Research On Medical Image Privacy Protection Algorithm And Sharing Mechanism

With the rapid development of science and technology,the convenient and fast network sharing platform has changed people's lifestyles greatly.The telemedicine platform has also been favored by many people.People can conduct medical consultation and medical diagnosis through the network platform.However,there are both advantages and disadvantages.In the information age,all kinds of information are exposed in the network.The current domestic and international medical information disclosure incidents common.The leaked information is likely to cause huge economic property losses to patients.In the context of the gradual development of telemedicine,more and more medical images are transmitted and stored through the network.Once these medical images are uploaded to the network,it means that patients lose control of the data.This phenomenon exacerbates the insecurities of patient privacy information.In response to this problem,in order to effectively protect the personal information related to the medical image of patients,so that patients can safely consult through telemedicine,this paper studies the data protection and sharing mechanism of medical images.Firstly,after extensive collection and careful analysis of medical images,targeted protection methods are proposed for the characteristics of medical images.The main framework is divided into two parts: text area detection of medical images and hierarchical access control encryption of medical images.Patients can upload their medical images to the network before uploading them to individuals on medical images.The information area is encrypted and different decryption privileges are set for different visitors.In the text region detection of medical image,this paper proposes a text region detection algorithm based on the combination of maximum stable extremum region algorithm and improved corner detection algorithm.The algorithm firstly filters the maximum stable extremum region using component feature analysis,then uses the improved SUSAN corner detection operator to detect the remaining maximum stable extremum,and finally combined all the detected text regions with expansion processing to obtain the text region of medical image.In the hierarchical access controlencryption of medical image,this paper uses chaotic algorithm to encrypt the previously extracted personal information text area into the current medical image,and then encrypts the entire medical image.When the patient seeks diagnosis,the doctor can only decrypt the image once,that is only the diseased area is visible in the decrypted image,and the personal information area is still encrypted.Only a person authorized by the patient can decrypt the entire text area of the medical image.Through a large number of simulation experiments,the results show that the proposed method basically achieves the expected results: in the text area detection of medical images,the accuracy,recall and overall performance of the proposed algorithm are 0.9,0.92 and 0.91,respectively,achieving the ideal detection effect.In the hierarchical access control encryption of medical images,the algorithm proposed in this paper effectively classifies the rights of visitors,and analyzes the encrypted key image from key space and key sensitivity analysis.The five aspects of histogram analysis,information entropy analysis and adjacent pixel correlation analysis are carried out.The analysis results show that the encryption algorithm used in this paper has good performance and high security.