Design And Implementation Of Intelligent PACS

Picture Archiving and Communication Systems(PACS)is a comprehensive application system for storing archived image files,accessing image data and managing patient examination reports,which plays an important role in the diagnosis and treatment of major diseases such as cancer.With the rapid growth of image examination demand in recent years,radiologists are relatively insufficient,and the image diagnosis method relying on artificial reading faces a great challenge.Therefore,in the trend of human-computer integration,the clinical application of intelligent image diagnosis method is more urgent.The purpose of this paper is to design an intelligent PACS system.In addition to the common basic functions of PACS system,three new technical modules are added: data set annotation,intelligent image detection and auxiliary teaching.The main work of this paper includes:1.Aiming at the lack of professional medical image annotation dataset for deep learning research,In this paper,the data set annotation module is introduced to support doctors to annotate real-time in the work of diagnostic film reading,and generate XML annotation files to provide researchers with high-quality data sets.MSXML parsing tools are used to parse XML files,access and edit XML file nodes,and realize the functions of saving,loading and modifying annotations.;2.In view of the problem that the pressure of reading film is increasing rapidly in domestic imaging doctors,this paper designs an intelligent image detection module.In this module,deep learning model is introduced to detect the focus of medical image,asynchronous call of deep learning model is realized with the help of Active MQ message middleware technology,and the detection results of the model are stored in Redis cache for doctors' reference,so as to achieve the purpose of auxiliary image diagnosis.At the same time,the module implements data read-write separation by deploying Redis master-slave architecture,so as to relieve the pressure of Redis master server;3.In order to solve the problem that the existing system functions can not meet the needs of personnel training in the imaging department,this paper designs the auxiliary teaching module,and puts forward the three-level audit method of the diagnosis report,so that the interns can deeply participate in the diagnosis and treatment process.In this paper,a reasonable report processing process is designed to make the three-level audit process compatible with the traditional two-level audit process,and to provide a practice platform for imaging interns and regulatory doctors.At the same time,a learning platform is designed for informal doctors to support the comparative learning of image diagnosis reports.In theory,the PACS system in this paper enables doctors to mark the lesion sites during diagnosis,which is helpful to provide high-quality data sets and promote the development of deep learning in the field of medical image recognition.In practical application,the system preliminarily realizes the function of intelligent image detection,which is helpful to improve the accuracy of early cancer screening and diagnosis.The auxiliary teaching function of this system is helpful to improve the training effect of the imaging professionals and has important clinical application value.