Research On Medical Image Data Management Methods

In today's era,the application of medical big data is developing rapidly.The digitization of medical information and the electronic processing of medical treatments have become an important means to meet the rapid increase in population,the aging trend,and the growth of chronic disease patients.Medical imaging data is an important source of information generated throughout the medical process,and has become an important reference for doctors to diagnose and treat diseases.The in-depth and effective organization and utilization of medical images has become an increasingly important requirement for the times and the future.At present,the processing of medical image data is mainly directly deposited in the information system,in order to call the access to historical data,resulting in a significant waste of medical information resources.In fact,a large number of medical imaging data contain the law of disease development.If it can be deeply explored and correlated,it will provide scientific reference and important decision support for the diagnosis,prediction,monitoring and prevention of diseases.In order to achieve the above objectives,this study uses convolutional neural networks and long short-term memory to extract and classify image data,and proposes a general method for medical image data management,which mainly includes the following four aspects.(1)Analyze the current status of medical image data management research.Through domestic and foreign literature research,we can understand and analyze the current methods of medical image data management and the problems to be solved.At the same time,the current processing of image data is mainly based on convolutional neural network for feature extraction,and the structured timing description of the diagnosis results is not possible.In response to this,this study introduces a hybrid model of convolutional neural networks and long short-term memory in deep learning.(2)Analyze the needs of medical workers for medical image data management.Through questionnaires,we learned the deep demand of medical workers for medical image data management.The analysis of 146 valid survey data shows that the current domestic hospital data environment has the characteristics of large data volume,high data input andlow digitization;medical workers generally feel heavy workload,high difficulty and high pressure.In the process of reading the film for the diagnosis of the disease,the inexperienced young doctor is difficult to diagnose alone,and it is necessary to consult an expert,and sometimes it is difficult to diagnose the consultation.Workers generally demand effective data management methods that can provide a decision-making reference for their diagnosis.(3)Construct a general method of medical image data management,organize and manage medical images and diagnostic information,and provide information basis for the application of medical images and patient information,medical record information,disease type information storage and medical image data management methods.Using the hybrid model of convolutional neural network and long short-term memory in deep learning,through the continuous adjustment of convolutional and down-sampling parameters in convolutional neural networks,combined with long short-term memory and Softmax classifiers,the disease is optimal.The recognition model is further exemplified by the identification of benign and malignant thyroid nodules,and the detailed process of medical imaging and diagnostic information organization management is described in detail.In addition,based on the theory of data lifecycle management theory,the life-value of medical image data in each stage is analyzed,and the life-value curve of medical image data is drawn.(4)Research on multimodal medical big data associative storage and application.Based on the unified DICOM medical digital image and communication standard,build a medical image intelligent database to realize the organic connection of medical macro data such as medical images,patient information,medical record information and disease information;medical image data management methods can be applied to intelligent medical auxiliary diagnosis The medical image metadata format and transmission standard specifications,the prediction and prevention of human epidemics,and the construction and improvement of personal health files.