Patient privacy data protection method used in operating room

Abstract

The invention relates to the technical field of medical information security, and discloses a patient privacy data protection method used in an operating room. The method comprises the following steps: uniformly coding patient privacy into a binary system, zero-filling the binary system into a square matrix, and filling according to rows; a disturbance seed is generated through combination of operating room environment temperature and noise quantification, and chaotic mapping is driven to carry out two-dimensional replacement on a matrix; applying position-related mask XOR to the sub-blocks, and overturning odd columns or even rows according to global parameters; and then linearization is carried out according to rows and filling is eliminated, and a final ciphertext is output. Meanwhile, disturbance control parameters are constructed and bound, and then Hash check is calculated and structured packaging is carried out. Through global and local double-layer disturbance, field seed driving and reversible coding, the scheme enhances the resistance to statistical analysis, plaintext attack and replay attack on the premise of no fixed key, and gives consideration to real-time performance, security and traceability.

Description

Technical Field

[0001] This invention relates to the field of medical information security technology, specifically to a method for protecting patient privacy data in the operating room. Background Technology

[0002] As one of the highest-security scenarios in a hospital information system, the operating room generates and exchanges various types of patient privacy data in real time, including preoperative assessment forms, identity and insurance information, intraoperative vital signs data, anesthesia logs, instrument and implant codes, surgical field videos and image clips, and immediate postoperative medical orders and medication records. This data is characterized by high sensitivity, strong real-time requirements, and multi-source heterogeneity: on the one hand, it is continuously updated on a timescale of milliseconds to seconds; on the other hand, it exhibits structural differences at the bit level (text, numerical values, waveforms, image frames, and event logs coexist), and spans multiple devices and buses (patient monitors, anesthesia machines, endoscopic camera systems, surgical navigation terminals, nurse workstations, etc.). Operating room networks are often in semi-closed or isolated network segments, and field devices may be embedded / low-power processors, resulting in limited network bandwidth and computing resources; breakpoint resume, temporary network outages, offline caching, and sudden batch data transmissions are also common. These engineering constraints collectively impose three requirements on privacy protection solutions: low latency, low computational cost, reversibility, and auditability.

[0003] Current practices often rely on centralized access control, rule-based data masking, static encryption, and log tracing, but these methods face several limitations in the operating room setting: coarse-grained processing at the field or file level struggles to cover the streaming and fragmented nature of intraoperative data, especially regarding the insufficient protection against leaks of details in continuous waveforms and video frames; complex key distribution, session negotiation, or highly complex transformations can introduce latency jitter on low-power terminals, affecting the synchronization of vital sign acquisition and display; and using completely homogeneous processing paths and fixed parameters across multiple devices makes them vulnerable to correlation analysis and statistical feature attacks, leading to re-identification vulnerabilities. There are known / selected plaintext risks; most solutions are decoupled from environmental conditions and cannot incorporate instantaneous physical disturbances in the operating room into the protection path, resulting in convergence of "protection trajectories" at different times and in different operating rooms; to meet the needs of medical and legal departments for traceability, review, and medical record restoration, the protection solution must be reversible under strict authorization, while ensuring that the original text cannot be reconstructed by statistical methods in an unauthorized state; many traditional methods often cannot achieve both of these; under conditions such as non-contiguous data length, fragmentation misalignment caused by network jitter, and extreme value input, existing processes are prone to unrecoverable or ambiguous states.

[0004] Therefore, this case aims to propose a method for protecting patient privacy data in the operating room. First, various heterogeneous privacy information is uniformly encoded and mapped onto a matrix structure. Perturbation parameters are dynamically generated using physical random factors such as ambient temperature and noise in the operating room. Then, through multiple perturbations based on the principle of chaos, such as two-dimensional permutation, local window-level mask XOR, and global row and column flipping, high-intensity diffusion and confusion of data in the matrix space are achieved. Finally, the data is restored to a linear sequence and the padding is removed. The perturbation parameters and encryption results are then recorded, hashed, verified, and encapsulated for storage. Summary of the Invention

[0005] This invention provides a method for protecting patient privacy data in the operating room, thereby helping to solve the problems mentioned in the background section.

[0006] This invention provides the following technical solution: a method for protecting patient privacy data in the operating room, comprising:

[0007] The patient privacy information to be protected is uniformly encoded into a binary sequence, and zeros are padded to the end of the sequence to meet the preset matrix size requirements. Then, it is filled into the matrix in row-major order.

[0008] Perturbation seeds are generated by quantizing the measured values ​​of operating room ambient temperature and noise, and the intensity of the chaotic mapping is calculated accordingly.

[0009] Two-dimensional coordinate rearrangement mapping is generated using chaotic mapping intensity, and each element in the matrix is ​​replaced to the target position according to the mapping relationship;

[0010] The permuted matrix is ​​divided into multiple sub-blocks, and an XOR operation is performed on the sub-block based on the position-related mask.

[0011] Based on the global disturbance control parameters, perform bit-flipping operations sequentially on the odd columns or even rows of the matrix;

[0012] Extract the bit sequence from the perturbation matrix in row-major order and remove the aforementioned padding zeros to obtain the final binary output;

[0013] Based on the data perturbation process, perturbation control parameters, including temperature perturbation seed, noise perturbation seed, window size parameters, mask generation parameters, and global perturbation parameters, are constructed and recorded, and these parameters are bound and stored with patient identification information.

[0014] The hash check value is calculated for the final encrypted bit sequence, and the encryption result, perturbation parameters, and check information are encapsulated in a structured manner.

[0015] Optionally, the step of uniformly encoding the patient privacy information to be protected into a binary sequence, padding the end of the sequence with zeros to meet the preset matrix size requirements, and then filling it into the matrix in row-major order, specifically includes:

[0016] The patient's private information to be encrypted is uniformly encoded and converted into a binary bit sequence. Among them, D (0) It is a linear binary sequence of the initial state; The initial sequence is represented by the k-th bit; the k-th bit is the linear index; N is the total number of bits in the binary sequence to be protected.

[0017] Select one that satisfies L 2 Let L be the smallest positive integer ≥ N. Where l is a temporary variable used to define the conditions for the value of L; It is a set of positive integers;

[0018] Build length L 2 The extended bit sequence D (1) Fill any insufficient parts with 0:

[0019] in, For the k-th bit after expansion;

[0020] D in row priority (1) Reconstruct an L×L matrix M (0) , specific bits:

[0021] u,v∈{1,...,L}; where M (0) This is the initial matrix; Let be the element in row u and column v; (u-1)·L+v is the row-major index mapping; u is the matrix row index; v is the matrix column index.

[0022] Optionally, the step of generating a perturbation seed based on the quantization of operating room ambient temperature and noise measurements, and calculating the intensity of the chaotic map accordingly, specifically includes:

[0023] The ambient temperature T in the operating room was measured using an infrared temperature sensor and an ambient noise sensor, respectively. room and noise level A room And set the perturbation seed as:

[0024] in, A temperature seed for a 16-bit domain; This is the noise seed for a 16-bit field; mod is the modulo operation.

[0025] The chaotic mapping intensity parameter α is calculated based on the perturbation seed, specifically as follows:

[0026]

[0027] Optionally, the step of generating a two-dimensional coordinate rearrangement mapping using chaotic mapping intensity, and replacing each element in the matrix with the target position according to the mapping relationship, specifically includes:

[0028] Construct an empty L×L dimensional matrix M (1) And mark all locations as unoccupied;

[0029] For M (0) Determine the sequence in row order

[0030] For any Set the two-dimensional chaotic mapping Φ:

[0031]

[0032] Where Φ is the chaotic mapping function from two-dimensional coordinates to target coordinates; u′ and v′ are the mapped coordinates;

[0033] The construction length is L 2 The target candidate sequence, let δ u (m) = m mod L, m∈{0,1,...,L 2 -1};where, δ u (m) represents the offset of the m-th candidate in the row direction; δ v (m) represents the offset of the m-th candidate in the column direction; m is the candidate position enumeration counter;

[0034] And generate candidate coordinates m at a time:

[0035] (u (m) ,v (m) )=[(u′+δ u (m))mod L+1,(v′+δ v (m))mod L+1];where, (u (m) ,v (m) ) represents the m-th candidate position;

[0036] Starting from m=0, check sequentially and take the first unoccupied position. Place: Where, m * This is the first unoccupied candidate index found. To perform a matrix assignment operation, The bits are written to the target location;

[0037] and position Marked as occupied.

[0038] Optionally, dividing the permuted matrix into multiple sub-blocks and performing an XOR operation on the content of each sub-block based on a position-related mask specifically includes:

[0039] Take odd numbers Set the number of rows and columns for the sub-block as follows: Where G is the number of sub-blocks in the row direction; J is the number of sub-blocks in the column direction;

[0040] The actual row and column dimensions of the (g,h)th sub-block are:

[0041] r u (g)=min{r,Lr·(g-1)},r v (h) = min{r, Lr·(h-1)}; where g ∈ {1, ..., G} is the block number of the sub-block in the row direction; h ∈ {1, ..., J} is the block number of the sub-block in the column direction; r u (g) gives the actual height of the g-th row sub-block in the row direction; r v (h) gives the actual width of the h-th column sub-block in the column direction;

[0042] Its top-left corner coordinates: (u s (g),v s (h))=(r·(g-1)+1,r·(h-1)+1); where, u s (g), v s (g) Give the global coordinates of the top left corner of the sub-block in row g and column h, respectively;

[0043] Let C be a constant:

[0044]

[0045] Set a mask bit for any offset (s,t) within this sub-block:

[0046] w g,h (s,t)=((u s (g)+s)·(v s (h)+t)+C)mod 2; where s is the row offset relative to the top left corner of the sub-block, and its value ranges from [0, r]. u (g)-1]; t is the column offset within the sub-block relative to the top left corner, with a value range of [0, r]. v [(h)-1];w g,h (s,t) gives the position mask bits located within the (g,h)th sub-block at offset (s,t);

[0047] Construct an L×L dimensional empty matrix M (2) Specifically, it represents the matrix after XORing the sub-block level mask;

[0048] Let (u,v) = (u s (g)+s,v s Given (h)+t), apply a change to the elements within the domain (s,t), while keeping the elements outside the domain unchanged, and obtain... in, Let M be a matrix (2) The bit at position (u,v); This is a bitwise XOR operation.

[0049] Optionally, the step of performing bit-flipping operations on the odd columns or even rows of the matrix sequentially according to the global disturbance control parameters specifically includes:

[0050] Construct an L×L dimensional empty matrix M (3) Specifically, it represents the matrix after applying a global row or column perturbation;

[0051] Calculate the global branch bit Λ:

[0052]

[0053] If Λ = 0, execute the following for all cells in the odd-numbered sequence v mod 2 = 1. otherwise in, Let M be a matrix (3) Elements in (u,v);

[0054] If Λ=1, execute the following for all units in the even-numbered rows where u mod 2=0. otherwise,

[0055] Optionally, the step of extracting the bit sequence from the perturbation matrix in row-major order and removing the aforementioned padding zeros to obtain the final binary output specifically includes:

[0056] Extract M in row-major order (3) All bits in the array are a linear sequence:

[0057] k = (u-1)·L + v; where, D (2) This is the linearized result after local and global perturbations; For the k-th bit after linearization; k = (u-1)·L+v bit row-major flattening formula;

[0058] Extract the first N bits as the final encrypted output sequence: Among them, D (enc) A linear sequence of bit-encrypted output.

[0059] Optionally, based on the data perturbation process, constructing and recording perturbation control parameters including temperature perturbation seeds, noise perturbation seeds, window size parameters, mask generation parameters, and global perturbation parameters, and binding and storing them with patient identification information, specifically includes:

[0060] Constructing the perturbation parameter tuple

[0061] Define and execute the following mappings:

[0062] S701, 8-bit alpha quantization:

[0063] S702, Seed directly uses 16 bits: already configured

[0064] S703, Fixed bit width of each component:

[0065] N [32-bit], L [16-bit], r [16-bit], a [8-bit] [16-bit], [16-bit];

[0066] S704. Let the bitwise XOR mask constant vector be Ξ:

[0067] Ξ=(2 32 -1,2 16 -1,2 16 -1,2 8 -1,2 16 -1,2 16 -1);

[0068] Component-wise definition in, Bit parameter encoding result;

[0069] Will It is linked to the patient's data ID and stored in the hospital's security information system.

[0070] Optionally, the step of calculating the hash check value of the final encrypted bit sequence and encapsulating the encryption result, perturbation parameters, and check information in a structured manner specifically includes:

[0071] Calculate the 16-bit hash check value

[0072] The final encrypted result is divided into triples:

[0073] The present invention has the following beneficial effects:

[0074] 1. Patient privacy information from different sources (text, numerical values, physiological curve sampling, etc.) is uniformly mapped into a binary sequence, with necessary zero padding at the end of the sequence, and then padded in row-major order to meet the preset specifications. This processing method not only eliminates the risk of inconsistency caused by differences in various data formats and improves the robustness of subsequent algorithms to adapt to heterogeneous data sources, but also lays the physical layout foundation for chaotic permutation and sub-block perturbation based on spatial coordinates through the introduction of the matrix structure. Compared with traditional techniques that only perform bit-level operations in the linear domain, the matrixization of this scheme makes the data have visualized spatial correlation, which facilitates local and global perturbation algorithms to perform hierarchical processing based on rows, columns, and sub-blocks, thereby achieving better diffusion and obfuscation effects with the same amount of computation, and providing a solid framework for subsequent security encryption.

[0075] 2. By combining infrared temperature and environmental noise sensor data from the operating room, the collected physical quantities are mapped to deterministic perturbation seeds, thereby generating chaotic mapping parameters. This approach breaks through the traditional model that relies solely on fixed or pre-shared keys, leveraging the dynamic randomness of the physical environment to enhance the unpredictability and resistance to reproducible experiments in key generation. Simultaneously, this environment-driven parameter generation method eliminates the need to store high-value original readings in the system, reducing the exposure of sensitive information, and ensures consistency in cross-platform transmission and replay through parameter quantization and fixed-width encoding. Compared to existing technologies, this step significantly increases the difficulty for attackers to reproduce the algorithm based on historical keys or known pseudo-random number generation algorithms, thereby improving the security level of data protection from the source.

[0076] 3. The scheme employs a two-dimensional coordinate chaotic mapping function to permutate bits within the matrix, and uses a collision detection mechanism to ensure the injectivity and invertibility of the mapping. This spatial shuffling method performs a highly nonlinear rearrangement of data in the two-dimensional domain, completely dispersing adjacent bits in the same row or column across the entire domain, increasing the unpredictability of the bit sequence and the difficulty of statistical analysis. Compared to simple linear shifts or simple permutations, the chaotic permutation in this scheme utilizes environment-driven parameters to achieve diverse paths, while ensuring that the mapping process can be completed within a finite number of steps without additional backtracking logic, thus achieving good implementation efficiency in hardware or real-time computing environments. Furthermore, through global traversal and priority selection of collision candidates, this scheme ensures that any collision can quickly locate empty slots, avoiding information loss and ensuring the robustness and invertibility of the algorithm.

[0077] 4. In the permuted matrix, this scheme divides it into multiple equal-sized sub-blocks and generates a unique reversible mask based on the relative coordinates of the elements in each sub-block. Local perturbation is then achieved through an XOR operation. This fine-grained sub-block-level randomization destroys local statistical features while ensuring the reversibility of the perturbation—the original data can be recovered by XORing the same mask again. Compared to traditional global one-time XOR or single-key masking methods, position-dependent sub-block masks increase the difficulty for attackers to recover the key in known or chosen plaintext scenarios. Furthermore, the perturbation strength can be flexibly adjusted according to the sub-block size, balancing security and computational cost. In addition, adaptive boundary handling allows the algorithm to adapt to matrices of arbitrary sizes without requiring additional special design for incomplete blocks at the tail, improving versatility and ease of implementation.

[0078] 5. Following the local sub-block perturbation, this scheme further selects specific rows or columns of the entire matrix for flipping operations based on global control parameters to create a macro-scale consistent perturbation effect. This scheme spreads the impact of local perturbation to a wider range through simple row and column flipping, resulting in data distribution with both intra-block and global levels of obfuscation. Compared to schemes that rely solely on sub-block randomization, this scheme enhances the overall data mixing degree while ensuring the reversibility of the perturbation, further improving its ability to resist known-plaintext or differential analysis attacks. Furthermore, the global perturbation is based on previously environment-driven seed parameters, ensuring reproducibility and traceability, meeting the compliance requirements of the operating room for data consistency verification and auditing.

[0079] 6. After completing the multi-layer perturbation, the scheme re-linearizes the bits in the matrix using row-major order and removes the invalid zero bits previously used for matrix padding, thereby recovering the pure encrypted bit sequence. This method not only achieves lossless mapping between the two-dimensional and one-dimensional domains but also automatically removes redundant padding, ensuring that the length of the output sequence is consistent with the meaning of the original data, facilitating transparent use in subsequent hospital medical record systems or communication protocols. Compared to traditional methods that require additional recording of padding length or execution of complex backfilling logic after matrix perturbation, this scheme's unified row-major recovery and one-time removal strategy simplifies the implementation process, reduces the risk of errors, and improves overall processing efficiency.

[0080] 7. Regarding the environmental seeds, chaotic mapping parameters, sub-block sizes, and mask generation rules involved in the entire perturbation process, this solution centrally quantifies and encodes all key parameters according to a predetermined format and binds them to patient identifiers, storing them in the secure information management system. This forms the necessary "credential package" for data descrambling, ensuring accurate restoration of the encryption process within authorized permissions, and also avoids the security risks associated with scattered key management. Unlike traditional methods of distributing keys or relying on external key management services, this solution's "centralized parameter storage + reversible verification" model is self-contained, facilitates auditing and multi-level access control, reduces the overall confidentiality of the system, and provides dual protection for compliance and maintainability.

[0081] 8. Finally, the scheme generates a lightweight hash checksum for the encrypted output bit sequence and packages the encrypted data, perturbation control parameters, and checksum information into a structured triplet. This design not only enables rapid data integrity verification without descrambling, effectively detecting tampering and single-bit errors during transmission or storage, but also reduces the risk of single-point leakage by separating the data body and parameter body and employing a hierarchical storage strategy. Compared to existing schemes that only perform integrity checks during the decryption stage, this scheme's pre-structured encapsulation provides an inherent anti-tampering mechanism during data transmission, enhancing end-to-end security and simplifying the calling interface of upper-layer application systems. Attached Figure Description

[0082] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0083] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0084] Example, refer to Figure 1 A method for protecting patient privacy data in the operating room, comprising:

[0085] The patient privacy information to be protected is uniformly encoded into a binary sequence, and zeros are padded to the end of the sequence to meet the preset matrix size requirements. Then, it is filled into the matrix in row-major order.

[0086] Perturbation seeds are generated by quantizing the measured values ​​of operating room ambient temperature and noise, and the intensity of the chaotic mapping is calculated accordingly.

[0087] Two-dimensional coordinate rearrangement mapping is generated using chaotic mapping intensity, and each element in the matrix is ​​replaced to the target position according to the mapping relationship;

[0088] The permuted matrix is ​​divided into multiple sub-blocks, and an XOR operation is performed on the sub-block based on the position-related mask.

[0089] Based on the global disturbance control parameters, perform bit-flipping operations sequentially on the odd columns or even rows of the matrix;

[0090] Extract the bit sequence from the perturbation matrix in row-major order and remove the aforementioned padding zeros to obtain the final binary output;

[0091] Based on the data perturbation process, perturbation control parameters, including temperature perturbation seed, noise perturbation seed, window size parameters, mask generation parameters, and global perturbation parameters, are constructed and recorded, and these parameters are bound and stored with patient identification information.

[0092] The hash check value is calculated for the final encrypted bit sequence, and the encryption result, perturbation parameters, and check information are encapsulated in a structured manner.

[0093] This method first encodes the patient's privacy information into a binary sequence and fills it into a preset matrix dimension. Then, it generates a random perturbation seed by combining the temperature and noise measurement results of the operating room. The seed is used to drive the chaotic mapping to perform two-dimensional permutation on the matrix elements. On this basis, mask XOR and bit flip perturbation are applied in two levels - local sub-block level and global row and column level. Finally, the linear sequence is restored and the padding zeros are removed. The perturbation parameters and hash check values ​​are then encapsulated together. Through this series of steps, the problems of data encryption in an operating room environment relying on fixed keys and being vulnerable to statistical analysis or known-plaintext attacks are solved: First, mapping the original data to a matrix structure and filling it ensures the spatial coherence of global and local perturbations, so that subsequent scrambling is no longer limited to the linear domain; Second, the perturbation seed driven by the on-site environment makes each encryption unpredictable and random, eliminating the need to store highly sensitive keys for a long time and reducing the risk of key leakage; Third, the layered perturbation strategy breaks the statistical characteristics of the bit sequence while ensuring reversibility and traceability; Fourth, hash verification and structured encapsulation add integrity verification between encrypted input and output, enabling any unauthorized modifications to be detected quickly. Compared with traditional single-key or single-XOR methods, this method achieves multi-dimensional obfuscation and security at the same computational cost, improving the ability to resist differential analysis, plaintext attacks, and key replay attacks, and meeting the comprehensive requirements of real-time performance, security, and compliance.

[0094] The process of uniformly encoding the patient privacy information to be protected into a binary sequence, padding the end of the sequence with zeros to meet the preset matrix size requirements, and then filling the matrix in row-major order, specifically includes:

[0095] The patient's private information to be encrypted is uniformly encoded and converted into a binary bit sequence. Among them, D (0) It is a linear binary sequence of the initial state; The initial sequence has the k-th bit; the k-th linear index; and N is the total number of bits in the binary sequence to be protected. Uniform encoding ensures that privacy information from different sources (text, numerical values, and physiological curve sampling) forms an operable and consistent data structure at the bit level, providing standardized input for subsequent matrixing and permutation algorithms and avoiding uncontrollable behavior caused by format differences.

[0096] Select one that satisfies L 2 Let L be the smallest positive integer ≥ N. Where l is a temporary variable used to define the conditions for the value of L; It is a set of positive integers; minimizing the matrix order can reduce the number of padding bits, reduce the overhead of invalid computation, and at the same time ensure that the square matrix is ​​large enough to hold all bits, which facilitates subsequent two-dimensional coordinate transformation and uniform sub-block division;

[0097] Build length L 2 The extended bit sequence D (1) Fill any insufficient parts with 0:

[0098] in, For the extended k-th bit; zero padding realizes length normalization, so that the linear sequence corresponds one-to-one with the L×L matrix, which facilitates the operation of permutation and sub-block perturbation on the complete grid and does not require separate logic design for "endless blocks";

[0099] D in row priority (1) Reconstruct an L×L matrix M (0) , specific bits:

[0100] u,v∈{1,...,L}; where M (0) This is the initial matrix; Let u be the element in the u-th row and v-th column; (u-1)·L+v is the row-major index mapping; u is the matrix row index; v is the matrix column index; after two-dimensionalization, the data is bound to the spatial coordinates, providing the necessary geometric structure for subsequent "position-related mask" and "chaotic permutation", so that the disturbance can be diffused and confused by the structural characteristics of rows, columns, blocks, etc.

[0101] By uniformly formatting and encoding patient privacy data (including text, numerical values, physiological curves, etc.) from different sources, the resulting binary sequence is then filled into a pre-determined matrix row-wise, with any insufficient parts padded with zeros. This measure first solves the inconsistency problem caused by heterogeneous data formats: traditional encryption often only works for specific formats or single types of data, while this method, through unified encoding, connects multiple data sources, allowing subsequent matrix transformation and permutation operations to be applied indiscriminately to all privacy information, thereby improving the algorithm's versatility and the ease of system deployment. Second, by reasonably selecting the minimum matrix size and padding with zeros, the computational overhead caused by redundant padding is reduced, while ensuring that each data bit has a definite spatial location, providing a regular physical distribution for subsequent local and global perturbations. When the matrix row and column size is determined, sub-block partitioning does not require additional handling of "tail shortages," avoiding the need for special indexes or branching logic and simplifying implementation. In terms of security, matrix transformation brings data from a linear domain to a two-dimensional space, breaking the statistical regularity of traditional linear bit partitioning, making any subsequent attacks based on frequency or histogram analysis ineffective due to the lack of spatial coherence. Compared with existing technologies, this method not only solves the format and distribution problems in the preprocessing stage, but also lays a solid foundation for multi-layer perturbation, tightly coupling data input with spatial structure, and improving the robustness and security of the overall encryption scheme.

[0102] The process of generating perturbation seeds based on quantized measurements of operating room ambient temperature and noise, and calculating the intensity of the chaotic mapping accordingly, specifically includes:

[0103] The ambient temperature T in the operating room was measured using an infrared temperature sensor and an ambient noise sensor, respectively. room and noise level A room And set the perturbation seed as:

[0104] in, A temperature seed for a 16-bit domain; This is a noise seed for a 16-bit field; mod is the modulo operation; it maps real-time physical quantities to fixed-width integer seeds, providing reproducible, storable, and transmittable "on-site randomness" for all subsequent parameter and branch calculations, while avoiding privacy leaks caused by directly saving the original physical quantities;

[0105] The chaotic mapping intensity parameter α is calculated based on the perturbation seed, specifically as follows:

[0106] α determines the nonlinear amplitude and "variability" of chaotic permutation, ensuring that the permutation path is uniquely determined at the same time under the same environment, while causing different readings under different environments to lead to different paths, increasing the difficulty for attackers to reproduce.

[0107] Using the temperature and noise of the operating room environment as physical random sources, the measured values ​​are collected by sensors and quantized into a fixed-length perturbation seed, which is then used to generate chaotic mapping strength parameters. This effectively solves the problem of static or predictable random number generator seeds in traditional encryption schemes: common pseudo-random generation relies on internal algorithms and fixed initial values; once the initial values ​​are recovered, the entire key sequence can be reproduced. This method, however, fully utilizes the dynamic variability of environmental physical quantities, allowing the same algorithm to generate different parameters at different times and in different scenarios, thus improving the unpredictability of randomness. Furthermore, quantizing and storing physical quantities with fixed-width encoding not only avoids privacy leaks caused by directly storing the original measured values ​​but also ensures the reproducibility and cross-platform consistency of the perturbation parameters during decryption, unaffected by floating-point errors or transmission errors. Compared to existing technologies, this method improves security by an order of magnitude at the seed generation level. It eliminates the need for pre-shared keys and hardware security modules; as long as the physical conditions measured on-site are legal and reliable, consistent perturbation parameters can be obtained synchronously, fundamentally enhancing the system's protection against measurement interference and key recovery attacks.

[0108] The method of generating a two-dimensional coordinate rearrangement mapping using chaotic mapping intensity, which replaces each element in the matrix with the target position according to the mapping relationship, specifically includes:

[0109] Construct an empty L×L dimensional matrix M (1) And mark all positions as unoccupied; initialize the occupancy table to ensure that conflicts can be detected and unique positions can be found when replacement is written, so that the mapping process is strictly injective and reversible, avoiding overwriting and data loss;

[0110] For M (0) The sequence O = {(u,v)|u = 1...L; v = 1...L} is determined by row priority; the fixed traversal order ensures that the algorithm produces a definite output sequence under the same input and parameters, avoids inconsistencies caused by traversal differences in different implementations, and ensures engineering reproducibility and legal verifiability.

[0111] For any Set the two-dimensional chaotic mapping Φ:

[0112]

[0113]

[0114] Wherein, Φ is the chaotic mapping function from two-dimensional coordinates to target coordinates; u′ and v′ are the mapped coordinates; Φ is the core permutation generator, which maps the regular grid to the perturbed grid, forming a highly complex spatial scramble, reducing the predictability of bit sequences and the risk of statistical feature leakage;

[0115] The construction length is L 2The target candidate sequence, let δ u (m) = m mod L, m∈{0,1,...,L 2 -1};where, δ u (m) represents the offset of the m-th candidate in the row direction; δ v (m) represents the offset of the m-th candidate in the column direction; m is the candidate position enumeration counter, which is used to generate candidate coordinates in a definite and full-domain order when a target position conflict occurs, until the first unoccupied position is found.

[0116] And generate candidate coordinates m at a time:

[0117] (u (m) ,v (m) )=[(u′+δ u (m))mod L+1,(v′+δ v (m))mod L+1];where, (u (m) ,v (m) ) represents the m-th candidate position;

[0118] Starting from m=0, check sequentially and take the first unoccupied position. Place: Where, m * This is the first unoccupied candidate index found. To perform a matrix assignment operation, The bits are written to the target location;

[0119] and position Marked as occupied;

[0120] The collision handling ensures that any collision can find a unique empty space within a finite number of steps, without losing data or causing overwriting; the candidate overwrite over the entire domain gives the algorithm deterministic termination, enhancing the robustness and reversibility of the permutation.

[0121] Using chaotic mapping parameters calculated from the in-situ perturbation seed, the coordinates of each element in the matrix are nonlinearly permuted, and a collision detection mechanism ensures the uniqueness and reversibility of the write operation. This process resolves the contradiction between complexity and reversibility in simple row and column permutations or cyclic shifts: traditional shift methods typically only change the position of rows or columns, making them easily identifiable by frequency analysis and pattern matching; while this method, based on chaos theory, generates highly nonlinear coordinate mappings, causing previously adjacent or grouped data to be globally scattered, increasing the difficulty of statistical attacks. Simultaneously, collision detection and candidate position enumeration ensure that when the target coordinates are occupied, the next empty space can be quickly found, avoiding data loss and overwriting, and maintaining a strict one-to-one mapping relationship. Compared to existing technologies that require additional recording of mapping tables or backoff logic, this scheme tightly integrates the mapping process with parameter generation; only the perturbation seed needs to be recorded to accurately reconstruct the mapping path during the decryption stage, without relying on external index structures. This method achieves efficient and reversible spatial permutation without increasing storage burden, providing a uniformly distributed input matrix for subsequent local and global perturbations.

[0122] The step of dividing the permuted matrix into multiple sub-blocks and performing an XOR operation on the content of each sub-block based on a position-related mask specifically includes:

[0123] Take odd numbers Set the number of rows and columns for the sub-block as follows: Where G is the number of sub-blocks in the row direction; J is the number of sub-blocks in the column direction;

[0124] The actual row and column dimensions of the (g,h)th sub-block are:

[0125] r u (g)=min{r,Lr·(g-1)},r v (h) = min{r, Lr·(h-1)}; where g ∈ {1, ..., G} is the block number of the sub-block in the row direction; h ∈ {1, ..., J} is the block number of the sub-block in the column direction; r u (g) gives the actual height of the g-th row sub-block in the row direction; r v (h) gives the actual width of the h-th column sub-block in the column direction;

[0126] Its top-left corner coordinates: (u s (g),v s (h))=(r·(g-1)+1,r·(h-1)+1); where, u s (g), v s (g) Give the global coordinates of the top left corner of the sub-block in row g and column h, respectively;

[0127] The sub-blocks are divided into sub-blocks to provide a scope for subsequent "local and position-dependent" masks, so that the perturbation not only spreads globally, but also forms a fine-grained random structure within the block, balancing security and controllability; the boundary adaptation ensures that any combination of L and r can be executed.

[0128] Let C be a constant:

[0129]

[0130] Set a mask bit for any offset (s,t) within this sub-block:

[0131] w g,h (s,t)=((u s (g)+s)·(v s (h)+t)+C)mod 2; where s is the row offset relative to the top left corner of the sub-block, and its value ranges from [0, r]. u (g)-1]; t is the column offset within the sub-block relative to the top left corner, with a value range of [0, r]. v [(h)-1];w g,h (s,t) gives the position mask bits located within the (g,h)th sub-block at offset (s,t);

[0132] Construct an L×L dimensional empty matrix M (2) Specifically, it represents the matrix after XORing the sub-block level mask;

[0133] The mask is independent of the data value and is determined entirely by parameters and coordinates, ensuring that the perturbation is reversible; different masks are generated for different sub-blocks and different positions, increasing the uncertainty at the level of detail and reducing the risk of local patterns being identified.

[0134] Let (u,v) = (u s (g)+s,v s Given (h)+t), apply a change to the elements within the domain (s,t), while keeping the elements outside the domain unchanged, and obtain... in, Let M be a matrix (2) The bit at position (u,v); For bitwise XOR;

[0135] The bitwise XOR operation ensures that the perturbation is "self-reversible" (it can be restored by applying the same mask again) and injects sub-block-level randomness into the data, which breaks down local statistical features and forms the first layer of strong obfuscation of privacy information.

[0136] After the global permutation, this scheme further divides the matrix into multiple sub-blocks of the same or approximately the same size. Within each sub-block, a reversible mask is generated based on positional correlation rules, and the data within the sub-block is XORed bitwise. This approach overcomes the problem of local pattern remnants in simple global permutations: when the matrix is ​​large, relying solely on permutations may still retain the relative nearest neighbor characteristics of the original bits in a local area, providing a breakthrough for differential analysis; however, this method injects fine-grained randomness into each local region through sub-block-level mask perturbation, completely destroying the original correlation of all bits within the sub-block. Furthermore, each mask is only related to its position coordinates and pre-calculated parameters, without relying on any external key, ensuring the reversibility of the perturbation—the original sub-block data can be restored by reapplying the same mask. Unlike traditional one-time masks or global XOR, this scheme's block-based XOR strategy strikes a balance between security and controllability, allowing for flexible adjustment of the sub-block size according to performance or security requirements, achieving different levels of perturbation strength. This method enhances the concealment effect of local statistical features and improves the ability to resist known plaintext and local pattern recognition attacks.

[0137] The step of performing bit-flipping operations on the odd columns or even rows of the matrix sequentially according to the global disturbance control parameters specifically includes:

[0138] Construct an L×L dimensional empty matrix M (3) Specifically, it represents the matrix after applying a global row or column perturbation;

[0139] Calculate the global branch bit Λ:

[0140] Λ condenses the field parameters into a global binary switch, enabling the system to choose between two global flip strategies, "row or column," during runtime. This increases the path uncertainty for attackers while maintaining parameter reproducibility, facilitating descrambling.

[0141] If Λ = 0, execute the following for all cells in the odd-numbered sequence v mod 2 = 1. otherwise in, Let M be a matrix (3) Elements in (u,v);

[0142] If Λ=1, execute the following for all units in the even-numbered rows where u mod 2=0. otherwise,

[0143] Based on the local perturbation of sub-blocks, a structured global flip is applied to create consistent perturbation in the row or column direction, further spreading the bit flip to a larger scale and improving the security redundancy against chosen plaintext / known plaintext.

[0144] Building upon sub-block-level masking perturbation, this scheme introduces a global row or column-oriented structured flip operation, reversing or swapping all bits in a specified row or column according to control parameters. This operation addresses the macroscopic arrangement patterns that may remain when perturbations are applied only within a small scope: if only local perturbation is performed, an attacker can search for residual row and column distribution patterns from a global perspective when comparing parts of the plaintext and ciphertext; however, with global flipping, the original row and column distribution is scrambled again, ensuring that any local recovery is difficult to reconstruct the overall distribution. The global perturbation strategy complements the sub-block perturbation, spreading the randomness introduced by sub-block perturbation to a wider range while retaining reversibility and one-time parameter-driven characteristics. Unlike existing schemes that typically apply perturbation only in one dimension, this method creates a superimposed perturbation effect at two levels—local and macroscopic—improving protection against wide-area differential analysis and multidimensional statistical attacks. Furthermore, it requires only a small amount of additional parameter identification, and the original arrangement can be restored simply by performing the same flip operation in reverse order during decryption.

[0145] The process of extracting the bit sequence from the perturbation matrix in row-major order and removing the aforementioned padding zeros to obtain the final binary output specifically includes:

[0146] Extract M in row-major order (3) All bits in the array are a linear sequence:

[0147] k = (u-1)·L + v; where, D (2) This is the linearized result after local and global perturbations; For the k-th bit after linearization; k = (u-1)·L+v bit row-major flattening formula; completes the lossless mapping from the two-dimensional domain back to the one-dimensional domain, so that subsequent storage, transmission and verification operations can be performed in a unified linear bit space, while maintaining the clear reversibility relationship;

[0148] Extract the first N bits as the final encrypted output sequence: Among them, D (enc) The output is a linear sequence of bit encryption; zero padding added to satisfy the matrix dimension is removed to restore the actual data length, ensuring that the length semantics and field boundaries of the business side (such as medical record system) are not violated.

[0149] After multiple perturbations, this scheme linearly extracts the matrix into a one-dimensional bit sequence in row-major order and removes the previously padded zeros, resulting in a pure encrypted output sequence. This design solves the conversion compatibility problem between matrix transformation and linear format while ensuring reversible recovery: traditional matrix perturbation schemes often face the problem of not being able to accurately distinguish the padded bits from the original data bits during recovery. This method, on the one hand, unifies the extraction order to ensure a stable and reliable mapping relationship from the two-dimensional field to the one-dimensional field; on the other hand, it accurately removes the padded zeros before storage or transmission, restoring the original data length and avoiding the subsequent processing complexity caused by the mismatch between the ciphertext length and field boundaries. Unlike existing schemes that require additional recording of the padded length or design of complex backfilling logic, this method has a built-in padded and removal mechanism, reducing the difficulty of managing additional metadata and lowering the risk of errors. At the same time, after completing the linearization and removal operations in one step, the encrypted output can be directly used for database or network transmission, simplifying the system integration interface and improving the deployment efficiency of terminal applications.

[0150] The process of constructing and recording perturbation control parameters, including temperature perturbation seeds, noise perturbation seeds, window size parameters, mask generation parameters, and global perturbation parameters, based on the data perturbation process, and binding and storing these parameters with patient identification information, specifically includes:

[0151] Constructing the perturbation parameter tuple The key parameters are packaged together to form a "descrambling credential" that corresponds one-to-one with the encryption result, providing the necessary and sufficient conditions for subsequent recovery under legitimate permissions without exposing the original environmental readings;

[0152] Define and execute the following mappings:

[0153] S701, 8-bit alpha quantization:

[0154] S702, Seed directly uses 16 bits: already configured

[0155] S703, Fixed bit width of each component:

[0156] N [32-bit], L [16-bit], r [16-bit], a [8-bit] [16-bit], [16-bit];

[0157] S704. Let the bitwise XOR mask constant vector be Ξ:

[0158] Ξ=(2 32 -1,2 16 -1,2 16 -1,2 8 -1,2 16 -1,2 16 -1);

[0159] Component-wise definition in, Bit parameter encoding result;

[0160] Fixed bit width and bit-reversible encoding ensure that parameters are transmitted without distortion between different systems / platforms; the original parameters can be recovered by XORing the same Ξ again, ensuring that the parameter packet can be verified without relying on third-party encryption mechanisms;

[0161] Will It is bound to the patient's data number and stored in the hospital's security information system; it achieves a unique correspondence between parameters and specific data objects, which facilitates auditing, traceability and access control; storing the parameter package separately can reduce the risk of ciphertext and parameters being exposed in the same database.

[0162] Key parameters throughout the perturbation process—including environmental perturbation seeds, sub-block partitioning sizes, mask generation rules, and global flip control identifiers—are centrally constructed into a parameter package and bound to patient identification information, then stored uniformly in the security management system. This design addresses the complexity and security risks arising from the dispersed management of parameters in multi-level perturbation processes: traditional solutions often scatter keys or parameters across different modules, resulting in high management and synchronization costs, and leakage at any stage can lead to a collapse of overall security; while this method's centralized packaging and binding storage strategy not only simplifies access control and auditing processes but also reduces the risk of "one-database leaks" by storing parameter packages separately from encrypted data. Furthermore, all parameters employ a fixed-width reversible encoding method, allowing integrity verification without additional encryption, and the original perturbation process can be restored using the same algorithm during legitimate decryption, ensuring the accuracy and traceability of data recovery.

[0163] The calculation of the hash check value for the final encrypted bit sequence, and the structured encapsulation of the encryption result, perturbation parameters, and check information, specifically includes:

[0164] Calculate the 16-bit hash check value A lightweight integrity verification method is provided, which can detect ciphertext tampering or transmission errors (such as single bit flips or missing segments) without descrambling, and can be used as a fast consistency verification method.

[0165] The final encrypted result is divided into triples: Unified packaging simplifies storage and transmission interfaces; separate parameter packets and data bodies can employ different security strategies (such as different security levels or different libraries) at the access control layer, reducing the systemic risk of single-point leakage.

[0166] A lightweight hash checksum is calculated on the final output encrypted bit sequence, and the encrypted data, perturbation parameter packet, and checksum information are uniformly encapsulated into a triple for integrity verification at any stage of transmission and storage. This design solves the problem of traditional integrity checks being performed only after decryption or relying on external digital signature services: the current solution completes checksum generation and encapsulation during the ciphertext generation stage, enabling rapid detection of any tampering or transmission errors without decryption, reducing system call complexity; simultaneously, the structured encapsulation separates the data body and parameter packet but carries them together, facilitating interface calls and supporting access control for different security levels. Unlike existing solutions that require additional signature certificates or third-party verification services, this solution achieves end-to-end data integrity protection using a simple hash algorithm, reducing deployment costs and operational complexity, and improving the overall system response speed and security.

[0167] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0168] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for protecting patient privacy data in the operating room, characterized in that, include: The patient privacy information to be protected is uniformly encoded into a binary sequence, and zeros are padded to the end of the sequence to meet the preset matrix size requirements. Then, it is filled into the matrix in row-major order. Perturbation seeds are generated by quantizing the measured values ​​of operating room ambient temperature and noise, and the intensity of the chaotic mapping is calculated accordingly. Two-dimensional coordinate rearrangement mapping is generated using chaotic mapping intensity, and each element in the matrix is ​​replaced to the target position according to the mapping relationship; The permuted matrix is ​​divided into multiple sub-blocks, and an XOR operation is performed on the sub-block based on the position-related mask. Based on the global disturbance control parameters, perform bit-flipping operations sequentially on the odd columns or even rows of the matrix; Extract the bit sequence from the perturbation matrix in row-major order and remove the aforementioned padding zeros to obtain the final binary output; Based on the data perturbation process, perturbation control parameters, including temperature perturbation seed, noise perturbation seed, window size parameters, mask generation parameters, and global perturbation parameters, are constructed and recorded, and these parameters are bound and stored with patient identification information. The hash check value is calculated for the final encrypted bit sequence, and the encryption result, perturbation parameters, and check information are encapsulated in a structured manner.

2. The method for protecting patient privacy data in the operating room according to claim 1, characterized in that, The process of uniformly encoding the patient privacy information to be protected into a binary sequence, padding the end of the sequence with zeros to meet the preset matrix size requirements, and then filling the matrix in row-major order, specifically includes: The patient's private information to be encrypted is uniformly encoded and converted into a binary bit sequence. Among them, D (0) It is a linear binary sequence of the initial state; The initial sequence is represented by the k-th bit; the k-th bit is the linear index; N is the total number of bits in the binary sequence to be protected. Select one that satisfies L 2 Let L be the smallest positive integer ≥ N. Where l is a temporary variable used to define the conditions for the value of L; It is a set of positive integers; Build length L 2 The extended bit sequence D (1) Fill any insufficient parts with 0: in, For the k-th bit after expansion; D in row priority (1) Reconstruct an L×L matrix M (0) , specific bits: Among them, M (0) This is the initial matrix; Let be the element in row u and column v; (u-1)·L+v is the row-major index mapping; u is the matrix row index; v is the matrix column index.

3. A method for protecting patient privacy data in an operating room according to claim 2, characterized in that, The process of generating perturbation seeds based on quantized measurements of operating room ambient temperature and noise, and calculating the intensity of the chaotic mapping accordingly, specifically includes: The ambient temperature T in the operating room was measured using an infrared temperature sensor and an ambient noise sensor, respectively. room and noise level A room And set the perturbation seed as: in, A temperature seed for a 16-bit domain; This is the noise seed for a 16-bit field; mod is the modulo operation. The chaotic mapping intensity parameter α is calculated based on the perturbation seed, specifically as follows:

4. A method for protecting patient privacy data in an operating room according to claim 3, characterized in that, The method of generating a two-dimensional coordinate rearrangement mapping using chaotic mapping intensity, which replaces each element in the matrix with the target position according to the mapping relationship, specifically includes: Construct an empty L×L dimensional matrix M (1) And mark all locations as unoccupied; For M (0) Determine the sequence in row order For any Set the two-dimensional chaotic mapping Φ: Where Φ is the chaotic mapping function from two-dimensional coordinates to target coordinates; u′ and v′ are the mapped coordinates; The construction length is L 2 The target candidate sequence, let δ u (m) = m mod L, Where, δ u (m) represents the offset of the m-th candidate in the row direction; δ v (m) represents the offset of the m-th candidate in the column direction; m is the candidate position enumeration counter; And generate candidate coordinates m at a time: (u (m) ,v (m) )=[(u′+δ u (m))mod L+1,(v′+δ v (m))mod L+1];where, (u (m) ,v (m) ) represents the m-th candidate position; Starting from m=0, check sequentially and take the first unoccupied position. Place: Where, m * This is the first unoccupied candidate index found. To perform a matrix assignment operation, The bits are written to the target location; and position Marked as occupied.

5. A method for protecting patient privacy data in an operating room according to claim 4, characterized in that, The step of dividing the permuted matrix into multiple sub-blocks and performing an XOR operation on the content of each sub-block based on a position-related mask specifically includes: Take odd numbers Set the number of rows and columns for the sub-block as follows: Where G is the number of sub-blocks in the row direction; J is the number of sub-blocks in the column direction; The actual row and column dimensions of the (g,h)th sub-block are: r u (g)=min{r,Lr·(g-1)},r v (h) = min{r, Lr·(h-1)}; where g ∈ {1, ..., G} is the block number of the sub-block in the row direction; h ∈ {1, ..., J} is the block number of the sub-block in the column direction; r u (g) gives the actual height of the g-th row sub-block in the row direction; r v (h) gives the actual width of the h-th column sub-block in the column direction; Its top-left corner coordinates: (u s (g),v s (h))=(r·(g-1)+1,r·(h-1)+1); where, u s (g), v s (g) Give the global coordinates of the top left corner of the sub-block in row g and column h, respectively; Let C be a constant: Set a mask bit for any offset (s,t) within this sub-block: w g,h (s,t)=((u s (g)+s)·(v s (h)+t)+C)mod 2; where s is the row offset relative to the top left corner of the sub-block, and its value ranges from [0, r]. u (g)-1]; t is the column offset within the sub-block relative to the top left corner, with a value range of [0, r]. v [(h)-1];w g,h (s,t) gives the position mask bits located within the (g,h)th sub-block at offset (s,t); Construct an L×L dimensional empty matrix M (2) Specifically, it represents the matrix after XORing the sub-block level mask; Let (u,v) = (u s (g)+s,v s Given (h)+t), apply a change to the elements within the domain (s,t), while keeping the elements outside the domain unchanged, and obtain... in, Let M be a matrix (2) The bit at position (u,v); This is a bitwise XOR operation.

6. A method for protecting patient privacy data in an operating room according to claim 5, characterized in that, The step of performing bit-flipping operations on the odd columns or even rows of the matrix sequentially according to the global disturbance control parameters specifically includes: Construct an L×L dimensional empty matrix M (3) Specifically, this represents the matrix after applying a global row or column perturbation; Calculate the global branch bit Λ: If Λ = 0, execute the following for all cells in the odd-numbered sequence v mod 2 = 1. otherwise in, Let M be a matrix (3) Elements in (u,v); If Λ=1, execute the following for all units in the even-numbered rows where u mod 2=0. otherwise, 7. A method for protecting patient privacy data in an operating room according to claim 6, characterized in that, The process of extracting the bit sequence from the perturbation matrix in row-major order and removing the aforementioned padding zeros to obtain the final binary output specifically includes: Extract M in row-major order (3) All bits in the array are a linear sequence: Among them, D (2) This is the linearized result after local and global perturbations; For the k-th bit after linearization; k = (u-1)·L+v bit row-major flattening formula; Extract the first N bits as the final encrypted output sequence: Among them, D (enc) A linear sequence of bit-encrypted output.

8. A method for protecting patient privacy data in an operating room according to claim 7, characterized in that, The process of constructing and recording perturbation control parameters, including temperature perturbation seeds, noise perturbation seeds, window size parameters, mask generation parameters, and global perturbation parameters, based on the data perturbation process, and binding and storing these parameters with patient identification information, specifically includes: Constructing the perturbation parameter tuple Define and execute the following mappings: S701, 8-bit alpha quantization: S702, Seed directly uses 16 bits: already configured S703, Fixed bit width of each component: N [32-bit], L [16-bit], r [16-bit], a [8-bit] [16-bit], [16-bit]; S704. Let the bitwise XOR mask constant vector be Ξ: Ξ=(2 32 -1, 2 16 -1, 2 16 -1, 2 8 -1, 2 16 -1, 2 16 -1); Component-wise definition in, Bit parameter encoding result; Will It is linked to the patient's data ID and stored in the hospital's security information system.

9. A method for protecting patient privacy data in an operating room according to claim 8, characterized in that, The calculation of the hash check value for the final encrypted bit sequence, and the structured encapsulation of the encryption result, perturbation parameters, and check information, specifically includes: Calculate the 16-bit hash check value The final encrypted result is divided into triples:

Images