CN-121999991-A - Remote artificial intelligent detection method and system for bone age digital X-ray image

CN121999991ACN 121999991 ACN121999991 ACN 121999991ACN-121999991-A

Abstract

The invention relates to a remote artificial intelligent detection method and a system for bone age digital X-ray images, wherein the method comprises the following steps of remotely acquiring X-ray image data, transmitting the X-ray image data through a 5G private network in a desensitization encryption mode, carrying out normalization processing on the received X-ray image data, judging required GPU computing power resources according to the normalized X-ray image data, calling at least one required AI (advanced technology) computing server based on a load balancing principle, acquiring bone age detection results of the AI computing server, and generating a detection report, wherein the X-ray image data and the detection report are stored in an AI storage device, and the storage of the X-ray image data and the detection report conforms to IHE integration specifications. Compared with the prior art, the invention effectively solves the problems of lack of bone age detection resources, high cost, easy leakage of privacy and the like in basic level and remote areas, and improves the detection efficiency and service coverage.

Inventors

YANG XIUJUN
SUN ZHONGQIANG
LI TINGTING
LAI SHUANG
Dang peng

Assignees

上海市儿童医院

Dates

Publication Date: 20260508
Application Date: 20251203

Claims (10)

1. A remote artificial intelligence detection method of bone age digital X-ray image is characterized by comprising the following steps: Remotely acquiring X-ray image data, wherein the X-ray image data is transmitted and acquired by a 5G private network in a desensitization encryption mode; Normalizing the received X-ray image data; judging required GPU computing power resources according to the normalized X-ray image data, and calling at least one required AI computing server based on a load balancing principle; acquiring a bone age detection result of an AI calculation server, and generating a detection report; Wherein the X-ray image data and the detection report are stored in an AI storage device, and the storage of the X-ray image data and the detection report complies with IHE integration specifications.
2. The method for remote artificial intelligence detection of bone age digital X-ray images of claim 1, wherein the X-ray image data is obtained through an API interface, a message queue, web Service docking or WebSocket docking mode.
3. The method for remote artificial intelligence detection of bone age digital X-ray images of claim 1, wherein the desensitizing encryption comprises: And performing desensitization treatment on the original data acquired by the X-ray bone age equipment, reserving minimum data required by AI calculation, and encrypting to form the X-ray image data.
4. The method for remote artificial intelligence detection of bone age digital X-ray images according to claim 1, wherein the normalization of the X-ray image data is performed using a quantile number suitable for the bone age of the child.
5. The method for remote artificial intelligence detection of bone age digital X-ray images according to claim 1, wherein the GPU computing power resources required for the determination are specifically: and determining the required GPU computing power resources according to the data scale and the effectiveness target of the X-ray image data.
6. The method according to claim 1, wherein if a fault occurs during the remote acquisition of the X-ray image data, the X-ray image data is continuously acquired by breakpoint continuous transmission after the fault is recovered.
7. A remote artificial intelligence detection system for bone age digital X-ray images, comprising: The infrastructure layer comprises a first front-end processor arranged at a base layer end, a second front-end processor arranged at an AI detection end, an AI database server, an AI storage device and an AI calculation server, wherein the first front-end processor and the second front-end processor are in remote communication connection through a 5G private network to realize remote transmission of X-ray image data, the second front-end processor, the AI database server, the AI storage device and the AI calculation server are connected through communication lines, and a plurality of AI calculation servers are arranged to form a cluster; The data resource layer comprises a data specification processing module and a data desensitization encryption processing module, wherein the data specification processing module is used for performing specification standardization processing on X-ray image data, report data and/or record data, and the data desensitization encryption processing module is used for performing desensitization encryption processing on the data desensitization encryption processing module; The platform supporting layer is deployed in the bone age service platform application server and is used for realizing the operation of an AI model for bone age detection and the cluster management and the distributed processing of an AI calculation server, wherein the distributed processing specifically comprises the steps of judging required GPU computing power resources according to X-ray image data and calling at least one required AI calculation server based on a load balancing principle; the application service layer is deployed on a plurality of terminal clients and comprises an application interface facing a user.
8. The system of claim 7, wherein the first pre-processor is configured with an image NET receive upload service and a collaborative platform interface service.
9. The remote artificial intelligence detection system of the bone age digital X-ray image according to claim 7, wherein the AI storage device synchronizes the received X-ray image data to the image storage management node according to a preset synchronization mechanism, simultaneously starts an image integrity verification process, compares the acquired image number with the image number in the regional PACS system, generates a verification passing identifier if the comparison passes, generates a supplementary transmission request if the comparison does not pass, and feeds back to the first front-end processor.
10. The remote artificial intelligence detection system of bone age digital X-ray images of claim 7, wherein the application interfaces include a bone age detection service interface, a doctor's conference service interface, and a patient report image service interface.

Description

Remote artificial intelligent detection method and system for bone age digital X-ray image Technical Field The invention relates to the field of artificial intelligence, in particular to a remote artificial intelligence detection method and system for bone age digital X-ray images. Background In pediatric routine testing, skeletal maturity testing is an important tool for diagnosing endocrine disorders in children by comparing bone age with true age. The traditional bone age detection adopts a manual film reading mode, so that the process is complicated, the time consumption is long, the problems of strong interpretation subjectivity, poor result consistency, lack of unified standards and the like exist, and the large-scale screening requirement is difficult to adapt. Meanwhile, the contradiction of the uneven distribution of pediatric medical resources is prominent, high-quality resources are concentrated in a central city, the total amount of pediatricians is insufficient in basic level and remote areas, professional bone age interpretation doctors are deficient, and the development of local bone age detection service is seriously restricted due to long culture period. Although some medical institutions have developed artificial intelligence bone age detection schemes that reach high annual energy expert levels, medical institutions in both the base and remote areas still face a number of dilemmas: Firstly, the quality of the scheme on the market is uneven, and the scheme is limited by the shortage of expenses, the lack of professional talents and the like, so that the adaptive scheme is difficult to screen; secondly, the localization deployment and later maintenance cost of the AI system is high, which exceeds the bearing capacity of the common basic mechanism; thirdly, the cloud service mode has the risk of patient privacy disclosure, and does not accord with the medical information safety specification; fourthly, the locally deployed bone age AI detection box is limited by hardware integration, and has the characteristics of slow response, large result error, difficult updating and maintenance and the like, and cannot meet the actual detection requirement. The research and development of the single-center bone age AI automatic evaluation detection scheme are finished in the early stage of the applicant team, the bone age detection is carried out by an artificial intelligence method, artificial subjective factors are avoided, and the detection efficiency and accuracy are improved. However, the scheme does not adapt to the basic-level medical cost and resource constraint, has high floor and maintenance cost, is complex to operate, lacks a medical data security guarantee mechanism, cannot avoid privacy disclosure risks, and also does not solve the problems of localized deployment of performance short boards and maintenance, and has insufficient compatibility to basic-level equipment and adaptation to special cases. Therefore, it is necessary to develop an AI bone age detection scheme that can cover a wider application range without additional cost, and that ensures data security. Disclosure of Invention The invention aims to overcome the defects of the prior art and provide a remote artificial intelligent detection method and a system for bone age digital X-ray images, which have the advantages of layered collaboration, large service coverage, standard detection flow, reliable detection result, high safety and small operation load. The aim of the invention can be achieved by the following technical scheme: a remote artificial intelligence detection method of bone age digital X-ray image comprises the following steps: Remotely acquiring X-ray image data, wherein the X-ray image data is transmitted and acquired by a 5G private network in a desensitization encryption mode; Normalizing the received X-ray image data; judging required GPU computing power resources according to the normalized X-ray image data, and calling at least one required AI computing server based on a load balancing principle; acquiring a bone age detection result of an AI calculation server, and generating a detection report; Wherein the X-ray image data and the detection report are stored in an AI storage device, and the storage of the X-ray image data and the detection report complies with IHE integration specifications. Further, the X-ray image data is obtained through an API interface, a message queue, web Service docking or a WebSocket docking mode. Further, the desensitizing encryption includes: And performing desensitization treatment on the original data acquired by the X-ray bone age equipment, reserving minimum data required by AI calculation, and encrypting to form the X-ray image data. Further, the normalization processing is performed on the X-ray image data by adopting a quantile suitable for the bone age of the children. Further, the GPU computing power resources required for the judgment are specifically: and determining t