Digital early warning system for lymphedema after breast cancer operation

Abstract

The invention relates to the technical field of health prediction, and discloses a breast cancer postoperative lymphedema digital early warning system comprising a data acquisition module used for acquiring breast cancer postoperative lymphedema monitoring data; the data application module is used for training a neural network correction model for correcting the lymphedema monitoring data based on the acquired lymphedema monitoring data; the model acquisition module is used for acquiring data to be decided from the lymphedema monitoring data and applying the data to be decided to the correction model based on the neural network, compared with a traditional method, misjudgment caused by various interference factors is reduced, and the accuracy of lymphedema diagnosis is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of health prediction, and in particular to a digital early warning system for lymphedema after breast cancer surgery. Background Art

[0002] Breast cancer surgery, especially axillary lymph node dissection, can disrupt the normal pathway of lymphatic drainage, causing local accumulation of lymph fluid and increasing the risk of lymphedema. Postoperative lymphedema not only affects the patient's limb function, but can also cause complications such as pain and infection, seriously reducing the patient's quality of life. Therefore, early warning is crucial to improving patient prognosis.

[0003] In the detection of lymphedema after breast cancer surgery, bioimpedance analysis (BIA) is an important basis for evaluating body composition and physiological status by measuring the electrical impedance characteristics of human tissue to weak alternating current. In human tissue, water, electrolytes, etc. are key factors affecting electrical impedance. When lymphedema occurs, the interstitial fluid increases, causing the electrical impedance of local tissues to change. By placing electrodes on the body surface, injecting weak alternating current into the body, and measuring the voltage and current, the electrical impedance value can be calculated, and then the change in tissue water content can be judged to indicate the possibility of lymphedema.

[0004] At present, patients are prone to electrolyte imbalance after surgery, such as dietary changes, drug side effects, and the impact of the disease itself, which can lead to electrolyte imbalance and then cause changes in impedance values. If the data is not corrected, it will affect the subsequent impedance measurement results. Therefore, a digital early warning system for lymphedema after breast cancer surgery is proposed. Summary of the invention

[0005] The main purpose of the present invention is to provide a digital early warning system for lymphedema after breast cancer surgery to solve the problems raised in the above background.

[0006] A digital early warning system for lymphedema after breast cancer surgery, comprising: Data collection module: used to collect lymphedema monitoring data after breast cancer surgery; Data application module: Based on the acquired lymphedema monitoring data, a neural network correction model is trained to correct the lymphedema monitoring data; Model acquisition module: obtains the data to be decided from the lymphedema monitoring data, and applies the data to be decided to the correction model based on the neural network.

[0007] The postoperative lymphedema monitoring data of breast cancer includes skin moisture value data and original electrical impedance value data.

[0008] The process of training a neural network correction model for correcting lymphedema monitoring data is as follows: Correct and normalize lymphedema monitoring data; Skin moisture value data is expressed as , the original electrical impedance data is expressed as , read the experimental data from the local database, remove abnormal data points, and for each skin moisture value The electrical impedance value under , get its average value and standard deviation ,when , then delete the data point; Skin moisture value data And the original electrical impedance data Perform normalization operation, the formula is expressed as: ,in, is the normalized skin moisture value data; Raw electrical impedance data Perform normalization operation: ,in, It is the normalized raw electrical impedance data.

[0009] The corrected and normalized lymphedema monitoring data were divided into training set, validation set and test set in the proportion of dividing the training data; Construct a feedforward neural network with an input layer, a hidden layer, and an output layer; The input layer node is the corrected and normalized humidity value , the output layer node is the corrected and normalized electrical impedance value , hidden layer settings neurons, through a Activation function, initialize the weights of the neural network and bias , the weight is arrive The bias is initialized to 0. The feedforward neural network is trained using the training set data to correct and normalize the humidity values ​​in the dataset. and electrical impedance Perform forward propagation; First, input the layer to the hidden layer, and set the weight of the input layer to be , the bias is , then the neuron input of the hidden layer is: ,in, is the neuron input of the hidden layer. When p=1, t=0, =1; Furthermore, the input signal of the hidden layer neuron is obtained by weighted summing the corrected and normalized humidity value of the input layer with the weight and adding the bias. Determines the degree of influence of the input humidity value on the hidden layer neurons. Different The value can make the model focus on different characteristics or changing trends of humidity values. An additional learnable parameter is provided for the neuron, so that the neuron can have a certain output when the input is zero, increasing the flexibility of the model; The output of the neurons in the hidden layer is: , is the activation function; pass The activation function performs a nonlinear transformation on the input signal to obtain the output of the hidden layer, which will serve as the input of the next layer (from the hidden layer to the output layer) to continue transmitting and processing information; Then from the hidden layer to the output layer, the weight of the input layer is set to , the bias is , then the neuron output of the hidden layer is: ,in, is the number of neurons in the hidden layer, Obtain a feed-forward neural network for the output layer output value; Furthermore, the output of the hidden layer and the weights from the hidden layer to the output layer Perform weighted summation and add bias , and obtain the final prediction output. By adjusting, the model can learn how to convert the features extracted by the hidden layer into accurate electrical impedance value predictions, providing an adjustable parameter that helps optimize the prediction results; Train a loss function, the formula is: ,in, is the number of samples in the training set; Furthermore, this loss function measures the difference between the model's predicted value and the true value. For each lymphedema monitoring data sample in the training set, the square of the difference between the predicted value and the true value is calculated, and then the square differences of all samples are summed and averaged. The smaller the value of the loss function, the better the prediction effect of the model. Finally, the validation set and test set are used to evaluate the trained neural network model and obtain the determination coefficient ,in, is the average value of the actual normalized electrical impedance value, and the model with the best performance is selected based on the evaluation results; Furthermore, The coefficient of determination reflects the degree of fit of the model to the data. Its value range is between 0 and 1. The closer it is to 1, the better the model fits the data. On the contrary, the closer it is to 0, the poorer the model fits the data. Represents the total square error between the model prediction value and the true value, the denominator It represents the total squared error between the true value and the mean of the true value. If the model cannot predict the data at all, it is 0; if the model can perfectly predict the data, it is 1.

[0010] The data to be decided includes normalized and corrected skin moisture value data representation And the original electrical impedance data ; Furthermore, by using the trained neural network correction model to process new lymphedema monitoring data, early warning of lymphedema can be achieved. After the normalized corrected skin moisture value is input into the model, the model can predict and correct the impedance value according to the relationship between humidity and impedance learned in the training stage. By transmitting the original impedance value, the model application submodule can perform a denormalization operation to convert the normalized corrected impedance value output by the neural network model into an actual corrected impedance value. This actual corrected impedance value combined with the original impedance value and other clinical information can more accurately determine whether the patient has a risk of lymphedema, providing doctors with more valuable diagnostic basis. Furthermore, the early warning module compares the predicted electrical impedance value with the patient's previous measured values ​​and the normal reference values ​​of patients of the same type based on the output decision results of the neural network correction model. If the predicted electrical impedance value exceeds the normal range or shows an abnormal change trend (such as multiple consecutive measured values ​​gradually deviating from the normal range), the system triggers the lymphedema warning mechanism.

[0011] The process of applying the decision-making data to the correction model based on the neural network is as follows: The normalized corrected skin moisture value data is expressed as Input into the neural network correction model, and obtain the predicted corrected electrical impedance value based on the neural network correction model ; Furthermore, the normalized corrected electrical impedance value is obtained through forward propagation calculation. The lymphedema monitoring data passes through each hidden layer from the input layer in sequence and finally reaches the output layer. In this process, the weighted summation and activation function of each layer of neurons are transmitted and transformed to finally obtain the predicted corrected electrical impedance value. ; Prediction and correction of electrical impedance value based on neural network correction model Perform reverse normalization: = ,in, It is the current in the process of bioelectrical impedance measurement; The system will then compare the corrected impedance value with the patient's previous measured values ​​and the normal reference values ​​for similar patients. If the predicted impedance value exceeds the normal range or shows an abnormal trend (such as multiple consecutive measured values ​​gradually deviating from the normal range), the system will trigger the lymphedema early warning mechanism and provide personalized early warning plans based on the patient's specific circumstances. By analyzing the patient's monitoring data and early warning results, the system can develop targeted rehabilitation training plans, stress treatment plans, and lifestyle adjustment recommendations for the patient to help reduce the risk of lymphedema and improve postoperative rehabilitation outcomes.

[0012] A digital early warning method for lymphedema after breast cancer surgery, comprising: S1: used to collect lymphedema monitoring data after breast cancer surgery; S2: Based on the acquired lymphedema monitoring data, a neural network correction model for correcting the lymphedema monitoring data is trained; S3: obtaining decision-making data from lymphedema monitoring data, and applying the decision-making data to a correction model based on a neural network; S4: The neural network-based revised model outputs a decision result, which is compared with the patient's previous measurements.

[0013] The present invention has the following beneficial effects: 0. In the present invention, through the constructed feedforward neural network structure, the input layer receives the normalized humidity value, the output layer outputs the normalized electrical impedance value, and the hidden layer sets multiple neurons and adopts The activation function can effectively learn the complex nonlinear relationship between skin humidity and electrical impedance value. During the training process, through the learning of a large amount of experimental data, the model can accurately capture the changing law of electrical impedance value under different humidity conditions. Therefore, in practical applications, it can more accurately predict the changing trend of electrical impedance value based on the input skin humidity value.

[0014] 1. In the present invention, the system can collect the patient's lymphedema monitoring data in real time, including skin moisture value and original electrical impedance value, and quickly input the normalized corrected skin moisture value into the trained neural network correction model. The model quickly calculates the predicted corrected electrical impedance value based on the learned relationship between humidity and electrical impedance.

[0015] 2. In the present invention, during the model application process, by transmitting the original electrical impedance value, the system can perform a denormalization operation in the model application submodule, and convert the normalized corrected electrical impedance value output by the neural network model into the actual corrected electrical impedance value, so as to more comprehensively and accurately judge the numerical accuracy of the electrical impedance value, and reduce the misjudgment caused by various interference factors (such as changes in skin moisture, electrolyte imbalance, etc.). BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a system block diagram of a digital early warning system for lymphedema after breast cancer surgery according to the present invention; Figure 2 This is a method step diagram of a digital early warning system for lymphedema after breast cancer surgery according to the present invention. DETAILED DESCRIPTION

[0017] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0018] Embodiment 1 Please refer to Figure 1 to Figure 2 As shown, a digital early warning system for lymphedema after breast cancer surgery includes: Data collection module: used to collect lymphedema monitoring data after breast cancer surgery; Data application module: Based on the acquired lymphedema monitoring data, a neural network correction model is trained to correct the lymphedema monitoring data; Model acquisition module: obtains the data to be decided from the lymphedema monitoring data, and applies the data to be decided to the correction model based on the neural network.

[0019] The lymphedema monitoring data after breast cancer surgery includes skin moisture value data and raw electrical impedance value data.

[0020] Furthermore, the data acquisition module is the basis of the entire early warning process, responsible for collecting lymphedema monitoring data of patients after breast cancer surgery, including skin moisture value data and raw electrical impedance value data. These data provide raw materials for subsequent model training and early warning judgment.

[0021] The process of training a neural network correction model for correcting lymphedema monitoring data is: Correct and normalize lymphedema monitoring data; Skin moisture value data is expressed as , the original electrical impedance data is expressed as , read the experimental data from the local database, remove abnormal data points, and for each skin moisture value The electrical impedance value under , get its average value and standard deviation ,when , then delete the data point; Skin moisture value data And the original electrical impedance data Perform normalization operation, the formula is expressed as: ,in, is the normalized skin moisture value data; Raw electrical impedance data Perform normalization operation: ,in, It is the normalized raw electrical impedance data.

[0022] The corrected and normalized lymphedema monitoring data were divided into training set, validation set and test set in the proportion of dividing the training data; Construct a feedforward neural network with an input layer, a hidden layer, and an output layer; The input layer node is the corrected and normalized humidity value , the output layer node is the corrected and normalized electrical impedance value , hidden layer settings neurons, through a Activation function, initialize the weights of the neural network and bias , the weight is arrive The bias is initialized to 0. The feedforward neural network is trained using the training set data to correct and normalize the humidity values ​​in the dataset. and electrical impedance Perform forward propagation; First, input the layer to the hidden layer, and set the weight of the input layer to be , the bias is , then the neuron input of the hidden layer is: ,in, is the neuron input of the hidden layer. When p=1, t=0, =1; Furthermore, the input signal of the hidden layer neuron is obtained by weighted summing the corrected and normalized humidity value of the input layer with the weight and adding the bias. Determines the degree of influence of the input humidity value on the hidden layer neurons. Different The value can make the model focus on different characteristics or changing trends of humidity values. An additional learnable parameter is provided for the neuron, so that the neuron can have a certain output when the input is zero, increasing the flexibility of the model; The output of the neurons in the hidden layer is: , is the activation function; pass The activation function performs a nonlinear transformation on the input signal to obtain the output of the hidden layer, which will serve as the input of the next layer (from the hidden layer to the output layer) to continue transmitting and processing information; Then from the hidden layer to the output layer, the weight of the input layer is set to , the bias is , then the neuron output of the hidden layer is: ,in, is the number of neurons in the hidden layer, Obtain a feed-forward neural network for the output layer output value; Furthermore, the output of the hidden layer and the weights from the hidden layer to the output layer Perform weighted summation and add bias , and obtain the final prediction output. By adjusting, the model can learn how to convert the features extracted by the hidden layer into accurate electrical impedance value predictions, providing an adjustable parameter that helps optimize the prediction results; Train a loss function, the formula is: ,in, is the number of samples in the training set; Furthermore, this loss function measures the difference between the model's predicted value and the true value. For each lymphedema monitoring data sample in the training set, the square of the difference between the predicted value and the true value is calculated, and then the square differences of all samples are summed and averaged. The smaller the value of the loss function, the better the prediction effect of the model. Finally, the validation set and test set are used to evaluate the trained neural network model and obtain the determination coefficient ,in, is the average value of the actual normalized electrical impedance value, and the model with the best performance is selected based on the evaluation results; Furthermore, The coefficient of determination reflects the degree of fit of the model to the data. Its value range is between 0 and 1. The closer it is to 1, the better the model fits the data. On the contrary, the closer it is to 0, the poorer the model fits the data. Represents the total square error between the model prediction value and the true value, the denominator It represents the total squared error between the true value and the mean of the true value. If the model cannot predict the data at all, it is 0; if the model can perfectly predict the data, it is 1.

[0023] The data to be decided includes normalized and corrected skin moisture value data representation And the original electrical impedance data ; Furthermore, by using the trained neural network correction model to process new lymphedema monitoring data, early warning of lymphedema can be achieved. After the normalized corrected skin moisture value is input into the model, the model can predict and correct the impedance value according to the relationship between humidity and impedance learned in the training stage. By transmitting the original impedance value, the model application submodule can perform a denormalization operation to convert the normalized corrected impedance value output by the neural network model into an actual corrected impedance value. This actual corrected impedance value combined with the original impedance value and other clinical information can more accurately determine whether the patient has a risk of lymphedema, providing doctors with more valuable diagnostic basis. Furthermore, the early warning module compares the predicted electrical impedance value with the patient's previous measured values ​​and the normal reference values ​​of patients of the same type based on the output decision results of the neural network correction model. If the predicted electrical impedance value exceeds the normal range or shows an abnormal change trend (such as multiple consecutive measured values ​​gradually deviating from the normal range), the system triggers the lymphedema warning mechanism.

[0024] The process of applying the decision-making data to the correction model based on neural network is: The normalized corrected skin moisture value data is expressed as Input into the neural network correction model, and obtain the predicted corrected electrical impedance value based on the neural network correction model ; Furthermore, the normalized corrected electrical impedance value is obtained through forward propagation calculation. The lymphedema monitoring data passes through each hidden layer from the input layer in sequence and finally reaches the output layer. In this process, the weighted summation and activation function of each layer of neurons are transmitted and transformed to finally obtain the predicted corrected electrical impedance value. ; Prediction and correction of electrical impedance value based on neural network correction model Perform reverse normalization: = ,in, It is the current in the process of bioelectrical impedance measurement; The system then compares the corrected impedance value with the patient's previous measured values ​​and the normal reference values ​​of similar patients. If the predicted impedance value exceeds the normal range or shows an abnormal trend of change (such as multiple consecutive measurements gradually deviating from the normal range), the system triggers the lymphedema early warning mechanism and provides personalized recommendations based on the patient's specific situation. By analyzing the patient's monitoring data and early warning results, the system can develop targeted rehabilitation training plans, stress treatment plans, and lifestyle adjustment recommendations for the patient to help patients reduce the risk of lymphedema and improve postoperative rehabilitation effects.

[0025] Please refer to Figure 1 to Figure 2 As shown, the digital early warning method for lymphedema after breast cancer surgery: S1: used to collect lymphedema monitoring data after breast cancer surgery; S2: Based on the acquired lymphedema monitoring data, a neural network correction model for correcting the lymphedema monitoring data is trained; S3: obtaining decision-making data from lymphedema monitoring data, and applying the decision-making data to a correction model based on a neural network; S4: The neural network-based revised model outputs a decision result, which is compared with the patient's previous measurements.

[0026] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A digital early warning system for lymphedema after breast cancer surgery, characterized in that: include: Data collection module: used to collect lymphedema monitoring data after breast cancer surgery; Data application module: Based on the acquired lymphedema monitoring data, a neural network correction model is trained to correct the lymphedema monitoring data; Model acquisition module: obtains the decision-making data from the lymphedema monitoring data, and applies the decision-making data to the correction model based on the neural network to obtain the predicted corrected electrical impedance value .

2. A digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 1, characterized in that: The postoperative lymphedema monitoring data of breast cancer includes skin moisture value data and original electrical impedance value data.

3. The digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 1, characterized in that: The training process of the neural network correction model is: Set the skin moisture value data to be expressed as , the original electrical impedance data is expressed as , correct and normalize the lymphedema monitoring data to obtain the corrected and normalized skin moisture value data , the original electrical impedance data is expressed as ; The corrected and normalized lymphedema monitoring data were divided into training set, validation set and test set in the proportion of dividing the training data; Construct a feedforward neural network with an input layer, a hidden layer, and an output layer; Based on the feedforward neural network training loss function, the trained neural network model is evaluated using the validation set and test set, and the determination coefficient evaluation results are obtained to select the model with the best performance.

4. A digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 3, characterized in that: The process of correcting and normalizing the lymphedema monitoring data is as follows: Read the experimental data from the local database, remove abnormal data points, and for each skin moisture value The electrical impedance value under , get its average value and standard deviation ,when , then delete the data point; Skin moisture value data And the original electrical impedance data Perform normalization operation, the formula is expressed as: ,in, is the normalized skin moisture value data; Raw electrical impedance data Perform normalization operation: ,in, It is the normalized raw electrical impedance data.

5. The digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 3, characterized in that: The process of constructing a feedforward neural network having an input layer, a hidden layer and an output layer is as follows; The corrected and normalized humidity value As the input layer node, the corrected and normalized electrical impedance value As the output layer node, the hidden layer is set neurons, through a Activation function, initialize the weights of the neural network and bias , the weight is arrive The bias is initialized to 0. The feedforward neural network is trained using the training set data, and the normalized humidity values ​​in the data set are and electrical impedance Perform forward propagation and finally obtain a feedforward neural network.

6. A digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 5, characterized in that: The corrected and normalized humidity values ​​in the dataset and electrical impedance The process of forward propagation is: First, input the layer to the hidden layer, and set the weight of the input layer to be , the bias is , then the neuron input of the hidden layer is: ,in, is the neuron input of the hidden layer. When p=1, t=0, =1; The output of the neurons in the hidden layer is: , is the activation function; Then from the hidden layer to the output layer, the weight of the input layer is set to , the bias is , then the neuron output of the hidden layer is: ,in, is the number of neurons in the hidden layer, Output value for the output layer.

7. The digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 5, characterized in that: The process of obtaining the determination coefficient evaluation result and selecting the model with the best performance is: Train a loss function: ,in, is the number of samples in the training set; Use the validation set and test set to evaluate the trained neural network model and obtain the determination coefficient ,in, is the average value of the actual normalized electrical impedance value, and the model with the best performance is selected based on the evaluation results.

8. The digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 1, characterized in that: The data to be decided include normalized and corrected skin moisture value data and original electrical impedance value data.

9. The digital early warning system for postoperative lymphedema after breast cancer surgery according to claim 1, characterized in that: The process of applying the decision-making data to the correction model based on the neural network is as follows: The normalized corrected skin moisture value data is expressed as Input into the neural network correction model, and obtain the predicted corrected electrical impedance value based on the neural network correction model ; Prediction and correction of electrical impedance value based on neural network correction model Perform reverse normalization: = ,in, It is the current during the bioelectrical impedance measurement process.

Images