Privacy protection medical data analysis method based on block chain and federal learning

Abstract

The invention relates to the field of medical data analysis, in particular to a privacy protection medical data analysis method based on a block chain and federal learning. Comprising the steps of collecting and preprocessing original medical data, and obtaining comprehensive features based on the preprocessed original medical data; training the medical model to obtain a prediction result; evaluating and optimizing the medical model, generating a medical model parameter update value, performing encryption and signature processing on the medical model parameter update value, and generating a signature; performing signature verification on the encryption parameter update values and the signature, integrating the encryption parameter update values to obtain an aggregation parameter update value, and encrypting and decrypting the aggregation parameter update value to obtain a final parameter update value; and optimizing the medical model based on the final parameter update value. The problem that federal learning may expose individual privacy during model training is solved; the block chain lacks security and expandability; and when the number of participants is huge and the medical data volume is huge, the calculation and communication overhead of model aggregation is high, and the efficiency and speed of medical data analysis are influenced.

Description

Technical Field

[0001] The present invention relates to the field of medical data analysis, and in particular to a privacy-preserving medical data analysis method based on blockchain and federated learning. Background Art

[0002] In the medical field, privacy protection and security analysis of medical data are crucial. Traditional medical data analysis methods usually involve the centralized storage of sensitive medical data, which may pose the risk of medical data leakage and privacy infringement. In order to solve these problems, federated learning and blockchain technology came into being and began to play a role in the field of medical data analysis, providing a new solution for the privacy protection and security sharing of medical data. Federated learning enables participants to jointly build medical models without sharing original medical data, while blockchain technology ensures the transparency and security of medical data. Combining the two technologies, privacy protection and security analysis of medical data can be achieved, bringing a more reliable and efficient data management method to the medical industry.

[0003] There are still some technical problems with existing federated learning and blockchain technologies: in terms of medical data privacy protection, although federated learning can train models in a distributed environment, individual privacy may still be exposed when the model is aggregated; in terms of medical data security sharing, although blockchain provides a distributed and tamper-proof way to store medical data, blockchain still lacks security; in terms of medical data analysis efficiency, although federated learning can train models in a distributed manner, when the number of participants is large or the amount of medical data is huge, the computational and communication overhead of model aggregation may become very high, affecting the efficiency and speed of medical data analysis. Therefore, improving the security, efficiency, and accuracy of medical data analysis remains a technical problem that needs to be solved urgently. Summary of the invention

[0004] The present invention proposes a privacy-preserving medical data analysis method based on blockchain and federated learning. The method introduces adaptive feature fusion technology, a dynamic blockchain size adjustment algorithm and a secure fusion aggregation algorithm to solve the problem that in terms of medical data privacy protection, although federated learning can perform model training in a distributed environment, individual privacy may still be exposed during model aggregation; in terms of medical data security sharing, although blockchain provides a distributed and tamper-proof medical data storage method, blockchain still lacks security; in terms of medical data analysis efficiency, although federated learning can train models in a distributed manner, when the number of participants is large or the amount of medical data is huge, the computational and communication overhead of model aggregation may become very high, affecting the efficiency and speed of medical data analysis. Technical problems.

[0005] A privacy-preserving medical data analysis method based on blockchain and federated learning includes the following steps:

[0006] S1: Use the dynamic blockchain resizing algorithm to monitor the data transmission load in real time and dynamically adjust the block size and the number of shards; collect and preprocess the raw medical data in medical institutions, and obtain comprehensive features based on the preprocessed raw medical data through adaptive feature fusion technology;

[0007] S2: Establish and train a medical model using comprehensive features and multi-layer hidden layer operations to obtain prediction results; evaluate and optimize the medical model based on the prediction results to generate medical model parameter update values; encrypt the medical model parameter update values ​​to generate encrypted parameter update values; sign the encrypted parameter update values ​​to generate signatures; verify the encrypted parameter update values ​​and signatures, and integrate the encrypted parameter update values ​​to obtain aggregated parameter update values; encrypt and then decrypt the aggregated parameter update values ​​to obtain aggregated parameter update values, and decrypt the aggregated parameter update values ​​to obtain final parameter update values; and optimize the medical model based on the final parameter update values.

[0008] Preferably, the S1 specifically includes:

[0009] Calculate the data transmission load at time t using the following formula:

[0010]

[0011] Among them, L t is the data transmission load at time t, N t-k is the number of transmissions of the encryption parameter update value within time tk, S t-k is the amount of data transmitted for the encryption parameter update value at time tk, F t-k is the transmission frequency of the encryption parameter update value at time tk, α, β, γ are weight coefficients used to adjust the impact of different load indicators on the data transmission load, and v is the size of the sliding window.

[0012] Preferably, the S1 specifically includes:

[0013] The original medical data includes structured medical original data and image medical original data; the preprocessed original medical data includes structured medical data and image medical data; structured medical data features and image medical data features are extracted from the structured medical data and image medical data respectively.

[0014] Preferably, the S1 specifically includes:

[0015] Adaptive feature fusion technology splices structured medical data features and image medical data features, and uses the attention mechanism to assign importance weights to the spliced ​​features.

[0016] Preferably, the S1 specifically includes:

[0017] The splicing features are weightedly fused based on the importance weights to obtain comprehensive features.

[0018] Preferably, the S2 specifically includes:

[0019] The medical model transforms the comprehensive features through multiple hidden layers and uses the output layer to output the prediction results; the calculation process of the hidden layer is as follows:

[0020] The calculation formula for the first hidden layer is:

[0021]

[0022] Among them, H (1) is the output of the first hidden layer, is the weight matrix of the first hidden layer, is the bias vector of the first hidden layer, is a comprehensive feature, ReLU is an activation function;

[0023] The calculation formula for the second to Xth hidden layers is:

[0024]

[0025] Among them, H (x) is the output of the x-th hidden layer, and is the output of the x-1-th hidden layer. (x-1) The result after the xth hidden layer is: is the weight matrix of the x-th hidden layer, is the bias vector of the xth hidden layer, and m is the weight matrix of the xth hidden layer The total number of elements in represents the number of weights of the x-th hidden layer. yes The A weight.

[0026] Preferably, the S2 specifically includes:

[0027] The calculation formula of the output layer is:

[0028]

[0029] Among them, H (X) is the output of the Xth hidden layer, where X represents the total number of hidden layers. is the weight matrix of the output layer, is the bias vector of the output layer, and q is the weight matrix of the output layer The total number of elements in , yes The pth weight in It is the prediction result when the input of the medical model is comprehensive features.

[0030] Preferably, the S2 specifically includes:

[0031] The secure fusion aggregation algorithm integrates the encryption parameter update values ​​of different medical nodes and introduces a smoothing factor to calculate the aggregate parameter update value; the calculation formula is as follows:

[0032]

[0033] in, is the aggregation parameter update value, M is the number of participating medical nodes, E(ΔW η ) is the encryption parameter update value of the medical node η, is the updated value of the aggregation parameter in the previous round; ω is the smoothing factor.

[0034] The beneficial effects of the technical solution of the embodiment of the present invention are:

[0035] 1. The constructed medical model can efficiently integrate the features of structured medical data and image medical data, enhance the predictive ability of the medical model, and weightedly fuse different features through the attention mechanism. The medical model can automatically identify and utilize important features, thereby improving the accuracy and robustness of the prediction results and providing more accurate and reliable support for clinical decision-making.

[0036] 2. Through the secure fusion aggregation algorithm, the original medical data of each medical institution remains private during the transmission and aggregation process. The secure fusion aggregation algorithm ensures that the medical data is not leaked or tampered with during the transmission process, allowing medical institutions to participate in training and optimization without sharing patient medical data, fully protecting patient privacy.

[0037] 3. The dynamic blockchain size adjustment algorithm dynamically adjusts the block size and the number of shards according to the medical data transmission load to ensure the efficient operation of the blockchain under high load conditions. The application of sharding technology improves the parallel processing capability of the blockchain, avoids the bottleneck problem of a single chain, and ensures high throughput and low latency medical data processing, making medical data analysis and medical model training more efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a flow chart of a privacy-preserving medical data analysis method based on blockchain and federated learning described in the present invention. DETAILED DESCRIPTION

[0039] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0040] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0041] The following is a detailed description of a specific scheme of a privacy-preserving medical data analysis method based on blockchain and federated learning provided by the present invention in conjunction with the accompanying drawings.

[0042] See attached Figure 1 , which shows a flow chart of a privacy-preserving medical data analysis method based on blockchain and federated learning provided by an embodiment of the present invention, the method comprising the following steps:

[0043] S1: Use the dynamic blockchain resizing algorithm to monitor the data transmission load in real time and dynamically adjust the block size and the number of shards; collect and preprocess the raw medical data in medical institutions, and obtain comprehensive features based on the preprocessed raw medical data through adaptive feature fusion technology;

[0044] In blockchain, the fixed block size determines the amount of data that the block can hold, and the number of shards determines the number of blocks that can be processed and transmitted in parallel. When the block size is fixed and the number of shards is limited, a large number of encryption parameter update value transmission requests will lead to slower data transmission speed and reduced network transmission efficiency.

[0045] Use a dynamic blockchain size adjustment algorithm to monitor the data transmission load in real time. The data transmission load includes the number of transmissions, the amount of data transmitted, and the transmission frequency of the encrypted parameter update values ​​sent and received by the medical node, and dynamically adjust the block size and the number of shards. When the data transmission load is high, increase the block size and the number of shards to ensure that more encrypted parameter update values ​​can be transmitted and processed in parallel, improve the processing speed of the encrypted parameter update values ​​and the network transmission efficiency, reduce transmission delays, and avoid network congestion; when the data transmission load is low, reduce the block size and the number of shards to reduce the resource consumption of the blockchain and federated learning system, and optimize storage and processing efficiency.

[0046] Calculate the data transmission load at time t using the following formula:

[0047]

[0048] Among them, L t is the data transmission load at time t, N t-k is the number of transmissions of the encryption parameter update value within time tk, S t-k is the amount of data transmitted for the encryption parameter update value at time tk, F t-k is the transmission frequency of the encryption parameter update value at time tk, α, β, γ are weight coefficients used to adjust the impact of different load indicators on the data transmission load, v is the size of the sliding window, which is used to smooth the load calculation, and α, β, γ, v can be specifically designed according to the specific implementation scenario.

[0049] The formula for dynamically adjusting block size is as follows:

[0050]

[0051] in, is the initial result of the block size at time t, B t is the final result of the block size at time t, B t-1 is the final result of the block size at time t-1, B min is the minimum value of the block size, indicating the block size under the lowest data transmission load, B max is the maximum block size, indicating the block size under the highest data transmission load, ε is the adjustment coefficient, used to control the adjustment speed, L target is the target data transmission load, which represents the ideal data transmission load level that blockchain and federated learning systems hope to achieve. target , B min , B max ,ε can be designed according to the specific implementation scenario.

[0052] The formula for dynamically adjusting the number of shards is as follows:

[0053]

[0054]

[0055] Among them, P t is the actual processing capacity at time t, N t is the number of transmissions of updated encryption parameter values ​​within time t, T t is the time required to process all the data within time t. is the moving average processing capacity at time t, The value is the actual processing capacity measured at the initial time point t = 0, λ is the sliding average coefficient, U t is the number of shards at time t, N max is the maximum amount of data that a shard can handle, P max is the maximum data processing capacity of the blockchain, Pmax ,λ,N max It can be set according to the specific implementation scenario.

[0056] Medical nodes represent medical institutions that actually provide medical services. They act as independent participants in blockchain and federated learning systems and collect original medical data within medical institutions. Original medical data include: structured medical original data in hospital information systems and electronic medical record systems, and image medical original data in medical imaging devices. Structured medical original data refers to medical data organized in a certain format and rules, with clear fields and types for easy processing and analysis. Common structured medical original data include patients' electronic medical records, laboratory test results, etc. Image medical original data refers to image data obtained through various medical imaging devices in the form of pictures. Common image medical original data include X-rays, CT scans, ultrasound images, etc.

[0057] First, the original medical data is preprocessed, and the preprocessing steps include data denoising, data standardization, and data format conversion. The structured medical raw data is preprocessed to obtain structured medical data, and the image medical raw data is preprocessed to obtain image medical data. For structured medical data, statistical methods such as calculating the mean, standard deviation, correlation coefficient, etc. are used to extract the structured medical data features f struct , the specific statistical method can be set according to the specific implementation scenario; for image medical data, convolutional neural network is used to extract image medical data features f ing The data denoising, data standardization, data format conversion, statistical methods and convolutional neural networks are all existing technologies and will not be elaborated here.

[0058] Furthermore, the medical node uses adaptive feature fusion technology to obtain comprehensive features. Adaptive feature fusion technology achieves maximum data utilization at the data level by splicing structured medical data features and image medical data features, and using the attention mechanism to assign importance weights to each feature, thereby improving the predictive ability and robustness of subsequent medical models.

[0059] The comprehensive features are obtained by splicing the structured medical data features and the image medical data features and weighted fusion using the attention mechanism, which is used for subsequent medical model training and prediction. The specific process is as follows:

[0060] The medical node first splices the structured medical data features and the image medical data features to obtain the spliced ​​features. The expression is as follows:

[0061]

[0062] Then use the attention mechanism to distribute the concatenated features Importance weights:

[0063]

[0064] Among them, a is the attention weight vector, which represents the concatenation feature is a collection of importance weights of each feature in , and softmax is the activation function. is the attention weight matrix, is the attention bias vector, the attention weight matrix and the attention bias vector The derivation process involves deep learning technology, including neural network training, back propagation algorithm and gradient descent optimization. The deep learning technology is existing technology and will not be described in detail here.

[0065] According to the calculated attention weight a, the concatenated features For weighted fusion, the formula is as follows:

[0066]

[0067] in, It is a comprehensive feature. is the attention weight vector a The importance weight of the feature, It is the first features, n is the concatenation feature The number of features in Represents the first The importance weights of each feature are exponentially calculated.

[0068] s2: Establish and train a medical model using comprehensive features and multi-layer hidden layer operations to obtain prediction results; evaluate and optimize the medical model based on the prediction results to generate medical model parameter update values; encrypt the medical model parameter update values ​​to generate encrypted parameter update values; sign the encrypted parameter update values ​​to generate signatures; verify the encrypted parameter update values ​​and signatures, and integrate the encrypted parameter update values ​​to obtain aggregated parameter update values; encrypt and then decrypt the aggregated parameter update values ​​to obtain aggregated parameter update values, and decrypt the aggregated parameter update values ​​to obtain final parameter update values; and optimize the medical model based on the final parameter update values.

[0069] Medical nodes use comprehensive features to build and train medical models. Medical models analyze input data to provide accurate prediction results. The prediction results are specific values, such as disease risk scores, patient recovery probability, or health indicators. The prediction results are used to assess the health status of patients, assist doctors in diagnosis and develop personalized treatment plans, thereby improving the quality and efficiency of medical services. The specific process is as follows:

[0070] In the process of building a medical model, the medical node will integrate the comprehensive features As input, feature transformation is performed through multiple hidden layers, and the predicted output of the medical model is calculated in the output layer. It is stipulated that there are X hidden layers, and the X value can be set according to the specific implementation scenario. The calculation process of the hidden layer is as follows:

[0071] The calculation formula for the first hidden layer is:

[0072]

[0073] Among them, H (1) is the output of the first hidden layer, The result after the first hidden layer is is the weight matrix of the first hidden layer, is the bias vector of the first hidden layer, and ReLU is the activation function.

[0074] The calculation formula for the second to Xth hidden layers is:

[0075]

[0076] Among them, H (x) is the output of the x-th hidden layer, and is the output of the x-1-th hidden layer. (x-1) The result after the xth hidden layer is: is the weight matrix of the x-th hidden layer, is the bias vector of the x-th hidden layer, is the exponential sum of all weights of the x-th layer, and m is the weight matrix of the x-th hidden layer The total number of elements in represents the number of weights of the x-th hidden layer. yes The A weight.

[0077] The calculation formula of the output layer is:

[0078]

[0079] Among them, H (X) is the output of the Xth hidden layer, is the weight matrix of the output layer, is the bias vector of the output layer, is the exponential sum of all weights of the output layer, and q is the output layer weight matrix The total number of elements in represents the number of weights of the output layer, yes The pth weight in It is the prediction result when the input of the medical model is comprehensive features.

[0080] After obtaining the prediction results through the medical model, the medical model needs to be evaluated and optimized, and the prediction performance is evaluated by calculating the mean square error loss function. If the value of the mean square error loss function is high, it means that the prediction error of the medical model is large and further optimization is needed. The optimization method uses the gradient descent technology to generate the medical model parameter update value ΔW of the medical node i i , the mean square error loss function and the gradient descent technique are existing technologies and will not be described in detail here. After each optimization, the prediction results and the value of the mean square error loss function are recalculated until the performance of the medical model reaches the expected standard.

[0081] To protect the privacy of medical data, medical node i updates the medical model parameter value ΔW i Use the Advanced Encryption Standard (AES) algorithm to encrypt and generate the encryption parameter update value E(ΔW i ), then, the digital signature algorithm (DSA) is used to i ) to perform signature processing and generate signature σ i , the advanced encryption standard algorithm and the digital signature algorithm are both existing technologies and will not be described in detail here. Medical node i converts E(ΔW i ) and σ i Sent to the aggregation node via blockchain.

[0082] The aggregation node extracts the encrypted parameter update value and signature of each medical node from the blockchain, and uses the digital signature algorithm to verify the signature. After verification, the aggregation node uses the secure fusion aggregation algorithm to integrate the encrypted parameter update values ​​of each medical node. The secure fusion aggregation algorithm can effectively integrate the encrypted parameter update values ​​of different medical nodes while protecting the privacy of medical data to obtain the aggregated parameter update value. Each medical node only shares the encrypted parameter update value to ensure that the patient's privacy is not leaked.

[0083]

[0084] in, is the aggregation parameter update value, M is the number of participating medical nodes, E(ΔW η ) is the encryption parameter update value of the medical node η, is the last round of aggregation parameter update value. When calculating the aggregation parameter update value in the first round, since there is no last round of aggregation parameter update value, it is necessary to is set to zero. ω is a smoothing factor used to balance and The impact of this model improves the stability and accuracy of the medical model, which can be set according to the specific implementation scenario.

[0085] Aggregation nodes use the Advanced Encryption Standard algorithm to update the aggregation parameter value The encrypted aggregate parameter update value is encrypted and sent back to each medical node through the blockchain. The medical node receives the encrypted aggregate parameter update value from the blockchain, decrypts it using the Advanced Encryption Standard algorithm to obtain the aggregate parameter update value, and then decrypts the aggregate parameter update value using the Advanced Encryption Standard algorithm to obtain the final parameter update value. Will Applied to the optimization of the medical model to complete this round of training cycle. After each round of training cycle, the loss function value of the medical model is calculated. If the loss function value is high, it means that the prediction error of the medical model is large, and it is necessary to continue iterative optimization until the performance of the medical model reaches the expected standard. The expected standard is set according to the specific implementation scenario. The final medical model has higher prediction accuracy and generalization ability, can accurately analyze and predict medical data, and provide reliable support for clinical decision-making. By analyzing the patient's historical data and current condition, it can provide prediction results to help doctors intervene early, develop personalized treatment plans, and evaluate treatment effects, thereby improving the quality and efficiency of medical services.

[0086] In summary, a privacy-preserving medical data analysis method based on blockchain and federated learning was completed.

[0087] The order of the embodiments of the invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0088] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0089] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A privacy-preserving medical data analysis method based on blockchain and federated learning, characterized in that: The following steps are involved: S1: Use the dynamic blockchain resizing algorithm to monitor the data transmission load in real time and dynamically adjust the block size and the number of shards; collect and preprocess the raw medical data in medical institutions, and obtain comprehensive features based on the preprocessed raw medical data through adaptive feature fusion technology; S2: Establish and train a medical model using comprehensive features and multi-layer hidden layer operations to obtain prediction results; evaluate and optimize the medical model based on the prediction results to generate updated values ​​of medical model parameters; Encrypting the medical model parameter update value to generate an encrypted parameter update value; signing the encrypted parameter update value to generate a signature; Perform signature verification on the encrypted parameter update value and the signature, and integrate the encrypted parameter update value to obtain an aggregated parameter update value; The aggregate parameter update value is encrypted and then decrypted to obtain the aggregate parameter update value, and the aggregate parameter update value is decrypted to obtain the final parameter update value; and the medical model is optimized based on the final parameter update value.

2. According to claim 1, a privacy-preserving medical data analysis method based on blockchain and federated learning is characterized in that: The S1 specifically includes: Calculate the data transmission load at time t using the following formula: Among them, L t is the data transmission load at time t, N t-k is the number of transmissions of the encryption parameter update value within time tk, S t-k is the amount of data transmitted for the encryption parameter update value at time tk, F t-k is the transmission frequency of the encryption parameter update value at time tk, α, β, γ are weight coefficients used to adjust the impact of different load indicators on the data transmission load, and v is the size of the sliding window.

3. According to claim 1, a privacy-preserving medical data analysis method based on blockchain and federated learning is characterized in that: The S1 specifically includes: The original medical data includes structured medical original data and image medical original data; the preprocessed original medical data includes structured medical data and image medical data; structured medical data features and image medical data features are extracted from the structured medical data and image medical data respectively.

4. According to claim 3, a privacy-preserving medical data analysis method based on blockchain and federated learning is characterized in that: The S1 specifically includes: Adaptive feature fusion technology splices structured medical data features and image medical data features, and uses the attention mechanism to assign importance weights to the spliced ​​features.

5. According to claim 4, a privacy-preserving medical data analysis method based on blockchain and federated learning is characterized in that: The S1 specifically includes: The splicing features are weightedly fused based on the importance weights to obtain comprehensive features.

6. According to claim 1, a privacy-preserving medical data analysis method based on blockchain and federated learning is characterized in that: The S2 specifically includes: The medical model transforms the comprehensive features through multiple hidden layers and uses the output layer to output the prediction results; the calculation process of the hidden layer is as follows: The calculation formula for the first hidden layer is: Among them, H (1) is the output of the first hidden layer, is the weight matrix of the first hidden layer, is the bias vector of the first hidden layer, is a comprehensive feature, ReLU is an activation function; The calculation formula for the second to Xth hidden layers is: Among them, H (x) is the output of the x-th hidden layer, and is the output of the x-1-th hidden layer. (x-1) The result after the xth hidden layer is: is the weight matrix of the x-th hidden layer, is the bias vector of the xth hidden layer, and m is the weight matrix of the xth hidden layer The total number of elements in represents the number of weights of the x-th hidden layer. yes The A weight.

7. A privacy-preserving medical data analysis method based on blockchain and federated learning according to claim 6, characterized in that: The S2 specifically includes: The calculation formula of the output layer is: Among them, H (X) is the output of the Xth hidden layer, where X represents the total number of hidden layers. is the weight matrix of the output layer, is the bias vector of the output layer, and q is the weight matrix of the output layer The total number of elements in , yes The pth weight in It is the prediction result when the input of the medical model is comprehensive features.

8. According to claim 1, a privacy-preserving medical data analysis method based on blockchain and federated learning is characterized in that: The S2 specifically includes: The secure fusion aggregation algorithm integrates the encryption parameter update values ​​of different medical nodes and introduces a smoothing factor to calculate the aggregate parameter update value; the calculation formula is as follows: in, is the aggregation parameter update value, M is the number of participating medical nodes, E(ΔW η ) is the encryption parameter update value of the medical node η, is the updated value of the aggregation parameter in the previous round; ω is the smoothing factor.

Images