Rib fracture rapid diagnosis system based on artificial intelligence technology

Abstract

The invention provides a rapid rib fracture diagnosis device based on an artificial intelligence technology. The rapid rib fracture diagnosis device comprises a spiral CT examination equipment main body, an intelligent host and a display device, the intelligent host comprises a fracture image sample acquisition module, a three-dimensional convolution analysis and learning module and a fracture image analysis and diagnosis module; the three-dimensional convolution analysis and learning module is used for carrying out feature analysis and learning memory on the obtained diagnosed fracture image samples to form a fracture feature image library; the fracture image analysis and diagnosis module is used for acquiring a multi-layer spiral CT examination image of the clinical rib fracture image output by the spiral CT examination equipment main body and carrying out artificial intelligence clinical diagnosis. According to the method, the fracture feature database is formed through artificial intelligence on the basis of the condition that the image features obtained through multi-layer spiral CT examination are finer and the number of the images is large, a doctor is helped to carry out clinical rib fracture diagnosis on the basis, the diagnosis speed and accuracy can be remarkably improved, and the diagnosis efficiency is improved. And the labor intensity of workers and misdiagnosis or missed diagnosis caused by human factors are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of orthopedic medical diagnosis, and in particular to a rapid diagnosis system for rib fractures based on artificial intelligence technology. Background Art

[0002] Rib fractures are common chest injuries in clinical practice, often caused by external forces such as collisions, falls, falls from heights, and car accidents. Rib fractures are often accompanied by lung damage. If not treated in time, it will affect prognosis or even life. Timely and accurate diagnosis is of great value to the treatment and prognosis of patients. Traditional examination and diagnosis of rib fractures are mainly carried out through X-ray examinations. Since X-ray examinations are overlapping images and are greatly related to the projection angle, the diagnostic efficiency is low and the missed diagnosis rate is relatively high. Therefore, when conditions permit, most rib fractures currently use multi-slice spiral CT examinations as the most important imaging examination method for chest trauma. Due to the increasingly thin scanning layer thickness during multi-slice spiral CT examinations, and the large increase in images, the workload of diagnostic physicians has increased dramatically. After working for a certain period of time, it will cause visual fatigue, which is easy to cause missed diagnosis or misdiagnosis; in addition, due to the different diagnostic levels of physicians, there is also a certain probability of missed diagnosis and misdiagnosis of rib fractures caused by human factors in actual clinical practice.

[0003] Using three-dimensional convolution for image analysis is a means in the field of artificial intelligence that takes into account both efficiency and results. The relevant algorithms and image processing chip technologies are relatively mature, and the current clinical judgment of fractures is also based on fracture images. The use of three-dimensional convolution and machine learning technology for artificial intelligence judgment of rib fractures should be a development trend in clinical treatment. It can improve the rapid diagnosis of clinical fractures, reduce the intensity of manual labor, and significantly reduce missed diagnoses or misdiagnoses caused by human factors. Summary of the invention

[0004] In view of the problem that the amount of image data in existing clinical examinations of rib fractures is large, which is easy to cause misdiagnosis or missed diagnosis due to film reading fatigue or insufficient examination experience, the present invention provides a rib fracture rapid diagnosis device based on artificial intelligence technology to achieve efficient on-site rib fracture diagnosis.

[0005] Specifically, the present invention is implemented by the following technical solutions: A rapid diagnosis device for rib fractures based on artificial intelligence technology, the rapid diagnosis device comprising a spiral CT inspection device body, an intelligent host and a display device; the intelligent host comprises a fracture image sample acquisition module, a three-dimensional convolution analysis and learning module and a fracture image analysis and diagnosis module; the fracture image sample acquisition module comprises one or more of a network connection module, a barcode recognition module, an image photography module and an image scanning module to acquire confirmed fracture images; the three-dimensional convolution analysis and learning module performs feature analysis and learning memory on the acquired confirmed fracture image samples to form a fracture feature image library; the fracture image analysis and diagnosis module acquires a multi-slice spiral CT inspection image of a clinical rib fracture image output by the spiral CT inspection device body, and performs a clinical diagnosis of the rib fracture based on the fracture image feature library formed by learning, and outputs the diagnosis result to the display device.

[0006] Preferably, the working process of the rapid diagnosis device for rib fracture is: S1, firstly obtain a set number of confirmed fracture sample images through the fracture image sample acquisition module; S2. After obtaining the confirmed fracture image sample, the three-dimensional convolution analysis and learning module obtains the three-dimensional features of the fracture characteristics by analyzing the image, and forms a fracture feature image database through machine learning; S3, performing a diagnostic test based on the formed fracture feature image database, and manually checking the diagnostic results. If the diagnostic success rate reaches or exceeds the target value, the diagnostic system can be applied to clinical diagnosis; if the diagnostic success rate is relatively low, the incorrectly diagnosed fracture images are used as new samples and further analyzed and machine-learned by the three-dimensional convolution analysis and learning module; S4. After the diagnostic success rate of the rib fracture rapid diagnosis device reaches the target value, the device can be used for clinical diagnosis. The clinical fracture image is obtained through the spiral CT examination equipment body. After the intelligent host obtains the fracture image, the clinical image is judged based on the fracture feature image database formed by machine learning to determine whether a fracture is formed, the number and location of the fracture, etc., and the machine diagnosis result is displayed on the display device.

[0007] More preferably, after the above-mentioned step S4, step S5 is further included, in which manual re-examination and screening are performed. If any misdiagnosed or missed cases are found, the fracture images of the misdiagnosed or missed cases are input into the system again through the fracture image sample acquisition module for re-analysis and learning.

[0008] More preferably, in the above step S1, the fracture sample image is acquired by retrieving existing fracture image data in a database through a network or a storage medium.

[0009] More preferably, in the above step S1, the fracture sample image is acquired by reading the two-dimensional code information containing the fracture image through a barcode recognition module.

[0010] More preferably, in the above step S1, the fracture sample image is obtained by photographing the fracture sample.

[0011] More preferably, in the above step S1, the fracture sample image is acquired by an image scanning module.

[0012] More preferably, in the above step S1, the fracture sample image is a two-dimensional image or a three-dimensional image or a reconstructed image formed by multi-slice spiral CT.

[0013] The rib fracture rapid diagnosis device based on artificial intelligence technology of the present invention is based on the existing multi-slice spiral CT examination for clinical fracture diagnosis. The image features obtained by the device are more detailed and the images are more numerous. The fracture feature database is formed through artificial intelligence analysis and learning, and clinical rib fracture diagnosis is performed based on the database. The speed and accuracy of diagnosis can be significantly improved, the workload of doctors' film reading can be greatly reduced, and the labor intensity of manual labor and misdiagnosis or missed diagnosis caused by human factors can be reduced. The combination of artificial intelligence and manual film reading by doctors can significantly improve the efficiency of clinical diagnosis, avoid misdiagnosis or missed diagnosis, and provide a reliable basis for subsequent treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The above and other objects, features and advantages of the present invention will become apparent from the following detailed description in conjunction with the accompanying drawings, in which: Figure 1 The figure shows the structure of the rapid diagnosis device for rib fracture based on artificial intelligence technology of the present invention; Figure 2 Shown Figure 1 Schematic diagram of the system architecture of the rapid diagnosis device for rib fractures based on artificial intelligence technology; Figure 3 Shown Figure 1 Flowchart of the working principle of the rapid diagnosis device for rib fractures based on artificial intelligence technology. DETAILED DESCRIPTION

[0015] With reference to the accompanying drawings, the system architecture features and working principles of the present invention are described in detail as follows.

[0016] See the schematic diagram of the structure of the rib fracture rapid diagnosis device based on artificial intelligence technology shown in FIG1. Figure 2The system architecture diagram of the device shown in the figure, the rapid diagnosis device includes a spiral CT inspection device body 10, an intelligent host 20 and a display device 30; the intelligent host 20 includes a fracture image sample acquisition module, a three-dimensional convolution analysis and learning module and a fracture image analysis and diagnosis module; the fracture image sample acquisition module includes one or more of a network connection module, a barcode recognition module 202, an image taking module 203 and an image scanning module 201 to obtain a confirmed fracture image; the three-dimensional convolution analysis and learning module performs feature analysis and learning memory on the acquired confirmed fracture image samples to form a fracture feature image library; the fracture image analysis and diagnosis module obtains a multi-slice spiral CT inspection image of a clinical rib fracture image output by the spiral CT inspection device body 10, and performs a clinical diagnosis of the rib fracture based on the fracture image feature library formed by learning, and the diagnosis result is output to the display device.

[0017] Combined with reference Figure 3 The working principle flow chart of the rib fracture rapid diagnosis device based on artificial intelligence technology of the present invention is shown. In the rib fracture rapid diagnosis device, the fracture image sample acquisition module first acquires a confirmed fracture sample image, which can be obtained through a wired or wireless network or a storage medium (such as Figure 1 As shown in , the smart host 20 is provided with a storage medium socket 206) to retrieve the existing fracture image sample data in the database, or to obtain the sample image by reading the two-dimensional code containing the fracture image sample through the barcode recognition module, or to obtain the fracture sample image through the image taking module, or to obtain the fracture sample image through the image scanning module; when different fracture sample image inputs are used according to different situations, the corresponding image sample input can be started through the corresponding start button 204 on the smart host 20. In another preferred embodiment, a plurality of indicator lights are also provided on the smart host 20 to let the doctor know the current working status of the device. The above-mentioned methods of obtaining fracture sample images are several commonly used technical means. If there are other methods to obtain fracture sample images, they should be regarded as technical means equivalent to the present application. In addition, the fracture sample image is a two-dimensional image or a three-dimensional image or a reconstructed image formed by multi-slice spiral CT.

[0018] After obtaining the confirmed fracture image samples, the three-dimensional convolution analysis and learning module analyzes the image samples to obtain the three-dimensional features of the fracture features, and forms a fracture feature image database through machine learning; in theory, the more fracture samples analyzed and learned, the more complete the fracture feature image database formed, and the more accurate the diagnosis based on the image database in clinical practice will be, but in order to improve the use efficiency of the device, the device cannot be endlessly analyzed and learned. In practice, a value can be set for the number of fracture image samples analyzed and learned by the system, such as 200 cases, and then a diagnostic test is performed based on the formed fracture feature image database, and the diagnostic results are manually checked. If the diagnostic success rate reaches more than 95%, the diagnostic system can be applied to the clinic; if the diagnostic success rate is relatively low, the misdiagnosed fracture image can be used as a new sample by the three-dimensional convolution analysis and learning module to continue to analyze and machine learn, and this is repeated until the diagnostic success rate reaches the target value or above; in this way, the use efficiency of the device can be taken into account and the diagnostic success rate of the device can be continuously improved. For the image processing method and chip based on the convolution algorithm, there are currently mature solutions available on the market, and the present invention will not be described in detail. Commercially available products can be directly applied to this system.

[0019] After the system's diagnostic success rate reaches the target value, the rib fracture rapid diagnosis device based on artificial intelligence technology can be used for clinical diagnosis. The clinical fracture image is acquired through the spiral CT examination equipment body 10. After the fracture image is obtained, the clinical image is judged based on the fracture feature image database formed by machine learning to determine whether a fracture is formed and the number and location of the fracture. The machine diagnosis result is displayed on the display device to provide basic data for treatment.

[0020] In actual application, in order to ensure that there is no misdiagnosis or missed diagnosis, manual re-examination and screening can be performed after the machine diagnosis. If a misdiagnosis or missed diagnosis is found, the fracture image of the misdiagnosis or missed diagnosis case will be input into the system again through the fracture image sample acquisition module for re-analysis and learning. In this way, the problem of misdiagnosis or missed diagnosis can be avoided and the diagnosis success rate of the system can be guaranteed to be higher and higher. The misdiagnosed or missed diagnosis fracture images can be input into the system through network data transmission, image scanning, etc.

[0021] Since multi-slice spiral CT images can be scanned more quickly and assisted by post-processing techniques such as multi-planar reconstruction and volume rendering, images of any plane can be reconstructed when analyzing the rib morphology. The images are clearer and more numerous, allowing for more three-dimensional, multi-angle and detailed analysis. This makes it easier to obtain fracture features for three-dimensional convolution analysis and learning modules, which can improve the diagnosis rate for clinical diagnosis and avoid misdiagnosis or missed diagnosis due to visual fatigue and lack of diagnostic experience of physicians.

[0022] The rib fracture rapid diagnosis device based on artificial intelligence technology of the present invention is based on the existing multi-slice spiral CT examination for clinical fracture diagnosis. The image features obtained by the device are more detailed and the images are more numerous. The fracture feature database is formed through artificial intelligence analysis and learning, and clinical rib fracture diagnosis is performed based on the database. The speed and accuracy of diagnosis can be significantly improved, the workload of doctors' film reading can be greatly reduced, and the labor intensity of manual labor and misdiagnosis or missed diagnosis caused by human factors can be reduced. The combination of artificial intelligence and manual film reading by doctors can significantly improve the efficiency of clinical diagnosis, avoid misdiagnosis or missed diagnosis, and provide a reliable basis for subsequent treatment.

[0023] The present invention is not limited to the described embodiments. Those skilled in the art may make some modifications or changes without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the scope defined in the claims.

Claims

1. A rapid diagnosis device for rib fractures based on artificial intelligence technology, characterized in that: The rapid diagnosis device includes a spiral CT inspection equipment body, an intelligent host and a display device; the intelligent host includes a fracture image sample acquisition module, a three-dimensional convolution analysis and learning module and a fracture image analysis and diagnosis module; the fracture image sample acquisition module includes one or more of a network connection module, a barcode recognition module, an image photography module and an image scanning module to obtain confirmed fracture images; the three-dimensional convolution analysis and learning module performs feature analysis and learning memory on the acquired confirmed fracture image samples to form a fracture feature image library; the fracture image analysis and diagnosis module obtains the multi-slice spiral CT inspection image of the clinical rib fracture image output by the spiral CT inspection equipment body, and performs clinical diagnosis of the rib fracture based on the fracture image feature library formed by learning, and the diagnosis result is output to the display device.

2. A rapid diagnosis device for rib fractures based on artificial intelligence technology as claimed in claim 1, characterized in that: The working process of the rapid diagnosis device for rib fracture is as follows: S1, firstly obtain a set number of confirmed fracture sample images through the fracture image sample acquisition module; S2. After obtaining the confirmed fracture image sample, the three-dimensional convolution analysis and learning module obtains the three-dimensional features of the fracture characteristics by analyzing the image, and forms a fracture feature image database through machine learning; S3, performing a diagnostic test based on the formed fracture feature image database, and manually checking the diagnostic results. If the diagnostic success rate reaches or exceeds the target value, the diagnostic system can be applied to clinical diagnosis; if the diagnostic success rate is relatively low, the incorrectly diagnosed fracture images are used as new samples and further analyzed and machine-learned by the three-dimensional convolution analysis and learning module; S4. After the diagnostic success rate of the rib fracture rapid diagnosis device reaches the target value, the device can be used for clinical diagnosis. The clinical fracture image is obtained through the spiral CT examination equipment body. After the intelligent host obtains the fracture image, the clinical image is judged based on the fracture feature image database formed by machine learning to determine whether a fracture is formed, the number and location of the fracture, etc., and the machine diagnosis result is displayed on the display device.

3. A rapid diagnosis device for rib fractures based on artificial intelligence technology as claimed in claim 2, characterized in that: After the above step S4, step S5 is also included, in which manual re-examination and screening are performed. If a case of misdiagnosis or missed diagnosis is found, the fracture image of this case is input into the system again through the fracture image sample acquisition module for re-analysis and learning.

4. The rapid diagnosis device for rib fracture based on artificial intelligence technology as claimed in claim 2, characterized in that: In the above step S1, the fracture sample image is obtained by retrieving the existing fracture image data in the database through the network or storage medium.

5. The rapid diagnosis device for rib fracture based on artificial intelligence technology as claimed in claim 2, characterized in that: In the above step S1, the barcode recognition module reads the two-dimensional code information containing the fracture image to obtain the fracture sample image.

6. The rapid diagnosis device for rib fracture based on artificial intelligence technology as claimed in claim 2, characterized in that: In the above step S1, the fracture sample image is acquired by photographing the fracture sample.

7. The rapid diagnosis device for rib fracture based on artificial intelligence technology as claimed in claim 2, characterized in that: In the above step S1, the fracture sample image is acquired through the image scanning module.

8. The rapid diagnosis device for rib fracture based on artificial intelligence technology as claimed in claim 2, characterized in that: In the above step S1, the fracture sample image is a two-dimensional image or a three-dimensional image or a reconstructed image formed by multi-slice spiral CT.

Images