Data Security Transmission Method and System for Hospital Digital Operation Platform

Abstract

The present invention relates to the technical field of data secure transmission, and proposes a data secure transmission method and system for a hospital digital operation platform, including: dividing core shards, secondary shards and salt shards in medical image data; determining target core sub-shards, adjusting pixel values in the salt shards to obtain adjusted salt shards for the target core sub-shards, determining the target core sub-shards with one-time salt addition and the core shards with one-time salt addition, and obtaining the secondary shards with one-time salt addition; using the secondary shards as salt shards, obtaining the core sub-shards with secondary division, obtaining the core shards with secondary salt addition and the core shards with secondary salt addition, using the salted secondary shards as salt shards to add salt to the core shards, and obtaining the core shards with three-time salt addition; obtaining a key and an encrypted result according to the core shards with three-time salt addition, the salted secondary shards and all salt shards, and performing secure transmission on the encrypted result. The present invention can shorten the transmission delay during the transmission of medical image data.

Description

Technical Field

[0001] The present invention relates to the technical field of data secure transmission, and particularly to a data secure transmission method and system for a hospital digital operation platform. Background Art

[0002] Medical images are an indispensable and important part of modern medical services, playing a key role in clinical diagnosis, treatment plan formulation, disease monitoring, etc. The introduction of digital technology enables medical images to be stored and transmitted in digital form, greatly improving the management efficiency and accessibility of image data. With the popularization of cross-institutional sharing of medical data such as remote consultations, the security issue of medical image transmission has become increasingly prominent. The salt addition algorithm can be used to encrypt medical image data, and then the encrypted data is transmitted to ensure the security of medical image transmission.

[0003] However, in order to ensure data security, when using the salt addition algorithm to encrypt medical image data, the data volume of the encrypted medical image data will increase, significantly increasing the encryption calculation time and resulting in an increase in transmission delay. Summary of the Invention

[0004] The present invention provides a data secure transmission method and system for a hospital digital operation platform to solve the problem of increased transmission delay caused by encrypting data using the salt addition algorithm during the transmission of medical image data. The specific technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present invention provides a data secure transmission method for a hospital digital operation platform. The method includes the following steps:

[0006] Divide the lesion area in the medical image data, obtain the core shards and secondary shards, and obtain the salt shard;

[0007] Fill and divide the core shards according to the size of the salt shard to obtain core sub-shards. Denote any one of the core sub-shards as the target core sub-shard. According to the difference in the pixel value distribution between the target core sub-shard and the salt shard, adjust the pixel values in the salt shard to obtain the adjusted salt shard of the target core sub-shard. Rotate the adjusted salt shard of the target core sub-shard at different angles. According to the consistency degree of the pixel value distribution included in the rotated adjusted salt shard and the target core sub-shard, determine the target core sub-shard for one-time salt addition, and then determine the core shards for one-time salt addition. Obtain the secondary shards for one-time salt addition according to the size of the salt shard. Use the salt-added core shards and salt-added secondary shards to update the core shards and secondary shards respectively;

[0008] Take the secondary shards as salt shards, fill and divide the core shards according to the size of the secondary shards to obtain the secondarily divided core sub-shards. Rotate the secondary shards at different angles respectively, and obtain the salted secondarily divided core sub-shards according to the entropy values of the superposition results of the secondary shards rotated at different angles and the secondarily divided core sub-shards, and then obtain the core shards with secondary salting. According to the consistency degree of the pixel value distributions included in the core shards divided from the medical image data and the core shards with secondary salting, use the salted secondary shards as salt shards to salt the core shards to obtain the core shards with tertiary salting;

[0009] Obtain the secret key and the encryption result according to the core shards with tertiary salting, the salted secondary shards and all the salt shards, and perform secure transmission on the encryption result.

[0010] Furthermore, the specific methods for obtaining the core shards and the secondary shards and obtaining the salt shards include:

[0011] Record the lesion area in the medical image data as the core shards;

[0012] Record all areas in the medical image data that are not core shards as secondary shards;

[0013] Divide the medical image data into sub-regions of a preset size, use the entropy calculation model to obtain the probability values corresponding to all sub-regions in the medical image data, select the sub-region with the largest probability value in the probability map, and record the ratio of the number of pixel points in the core shards contained in the sub-region to the total number of pixel points contained in the sub-region as the lesion overlap rate of the sub-region;

[0014] When the lesion overlap rate of the sub-region is less than or equal to the first judgment threshold, record the selected sub-region as the salt shard;

[0015] When the lesion overlap rate of the sub-region is greater than the first judgment threshold, remove all sub-regions with a lesion overlap rate greater than the first judgment threshold from the probability map and re-select the sub-region with the largest probability value. Compare the selected sub-region with the value of the first judgment threshold. When the lesion overlap rate of the sub-region is less than or equal to the first judgment threshold, record the selected sub-region as the salt shard area. When the lesion overlap rate of the sub-region is greater than the first judgment threshold, re-select the sub-region according to the same steps until the salt shard is determined.

[0016] Furthermore, the specific methods for filling and dividing the core shards according to the size of the salt shards to obtain the core sub-shards include:

[0017] Calculate whether the number of rows of the core shards is an integer multiple of the number of rows of the salt shards. If so, do not fill the rows of the core shards. If not, fill the rows of the core shards until the number of rows of the core shards is an integer multiple of the number of rows of the salt shards;

[0018] Calculate whether the number of columns of the core shard is an integer multiple of the number of columns of the salt shard. If it is, do not fill the columns of the core shard. If not, fill the columns of the core shard with random numbers until the number of columns of the core shard is an integer multiple of the number of columns of the salt shard;

[0019] Divide the core shard into core sub-shards of the same size as the salt shard.

[0020] Further, the specific method for adjusting the pixel values in the salt shard according to the difference in the pixel value distribution between the target core sub-shard and the salt shard to obtain the adjusted salt shard of the target core sub-shard includes:

[0021] Divide the target core sub-shard using the Otsu method, and determine the foreground mean and background mean of the target core sub-shard according to the division result;

[0022] Divide the salt shard using the Otsu method, and determine the foreground mean and background mean of the salt shard according to the division result;

[0023] According to the foreground mean and background mean of the target core sub-shard, and the foreground mean and background mean of the salt shard, calculate the adjusted pixel value of each pixel point in the salt shard respectively, and record the salt shard composed of the adjusted pixel values of all pixel points as the adjusted salt shard of the target core sub-shard.

[0024] Further, the calculation method for the adjusted pixel value of each pixel point in the salt shard is:

[0025] Denote the difference between the foreground mean and the background mean of the target core sub-shard as the first difference of the target core sub-shard, denote the difference between the foreground mean and the background mean of the salt shard as the second difference of the salt shard, and denote the ratio of the first difference of the target core sub-shard to the second difference of the salt shard as the first ratio of the target core sub-shard;

[0026] Denote the product of the pixel value of any pixel point in the target core sub-shard and the first ratio of the target core sub-shard as the first product of any pixel point in the target core sub-shard, and denote the floor value of the sum of the background mean of the target core sub-shard and the first product of any pixel point in the target core sub-shard as the adjusted pixel value of any pixel point in the target core sub-shard after adjustment.

[0027] Further, the specific method for determining the target core sub-shard for one-time salting according to the consistency degree of the pixel value distribution included in the rotated adjusted salt shard and the target core sub-shard includes:

[0028] Obtain the grayscale distribution histograms of the rotated adjusted salt tablets and the target core sub-fragments respectively. According to the number of occurrences of all grayscale values in the grayscale distribution histograms, obtain the grayscale value distribution sequences of the rotated adjusted salt tablets and the target core sub-fragments respectively. Denote the DTW distance between the grayscale value distribution sequences of the rotated adjusted salt tablets and the target core sub-fragments as the confusion effect between the rotated adjusted salt tablets and the target core sub-fragments;

[0029] Take the rotated adjusted salt tablet corresponding to the maximum value in the confusion effect corresponding to the target core sub-fragment as the salt tablet for confusing with the target core sub-fragment. Superimpose the salt tablet for confusing with the target core sub-fragment and the target core sub-fragment to obtain the salt-added target core sub-fragment.

[0030] Furthermore, the specific method for determining the core fragment with one-time salting includes:

[0031] Arrange all the salt-added core sub-fragments according to the positions of the corresponding core sub-fragments of the salt-added core sub-fragments in the core fragments to obtain the salt-added core fragment.

[0032] Furthermore, the specific method for obtaining the salt-added secondary partition core fragment based on the value of the information entropy of the superimposed result of the secondary fragment rotated at different angles and the secondary partition core fragment, and then obtaining the core fragment with secondary salting includes:

[0033] Calculate the information entropy of the superimposed result of the rotated secondary fragment and the secondary partition core fragment at each rotation angle respectively. Take the rotation angle of the secondary fragment corresponding to the maximum value of the information entropy of all the superimposed results of the secondary partition core fragments as the rotation angle for confusing the core fragment with the secondary fragment;

[0034] Denote the superimposed result of the secondary fragment corresponding to the maximum value of the information entropy of all the superimposed results of the secondary partition core fragments and the secondary partition core fragment as the salt-added secondary partition core fragment;

[0035] Arrange all the salt-added secondary partition core fragments according to the positions of the secondary partition core fragments in the core fragments to obtain the core fragment with secondary salting.

[0036] Furthermore, the specific method for using the salt-added secondary fragment as a salt tablet to salt the core fragment according to the consistency degree of the pixel value distributions included in the core fragment and the core fragment with secondary salting divided from the medical image data to obtain the core fragment with tertiary salting includes:

[0037] Obtain the grayscale distribution histograms of the core shards and the core shards with secondary salting obtained by partitioning the medical image data respectively. According to the grayscale distribution histograms, determine the grayscale value distribution sequences of the core shards and the core shards with secondary salting obtained by partitioning the medical image data respectively, and the matching relationship between the number of occurrences of grayscale values in the grayscale value distribution sequences. Denote the difference between the number of occurrences of the corresponding grayscale values in the core shards and the core shards with secondary salting obtained by partitioning the medical image data as the frequency difference of the number of occurrences of the corresponding grayscale values. Take the pixel value corresponding to the maximum value of all frequency differences as the central frequency difference, and establish a local window of the central frequency difference with a preset window length centered on the central frequency difference;

[0038] For the core shards with secondary salting, calculate the between-class variance of all pixel values included in the local window of the central frequency difference and all pixel values excluded from the local window of the central frequency difference. Increase the window length by the same preset step size until there are no pixel values outside the local window, and stop calculating the between-class variance corresponding to the value of each window length;

[0039] Select the local window of the central frequency difference corresponding to the maximum value of the between-class variance. Denote the minimum value of the pixel value interval corresponding to the local window as the secondary salting background mean of the core shards with secondary salting, and denote the maximum value of the pixel value interval corresponding to the local window as the secondary salting foreground mean of the core shards with secondary salting;

[0040] Use the maximum between-class variance method to partition the secondary shards, and determine the foreground mean and background mean of the secondary shards;

[0041] Take the secondary salting background mean of the core shards with secondary salting as the background mean of the core shards with secondary salting, take the secondary salting foreground mean of the core shards with secondary salting as the foreground mean of the core shards with secondary salting. Combine the foreground mean and background mean of the secondary shards, and use the salted secondary shards as salt shards to salt the core shards to obtain the core shards with tertiary salting.

[0042] In a second aspect, an embodiment of the present invention further provides a data security transmission system for a hospital digital operation platform, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0043] The beneficial effects of the present invention are:

[0044] This application first determines the core shards and secondary shards based on the lesion areas in the medical image data, and dynamically selects salt shards. The core shards are the lesion areas in the medical image data, the secondary shards are the other areas in the medical image data except the lesion areas, and the salt shards are the low information density areas in the medical image data. The salt shards are used to perform matching and obfuscation processing on the core shards and secondary shards respectively. The salt shards can be attached to the core shards and secondary shards, and then attached to the medical image data, mixing the effective information of the medical image data with the chaotic information of the salt shards to cover the original information of the medical image data. During the process of adding salt to the information on the core shards and secondary shards, salt shards with different rotation angles are used to mix the information of the sub-shards of the core shards and secondary shards respectively, and the core shards and secondary shards for the first salt addition are determined according to the mixing effect, improving the security of medical image data encryption. Further, since the information contained in the core shards is more important for the medical image data, the secondary shards are selected as the salt shards to perform the second salt addition processing on the core shards, further improving the security when the medical image data is transmitted. Then, according to the gray level distribution difference between the core shards divided from the medical image data and the core shards after the second salt addition, the third salt addition to the core shards is determined. The three times of salt addition can improve the security of medical image data encryption. At the same time, even if two users use the same password, due to the different salt values, the generated hash values will be different, effectively preventing rainbow table attacks and brute force cracking. Finally, the encrypted results are transmitted in different secure ways, which can solve the problem of increased transmission delay caused by encrypting data using the salt addition algorithm during the transmission of medical image data, and shorten the transmission delay during the transmission of medical image data. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0046] Figure 1 It is a flowchart of a data security transmission method for a hospital digital operation platform provided by an embodiment of the present invention;

[0047] Figure 2 It is a flowchart for obtaining adjusted pixel values provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0049] Please refer to Figure 1 , which shows a flowchart of a data security transmission method for a hospital digital operation platform provided by an embodiment of the present invention. The method includes the following steps:

[0050] Step S001, divide the lesion area in the medical image data, obtain the core shards and secondary shards, and obtain the salt shards.

[0051] Retrieve the medical image data to be transmitted from the PACS system.

[0052] Among them, the PACS system, namely the Picture Archiving and Communication Systems, is an integrated system for digital storage, management, transmission, and display of medical images; the output standard of the medical image data is a DICOM format file, and sensitive patient information fields such as the patient's name and date of birth in the DICOM tags need to be removed, and the medical image data is anonymized encoded using a hash algorithm.

[0053] Take the medical image data with the lesion area divided as the training set, use an improved U-Net architecture to train the neural network model, obtain the trained data model, use the trained data model to segment the lesion area in the medical image data, record the divided lesion area as the core shard, and record all areas that are not the core shard as secondary shards.

[0054] Taking the 32*32 image area as the preset area size of the salt slice area, dividing the medical image data into 32*32 sub-areas, using the information entropy calculation model to obtain the probability values corresponding to all sub-areas in the medical image data, selecting the sub-area with the largest probability value in the probability map, and recording the ratio of the number of pixel points in the core slice contained in the sub-area to the total number of pixel points contained in the sub-area as the lesion overlap rate of the sub-area. When the lesion overlap rate of the sub-area is less than or equal to the first judgment threshold, the selected sub-area is recorded as a salt slice; when the lesion overlap rate of the sub-area is greater than the first judgment threshold, remove all sub-areas with a lesion overlap rate greater than the first judgment threshold from the probability map and re-select the sub-area with the largest probability value, compare the selected sub-area with the value of the first judgment threshold, when the lesion overlap rate of the sub-area is less than or equal to the first judgment threshold, record the selected sub-area as the salt slice area, when the lesion overlap rate of the sub-area is greater than the first judgment threshold, re-select the sub-area according to the same steps until the salt slice is determined.

[0055] It can be understood that the core slice and the secondary slice together constitute the medical image data. The core slice is the lesion area in the medical image data, and the secondary slice is the other area in the medical image data except the lesion area. The salt slice is the low information density area in the medical image data, such as the area corresponding to the uniform background and non-critical structures.

[0056] Among them, in this embodiment, the 32*32 image area is selected as the preset area size of the salt slice area. The larger the salt slice area, the lower the security of encryption using the salt addition algorithm. The value of the first judgment threshold in this embodiment is 5%. In the actual application process, as other implementation manners, the implementer can determine the preset area size of the salt slice area and the value of the first judgment threshold according to the actual situation by himself / herself, and this application does not make special restrictions.

[0057] So far, the salt slice is obtained, and the core slice and the secondary slice in the medical image data are obtained.

[0058] Step S002, fill and segment the core slice according to the size of the salt slice to obtain core sub-slices. Denote any one of the core sub-slices as the target core sub-slice. According to the difference in the pixel value distribution between the target core sub-slice and the salt slice, adjust the pixel values in the salt slice to obtain the adjusted salt slice of the target core sub-slice. Rotate the adjusted salt slice of the target core sub-slice at different angles respectively. According to the consistency degree of the pixel value distribution contained in the rotated adjusted salt slice and the target core sub-slice, determine the target core sub-slice for one-time salt addition, and then determine the core slice for one-time salt addition. Obtain the secondary slice for one-time salt addition according to the size of the salt slice, and use the salt-added core slice and the salt-added secondary slice to update the core slice and the secondary slice respectively.

[0059] When using a salt algorithm to encrypt medical image data, salt tablets are needed to match and confuse the core shards and secondary shards. Among them, the confusion process is the salt addition process, that is, the encryption process. Then, a dynamic salt value is generated according to the salt tablets, and the matching and confusion information is placed in the data shards. Combining the pre-shared master secret key and the dynamic salt value, a shard encryption secret key is generated, and the medical image data is encrypted through the shard encryption secret key. Finally, the encrypted results are independently transmitted through different transmission channels. Among them, the salt tablets are transmitted through low-risk lines or intranet lines, and the other shards are transmitted through conventional transmission channels to reduce the transmission pressure on the intranet lines.

[0060] First, it is necessary to fill the core shard according to the size of the salt tablet. Calculate whether the number of rows of the core shard is an integer multiple of the number of rows of the salt tablet. If so, do not fill the rows of the core shard. If not, fill the rows of the core shard with random numbers until the number of rows of the core shard is an integer multiple of the number of rows of the salt tablet. Calculate whether the number of columns of the core shard is an integer multiple of the number of columns of the salt tablet. If so, do not fill the columns of the core shard. If not, fill the columns of the core shard with random numbers until the number of columns of the core shard is an integer multiple of the number of columns of the salt tablet. Secondly, the core shard is divided according to the size of the salt tablet, and the core shard is divided into core sub-shards of the same size as the salt tablet.

[0061] For example, the size of the salt tablet is 32*32, and the size of the core shard is 410*512. Then the number of rows of the salt tablet is 32, and the number of rows of the core shard is 410. 410 is not an integer multiple of 32. Therefore, the rows of the core shard are filled until 410 is increased to 416, which is an integer multiple of 32, that is, 6 rows of random numbers are filled in the core shard. The number of columns of the salt tablet is 32, and the number of columns of the core shard is 512. 512 is an integer multiple of 32, so the columns of the core shard are not filled.

[0062] Use the salt tablets to match and confuse the core shards and secondary shards respectively. The purpose is to attach the salt tablets to the core shards and secondary shards, and then attach the salt tablets to the medical image data, mixing the effective information of the medical image data and the chaotic information of the salt tablets to cover the original information of the medical image data.

[0063] Denote any one of the core sub - slices as the target core sub - slice. Use the Otsu method to divide all the pixel values contained in the target core sub - slice to obtain the first division threshold. Denote the mean value of all pixel points in the target core sub - slice that are greater than or equal to the first division threshold as the foreground mean value of the target core sub - slice, and denote the mean value of all pixel points in the target core sub - slice that are less than the first division threshold as the background mean value of the target core sub - slice. Use the Otsu method to divide all the pixel values contained in the salt slice to obtain the second division threshold. Denote the mean value of all pixel points in the salt slice that are greater than or equal to the second division threshold as the foreground mean value of the salt slice, and denote the mean value of all pixel points in the salt slice that are less than the second division threshold as the background mean value of the salt slice.

[0064] In order to obfuscate the information contained in the target core sub - slice, it is necessary to adjust the pixel value range of the pixel points contained in the salt slice to be similar to the pixel value distribution of the target core sub - slice.

[0065] According to the foreground mean value and background mean value of the target core sub - slice, and the foreground mean value and background mean value of the salt slice, calculate the adjusted pixel value of each pixel point in the salt slice respectively. Denote the salt slice composed of the adjusted pixel values of all pixel points as the adjusted salt slice of the target core sub - slice. Among them, the calculation formula for the adjusted pixel value of each pixel point in the adjusted salt slice is:

[0066]

[0067] In the formula: represents the adjusted pixel value after adjustment of the th pixel point in the target core sub - slice; represents the floor function symbol, whose function is to round down the value inside the symbol; represents the th pixel point in the target core sub - slice; represents the foreground mean value of the salt slice; represents the background mean value of the salt slice; represents the foreground mean value of the target core sub - slice; represents the background mean value of the target core sub - slice.

[0068] Among them, the th pixel point in the target core sub - slice is any pixel point in the target core sub - slice, takes values in the range of all natural numbers greater than or equal to 1 and less than or equal to the number of pixel points contained in the target core sub - slice.

[0069] Among them, it should be noted that the difference between the foreground mean value and the background mean value of the target core sub-slice is denoted as the first difference of the target core sub-slice, the difference between the foreground mean value and the background mean value of the salt slice is denoted as the second difference of the salt slice, and the ratio of the first difference of the target core sub-slice to the second difference of the salt slice is denoted as the first ratio of the target core sub-slice. The first ratio of the target core sub-slice is . The product of the pixel value of the -th pixel point in the target core sub-slice and the first ratio of the target core sub-slice is denoted as the first product of the -th pixel point in the target core sub-slice. The floor value of the sum of the background mean value of the target core sub-slice and the first product of the -th pixel point in the target core sub-slice is denoted as the adjusted pixel value of the -th pixel point in the target core sub-slice.

[0070] The flowchart for obtaining the adjusted pixel value is as Figure 2 shown.

[0071] It can be understood that the adjusted salt slice of the target core sub-slice is a salt slice whose pixel value distribution is adjusted to be similar to that of the target core sub-slice. The adjusted salt slice of any core sub-slice can be obtained in the same way. That is to say, for any core sub-slice, there is a corresponding adjusted salt.

[0072] When using the adjusted salt slice to perform confusion processing on the target core sub-slice, the result of direct confusion processing may not be the best. Therefore, the adjusted salt slice of the target core sub-slice is rotated clockwise by 0°, 90°, 180°, and 270° respectively, and the confusion effect of the rotated adjusted salt slice and the target core sub-slice is calculated according to the degree of consistency of the pixel value distributions included in the rotated adjusted salt slice and the target core sub-slice.

[0073] Specifically, the gray-scale distribution histograms of the rotated adjusted salt slice and the target core sub-slice are obtained respectively. The number of occurrences of all gray-scale values in the gray-scale distribution histogram of the rotated adjusted salt slice is arranged in ascending order of gray-scale value to obtain the gray-scale value distribution sequence of the rotated adjusted salt slice. The number of occurrences of all gray-scale values in the gray-scale distribution histogram of the target core sub-slice is arranged in ascending order of gray-scale value to obtain the gray-scale value distribution sequence of the target core sub-slice. The DTW distance between the gray-scale value distribution sequences of the rotated adjusted salt slice and the target core sub-slice is denoted as the confusion effect of the rotated adjusted salt slice and the target core sub-slice.

[0074] Among them, obtaining the gray-scale distribution histogram is a well-known technology and will not be elaborated here.

[0075] When the confusion effect between the rotated adjusted salt slice and the target core sub-slice is greater, the difference in the image gray value distribution between the rotated adjusted salt slice and the target core sub-slice is greater. The better the effect of using the rotated adjusted salt slice to mix information and cover the original information of the medical image data for the target core sub-slice, and the better the effect of selecting the rotated adjusted salt slice for confusion processing.

[0076] Take the rotated adjusted salt slice corresponding to the maximum value in the confusion effect corresponding to the target core sub-slice as the salt slice for confusion processing with the target core sub-slice, and superimpose the salt slice for confusion processing with the target core sub-slice and the target core sub-slice to obtain the salt-added target core sub-slice.

[0077] Among them, superimposing the salt slice and the target core sub-slice means summing the pixel values of the corresponding position pixel points. In the same way, the salt-added core sub-slice corresponding to any core sub-slice can be obtained. That is to say, for any core sub-slice, there is a corresponding salt-added core sub-slice.

[0078] Arrange all the salt-added core sub-slices according to the positions of the core sub-slices corresponding to the salt-added core sub-slices in the core slice to obtain the salt-added core slice.

[0079] Save the first ratio of the target core sub-slice, the background mean value of the target core sub-slice, and the rotation angle of the rotated adjusted salt slice corresponding to the maximum value in the confusion effect corresponding to the target core sub-slice in the key for decryption.

[0080] According to the method of filling and segmenting the core slice to obtain the core sub-slice and obtaining the salt-added core sub-slice, perform filling and segmenting processing on the secondary slice to obtain the secondary sub-slice and obtain the salt-added secondary sub-slice. Arrange the secondary sub-slices corresponding to all the salt-added secondary sub-slices according to their positions in the secondary slice to obtain the salt-added secondary slice.

[0081] Replace the core slice with the salt-added core slice to obtain the updated core slice; replace the secondary slice with the salt-added secondary slice to obtain the updated secondary slice.

[0082] It can be understood that the process of obtaining the updated core slice and the updated secondary slice is the process of performing the first salt addition on the core slice and the secondary slice.

[0083] So far, the updated core slice and the updated secondary slice are obtained.

[0084] Step S003: Take the secondary shards as salt shards, fill and divide the core shards according to the size of the secondary shards to obtain the secondarily divided core sub-shards. Rotate the secondary shards at different angles respectively, and obtain the salt-added secondarily divided core sub-shards according to the entropy values of the superposition results of the secondarily divided core sub-shards and the secondary shards rotated at different angles, and then obtain the core shards with secondary salt addition. According to the consistency degree of the pixel value distributions included in the core shards divided from the medical image data and the core shards with secondary salt addition, use the salt-added secondary shards as salt shards to add salt to the core shards to obtain the core shards with tertiary salt addition.

[0085] Among the core shards and the secondary shards, the information contained in the core shards is more important for the medical image data. Select the secondary shards as salt shards to perform obfuscation processing on the core shards to improve the security during data transmission of the medical image data.

[0086] The method of filling and dividing the core shards according to the size of the salt shards to obtain the core sub-shards is to fill and divide the core shards according to the size of the secondary shards to obtain the secondarily divided core sub-shards. It can be understood that the secondarily divided core sub-shards are the same size as the secondary shards. Perform histogram equalization on the secondarily divided core sub-shards and the secondary shards respectively to reduce the information difference between the secondarily divided core sub-shards and the secondary shards.

[0087] Rotate the secondary shards clockwise by 0°, 90°, 180°, and 270° respectively, and determine the rotation angle of the rotated secondary shards and the salt-added secondarily divided core sub-shards according to the entropy of the superposition results of the rotated secondary shards and the secondarily divided core sub-shards.

[0088] Specifically, calculate the entropy of the superposition results of the rotated secondary shards and the secondarily divided core sub-shards at each rotation angle respectively, determine the maximum value of the entropy of the superposition results of all the secondarily divided core sub-shards, take the rotation angle of the secondary shards corresponding to the maximum value of the entropy of the superposition results of all the secondarily divided core sub-shards as the rotation angle for obfuscating the core shards with the secondary shards, and record the superposition result of the secondary shards corresponding to the maximum value of the entropy of the superposition results of all the secondarily divided core sub-shards and the secondarily divided core sub-shards as the salt-added secondarily divided core sub-shards. Arrange all the salt-added secondarily divided core sub-shards according to the positions of the secondarily divided core sub-shards in the core shards to obtain the core shards with secondary salt addition.

[0089] Among them, performing histogram equalization and calculating entropy are both well-known technologies and will not be elaborated here.

[0090] Obtain the grayscale distribution histograms of the core shards and the core shards with secondary salt addition respectively from the medical image data. Arrange the number of occurrences of all grayscale values in the grayscale distribution histogram of the core shards obtained from the medical image data in ascending order of grayscale values to obtain the grayscale value distribution sequence of the core shards obtained from the medical image data. Arrange the number of occurrences of all grayscale values in the grayscale distribution histogram of the core shards with secondary salt addition in ascending order of grayscale values to obtain the grayscale value distribution sequence of the core shards with secondary salt addition. Use the DTW algorithm to match the number of occurrences of grayscale values in the grayscale value distribution sequences of the core shards and the core shards with secondary salt addition obtained from the medical image data. Denote the difference between the number of occurrences of the corresponding grayscale values in the core shards and the core shards with secondary salt addition obtained from the medical image data as the frequency difference of the number of occurrences of the corresponding grayscale values. Take the pixel value corresponding to the maximum value of all frequency differences as the central frequency difference. With the central frequency difference as the center and 2 as the window length, establish a local window of the central frequency difference. It should be noted that the corresponding values within the local window of the central frequency difference are all pixel values. For example, when the central frequency difference is 50, the local window of the central frequency difference is 。

[0091] For the core shards with secondary salt addition, calculate the between-class variance of all pixel values included within the local window of the central frequency difference and all pixel values included outside the local window of the central frequency difference. Denote the calculated between-class variance as the between-class variance corresponding to the value of the window length. Further, take the number 2 as the step size for increasing the window length, and calculate the between-class variance corresponding to each value of the window length respectively until there are no pixel values outside the local window to stop the calculation. For example, if the local window of the central frequency difference is ,taking the number 2 as the step size for increasing the window length, then after increasing the step size once, the local window of the central frequency difference is 。

[0092] Among them, the calculation of the between-class variance is a well-known technology and will not be elaborated here.

[0093] Select the local window of the central frequency difference corresponding to the maximum value of the between-class variance. Denote the minimum value of the pixel value interval corresponding to the local window as the secondary salt addition background mean of the core shards with secondary salt addition, and denote the maximum value of the pixel value interval corresponding to the local window as the secondary salt addition foreground mean of the core shards with secondary salt addition.

[0094] Use the maximum between-class variance method to divide the secondary shards to obtain the third division threshold. Denote the mean value of all pixel points in the secondary shards that are greater than or equal to the third division threshold as the foreground mean of the secondary shards, and denote the mean value of all pixel points in the secondary shards that are less than the third division threshold as the background mean of the secondary shards.

[0095] According to the method of salting the core sub - slices based on the salt slices to obtain the salted secondary sub - slices, the background mean value of the second - level salted core slice is used as the background mean value of the second - level salted core slice, and the foreground mean value of the second - level salted core slice is used as the foreground mean value of the second - level salted core slice. Combining the foreground mean value and the background mean value of the secondary slice, the salted secondary slice is used as a salt slice to salt the core slice, and the core slice with triple salting is obtained.

[0096] It can be understood that the core slice with triple salting is the final result obtained by encrypting the core slice.

[0097] Thus far, the core slice with triple salting is obtained.

[0098] Step S004: Obtain the key and the encryption result based on the core slice with triple salting, the salted secondary slice, and all salt slices, and perform secure transmission of the encryption result.

[0099] To ensure the security of the key when using the salting algorithm to encrypt medical image data, this application dynamically selects salt slices according to the medical image data, generates salt values through the salt slices, and generates a dynamic key in combination with the main key. In this way, even if two users use the same password, due to different salt values, the generated hash values will be different, thus effectively preventing rainbow table attacks and brute - force cracking.

[0100] Convert the salt slice into a grayscale image, obtain the data stream of the salt slice according to the row - first principle, and convert the salt slice data stream into a dynamic salt value through a hash algorithm. Use the key distribution system of the core slice with triple salting to obtain the pre - allocated shared main key, generate a slice encryption key through the KDF key derivation function, encrypt all salt slices respectively using the slice encryption key, transmit the encryption results of the core slice with triple salting, the salted secondary slice, and all salt slices through a public line, and transmit the medical image data through a third - party audit pipeline.

[0101] Among them, when encrypting all salt slices respectively using the slice encryption key, Huffman coding is selected for encryption. Converting the salt slice into a grayscale image, obtaining the data stream of the salt slice according to the row - first principle, converting the salt slice data stream into a dynamic salt value through a hash algorithm, obtaining the pre - allocated shared main key, generating a slice encryption key through the KDF key derivation function, encrypting all salt slices respectively using the slice encryption key, and transmitting data through a public line and a third - party audit pipeline are all well - known technologies and will not be elaborated here.

[0102] After receiving the encrypted results of the triple-salted core shards, salted secondary shards, and all salt shards, the receiver first decrypts the salt shards. Using the decrypted salt shards and the shared master secret key, the encryption keys for the core shards, secondary shards, and all salt shards are calculated. The core shards, secondary shards, and all salt shards are decrypted using the encryption keys. During the decryption process, based on the data for decryption such as the first ratio of the target core sub-shard combined in the key, the background mean of the target core sub-shard, and the rotation angle of the rotated adjusted salt shard corresponding to the maximum value in the confusion effect corresponding to the target core sub-shard, the medical image data is obtained.

[0103] Thus, the secure data transmission for the hospital digital operation platform is realized.

[0104] Based on the same inventive concept as the above method, an embodiment of the present invention further provides a secure data transmission system for a hospital digital operation platform, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above methods for secure data transmission for a hospital digital operation platform are implemented.

[0105] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data security transmission method for a hospital digital operation platform, characterized in that The method includes the following steps: Divide the lesion area in the medical image data to obtain a core slice, a secondary slice, and a salt slice. Among them, the core slice is the lesion area in the medical image data; the secondary slice is all areas in the medical image data that are not core slices, and the salt slice is the low information density area in the medical image data; Fill and segment the core slice according to the size of the salt slice to obtain core sub-slices. Denote any one of the core sub-slices as the target core sub-slice. According to the difference in the pixel value distribution between the target core sub-slice and the salt slice, adjust the pixel values in the salt slice to obtain the adjusted salt slice of the target core sub-slice. Rotate the adjusted salt slice of the target core sub-slice at different angles respectively. According to the consistency degree of the pixel value distribution contained in the rotated adjusted salt slice and the target core sub-slice, determine the target core sub-slice with salt addition once. Arrange the positions of the core sub-slices corresponding to all the core sub-slices with salt addition in the core slice to obtain the core slice with salt addition once. Obtain the secondary slice with salt addition once according to the size of the salt slice. Use the core slice with salt addition and the secondary slice with salt addition to update the core slice and the secondary slice respectively; Take the secondary slice as the salt slice, fill and segment the core slice according to the size of the secondary slice to obtain the core sub-slices after secondary division. Rotate the secondary slice at different angles respectively. According to the value of the information entropy of the superposition result of the secondary slice rotated at different angles and the core sub-slices after secondary division, obtain the core sub-slices after secondary division with salt addition. Arrange all the core sub-slices after secondary division with salt addition according to the positions of the core sub-slices after secondary division in the core slice to obtain the core slice with salt addition twice. According to the consistency degree of the pixel value distribution contained in the core slice divided from the medical image data and the core slice with salt addition twice, use the secondary slice with salt addition as the salt slice to add salt to the core slice to obtain the core slice with salt addition three times; Obtain the key and the encryption result according to the core slice with salt addition three times, the secondary slice with salt addition, and all the salt slices. Transmit the encryption results of the core slice with salt addition three times, the secondary slice with salt addition, and all the salt slices through a public line, and transmit the medical image data through a third-party audit pipeline.

2. The data security transmission method for a hospital digital operation platform according to claim 1, characterized in that The method for determining the salt slice is as follows: Divide the medical image data into sub-regions of a preset size, use the information entropy calculation model to obtain the probability values corresponding to all sub-regions in the medical image data, select the sub-region with the largest probability value in the probability map, and denote the ratio of the number of pixel points in the core slice contained in the sub-region to the total number of pixel points contained in the sub-region as the lesion overlap rate of the sub-region; When the lesion overlap rate of the sub-region is less than or equal to the first judgment threshold, denote the selected sub-region as the salt slice; When the lesion overlap rate of the sub-region is greater than the first judgment threshold, reselect the sub-region with the maximum probability value from all sub-regions other than those with a lesion overlap rate greater than the first judgment threshold in the probability map, and compare the selected sub-region with the value of the first judgment threshold. When the lesion overlap rate of the sub-region is less than or equal to the first judgment threshold, mark the selected sub-region as the salt tablet region. When the lesion overlap rate of the sub-region is greater than the first judgment threshold, reselect the sub-region according to the same steps until the salt tablet is determined.

3. The data security transmission method for a hospital digital operation platform according to claim 1, characterized in that, The specific method for filling and dividing the core shard according to the size of the salt tablet to obtain the core sub-shards includes: Calculate whether the number of rows of the core shard is an integer multiple of the number of rows of the salt tablet. If so, do not fill the rows of the core shard. If not, fill the rows of the core shard until the number of rows of the core shard is an integer multiple of the number of rows of the salt tablet; Calculate whether the number of columns of the core shard is an integer multiple of the number of columns of the salt tablet. If so, do not fill the columns of the core shard. If not, fill the columns of the core shard with random numbers until the number of columns of the core shard is an integer multiple of the number of columns of the salt tablet; Divide the core shard into core sub-shards of the same size as the salt tablet.

4. The data security transmission method for a hospital digital operation platform according to claim 1, wherein The specific method for adjusting the pixel values in the salt tablet according to the difference in the pixel value distribution between the target core sub-shard and the salt tablet to obtain the adjusted salt tablet of the target core sub-shard includes: Use the maximum inter-class variance method to divide the target core sub-shard, and determine the foreground mean and background mean of the target core sub-shard according to the division result; Use the maximum inter-class variance method to divide the salt tablet, and determine the foreground mean and background mean of the salt tablet according to the division result; According to the foreground mean and background mean of the target core sub-shard, and the foreground mean and background mean of the salt tablet, calculate the adjusted pixel value of each pixel point in the salt tablet respectively. Denote the salt tablet composed of the adjusted pixel values of all pixel points as the adjusted salt tablet of the target core sub-shard.

5. The data security transmission method for a hospital digital operation platform according to claim 4, wherein The calculation method for the adjusted pixel value of each pixel point in the salt tablet is: Denote the difference between the foreground mean and the background mean of the target core sub-shard as the first difference of the target core sub-shard, denote the difference between the foreground mean and the background mean of the salt tablet as the second difference of the salt tablet, and denote the ratio of the first difference of the target core sub-shard to the second difference of the salt tablet as the first ratio of the target core sub-shard; Denote the product of the pixel value of any pixel point in the target core sub-shard and the first ratio of the target core sub-shard as the first product of any pixel point in the target core sub-shard, and denote the downward rounding value of the sum of the background mean of the target core sub-shard and the first product of any pixel point in the target core sub-shard as the adjusted pixel value of any pixel point in the target core sub-shard after adjustment.

6. The data security transmission method for a hospital digital operation platform according to claim 1, wherein The specific method for determining the target core sub-shard for one-time salt addition according to the consistency degree of the pixel value distribution included in the adjusted salt tablet after rotation and the target core sub-shard includes: Obtain the grayscale distribution histograms of the rotated adjusted salt slice and the target core sub-slice respectively. According to the number of occurrences of all grayscale values in the grayscale distribution histogram, obtain the grayscale value distribution sequences of the rotated adjusted salt slice and the target core sub-slice respectively. Denote the DTW distance between the grayscale value distribution sequences of the rotated adjusted salt slice and the target core sub-slice as the confusion effect between the rotated adjusted salt slice and the target core sub-slice. Take the rotated adjusted salt slice corresponding to the maximum value in the confusion effect corresponding to the target core sub-slice as the salt slice for confusion processing with the target core sub-slice. Superimpose the salt slice for confusion processing with the target core sub-slice and the target core sub-slice to obtain the salt-added target core sub-slice.

7. The data security transmission method for a hospital digital operation platform according to claim 1, characterized in that The specific method for obtaining the salt-added secondary partition core sub-slice according to the value of the information entropy of the superimposed result of the secondary slice rotated at different angles and the secondary partition core sub-slice includes: Calculate the information entropy of the superimposed result of the rotated secondary slice and the secondary partition core sub-slice at each rotation angle respectively. Take the rotation angle of the secondary slice corresponding to the maximum value of the information entropy of all superimposed results of the secondary partition core sub-slice as the rotation angle for confusing the core slice with the secondary slice. Denote the superimposed result of the secondary slice corresponding to the maximum value of the information entropy of all superimposed results of the secondary partition core sub-slice and the secondary partition core sub-slice as the salt-added secondary partition core sub-slice.

8. The data security transmission method for a hospital digital operation platform according to claim 1, wherein The specific method for using the salt-added secondary slice as the salt slice to add salt to the core slice to obtain the three-time salt-added core sub-slice according to the consistency degree of the pixel value distributions included in the core slice and the twice-salt-added core sub-slice divided from the medical image data includes: Obtain the grayscale distribution histograms of the core slice and the twice-salt-added core sub-slice divided from the medical image data respectively. According to the grayscale distribution histogram, determine the matching relationship between the grayscale value distribution sequences of the core slice and the twice-salt-added core sub-slice divided from the medical image data and the number of occurrences of grayscale values in the grayscale value distribution sequence respectively. Denote the difference between the number of occurrences of the corresponding grayscale values in the core slice and the twice-salt-added core sub-slice divided from the medical image data as the frequency difference of the number of occurrences of the corresponding grayscale values. Take the pixel value corresponding to the maximum value of all frequency differences as the central frequency difference, and establish a local window of the central frequency difference with a preset window length centered on the central frequency difference. For the twice-salt-added core sub-slice, calculate the between-class variance of all pixel values included in the local window of the central frequency difference and all pixel values included outside the local window of the central frequency difference. Increase the window length by the same preset step size until there are no pixel values outside the local window, and stop calculating the between-class variance corresponding to each window length value. Select the local window of the central frequency difference corresponding to the maximum value of the between-class variance. Denote the minimum value of the pixel value interval corresponding to the local window as the twice-salt-added background mean of the twice-salt-added core sub-slice, and denote the maximum value of the pixel value interval corresponding to the local window as the twice-salt-added foreground mean of the twice-salt-added core sub-slice. The maximum inter-class variance method is used to divide the secondary shards, and the foreground mean and background mean of the secondary shards are determined; The secondary salting background mean of the core shard with secondary salting is used as the background mean of the core shard with secondary salting, and the secondary salting foreground mean of the core shard with secondary salting is used as the foreground mean of the core shard with secondary salting. Combining the foreground mean of the secondary shards and the background mean of the secondary shards, the salted secondary shards are used as salt tablets to salt the core shards, and the core shards with three times of salting are obtained.

9. A data security transmission system for a hospital digital operation platform, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, the steps of the data security transmission method for the hospital digital operation platform according to any one of claims 1-8 are implemented.

Images