s411: and when the compression rate falls into a first preset grading interval, diagnosing the vertebral body at the preset position according to the first-degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

Referring to fig. 3, therein;

step S41 includes: s412: and when the compression rate falls into a second preset grading interval, diagnosing the vertebral body at the preset position according to the second degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

Referring to fig. 4, therein;

step S41 includes: s413: and when the compression rate falls into a third preset grading interval, diagnosing the vertebral body at the preset position according to the three-degree fracture grade by a semi-quantitative diagnosis method, and generating a diagnosis report.

Specifically, in this embodiment, according to the compression ratio value falling in the interval classified into three levels, i.e., 0.2-0.25, 0.25-0.4, and 0.4-1.0, the division of vertebral fracture is diagnosed by referring to the semi-quantitative diagnosis method to be the first, second, and third degree fracture levels, and the zero degree fracture level is determined when the compression ratio value is lower than 0.2; whereby the fracture grade is determined by a preset grade interval in which the compression rate falls.

Wherein, step S4 is followed by the step of: s5: and uploading the vertebral body diagnosis report with the fracture grade more than or equal to one degree to a server for storage, and summarizing to form a clinical vertebral body fracture diagnosis report.

Specifically, in this embodiment, after the hierarchical diagnosis is completed, the vertebral bodies with the graduation equal to or greater than the first degree fracture level are listed individually by screening and summarizing according to the hierarchical situation, so as to form a clinically usable vertebral body fracture diagnosis report.

Meanwhile, the background program can automatically derive a summary table containing all information such as the individual condition of a patient, the fracture vertebral body, the graduation of the fracture vertebral body and the like for the purpose of subsequent information retrieval or scientific research; and finally, the whole film reading work for establishing the film shooting library is completed, and the standard clinical diagnosis report or the subsequent data for scientific research and analysis can be exported or backed up at any time.

Specifically, genetic or metabolic bone disease osteoporosis; any disease that is subject to severe trauma or vertebral body tumors that can lead to spinal surgery and treatments that include bone cement and other conditions that can affect the morphology of the spine. We will inform the subjects willing to participate about the risk of thoracolumbar lateral X-ray examination while signing informed consent and taking a thoracolumbar lateral X-ray photograph.

All X-ray films are stored in a DICOM or JPG format through an electronic edition, the resolution of the X-ray films is at least 1280X 1024, and the collected medical histories are collected to a database in a unified mode. AI aids based on the teaching of Genant et al to establish semi-quantitative diagnostics are used to diagnose vertebral fractures and the graduation of fractures.

And (3) reading the film: and (3) importing the screened 14075 photos in DICOM format into film reading software by using an automatic importing function of the software, and establishing a total photo database for each time of film reading or research.

Software is utilized to automatically identify 6 positioning points of the appearance of the cone shape, the highest lowest point of the front edge of the cone, the highest lowest point of the rear edge of the cone and the highest lowest point of the middle of the cone.

After the positioning is finished, the highest and lowest height values of each group are calculated, and the compression rate of the vertebral body is determined through the height values.

After the compression ratio of the vertebral body is obtained, the graded diagnosis of the fracture is carried out by grading the fractures of 0, 1, 2 and 3 degrees of the vertebral body fracture by referring to a semi-quantitative diagnosis method according to the percentage value of the compression ratio, and after the graded diagnosis is finished, the system can list out the vertebral bodies with the graduation of more than or equal to 1 degree independently according to the grading condition to form a clinically usable diagnosis report of the vertebral body fracture.

Meanwhile, a background program of the system can automatically derive a summary table containing all information such as the individual condition of a patient, the fracture vertebral body, the graduation of the fracture vertebral body and the like, and the summary table is used for subsequent information retrieval or scientific research.

And finally, the whole film reading work for establishing the film shooting library is completed, and the standard clinical diagnosis report or the subsequent data for scientific research and analysis can be exported or backed up at any time.

Specifically, the diagnosis principle of a semi-quantitative interpretation method is utilized, and automatic quantification and calculation are realized by utilizing a computer AI technology, so that a large amount of manpower and material resources can be saved, and the highest consistency result can be achieved.

Before film reading, all DICOM format films can be uniformly built into a total film taking library, the total film taking library is led into a system to build a single film reading library by using a software lead-in function, and different film reading libraries can be built to independently analyze different film reading in each time according to different patient groups or scientific research subject groups in each time.

After the introduction is finished, the single picture reading work can be started, after the picture reading is started, software firstly positions each vertebral body, namely the positions of thoracic vertebra T4-12 and lumbar vertebra L1-4 according to the different selected introduction parts, such as thoracic vertebra lateral position or lumbar vertebra lateral position X-ray radiography, and then automatically identifies 6 positioning points of the appearance of the vertebral body shape.

The boundary of the vertebral body can be divided by utilizing the different densities of the vertebral body, the intervertebral disc, the soft tissue and the like on the X-ray film, the boundary is a quasi-rectangular boundary, 6 positioning points are 4 vertexes of the rectangle and 2 middle points of the upper edge and the lower edge, and are the highest and lowest points of the front edge of the vertebral body, the highest and lowest points of the rear edge of the vertebral body and the middle highest and lowest points of the vertebral body at the same time; the positioning point can still be manually fine-tuned after being automatically confirmed.

After the positioning of the system is completed, the reader or the software user can manually adjust 6 positioning points again according to own experience or visual reading opinions so as to more accurately position the precise vertebral body shape, but the manual adjustment process can be omitted, so that the time can be saved, the reading efficiency can be improved, and the consistency of the whole reading can be improved.

After the positioning step is completed, the system can automatically calculate the height value between each group of positioning points, namely the height value of the front edge of the vertebral body, the rear height of the vertebral body and the height value of the middle position of the vertebral body, and then calculate the following 3 groups of data: (max-min)/max, which is the compression ratio of the vertebral body; meanwhile, if the minimum value is the height of the front edge of the cone, the minimum value is the height of the rear edge of the cone, the minimum value is compression (crush), and the minimum value is the height of the middle of the cone, the minimum value is biconcave.

The compression ratio value is classified into three levels of vertebral body height compression values of 0.2-025, 0.25-0.4 and 0.4-1.0, and the graduation of vertebral body fracture is 1, 2 and 3 degrees fracture corresponding to a semi-quantitative diagnosis method, so that the final diagnosis of positioning, qualitative and quantitative fracture can be completed.

After the grading diagnosis of the overall shooting library is completely finished, the system can automatically count the relevant shooting results as follows: the overall fracture of more than 1 degree, namely the detection rate of vertebral body fracture, the fracture graduation of each vertebral body, the fracture condition of each vertebral body under each graduation and the like, so that the reference and use can be conveniently carried out during clinical or subsequent scientific research.

All the documents in the reading library are read after the reading work is finished, and both standard clinical diagnosis reports and subsequent data for scientific research and analysis can be exported or backed up at any time, and paper-version reports can be formed, so that the working time and pressure of reading are greatly reduced, manpower and material resources are greatly reduced, and the consistency of the reading result is improved to the maximum extent.

The method solves the problems that the existing centrum fracture reading methods in clinical or scientific research work are not uniform, and the reading results are different due to different reading personnel or adopted methods when the same reading work or even the same shooting reading work is carried out, so that the effectiveness of clinical diagnosis and scientific research data is influenced.

The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.