Research On Three-dimensional Modeling Of Knee Joint And Detection Of Effusion Based On Biological Impedance

The incidence of knee joints is the highest among all joints in the body.Many acute and chronic knee injuries are accompanied by the emergence of knee effusions.These diseases severely affect patients’ quality of life.Therefore,the prediction of early knee joint disease is helpful for the diagnosis and treatment of the disease.The bioimpedance knee joint effusion detection is based on the principle that the physiological structure,pathological condition of the human body and the electrical characteristics of tissues and organs have a high correlation.A tiny current is injected through an electrode placed on the outer surface of the human knee joint to detect the impedance signal and obtain biomedical information related to the edema of the knee joint.This paper studies the key issues of knee bioimpedance detection,constructs an approximate human model of the knee joint,designs a method for knee joint impedance detection,and proposes a knee effusion severity classification algorithm.Firstly,the pathogenesis of knee effusion and common knee diseases that cause effusion are studied.Based on the working principle of bioimpedance detection technology and electrical parameters of healthy human tissue,relevant parameters of simulation and experimental detection system are designed.Secondly,based on the anatomy of the knee joint,and using CT pictures of the lower limbs of the human body,the required construction parts in the simulation model were selected and segmented to construct an approximate human knee simulation model.According to the principle of electromagnetic field calculation,the variation process of knee joint impedance under different fluid accumulation conditions was simulated.The simulation results show the potential distribution state when the severity of the fluid accumulation is different and the fluid accumulation position is different.After the fluid accumulation increases,the regional boundaries of each tissue structure are gradually blurred,and as the effusion generation area moves,the low potential area will also move with it.A detection electrode position and a fourelectrode knee joint detection system are designed.The test results show that an increase in the degree of effusion of the knee joint will cause the surface potential and impedance of the human body to decrease,and the local effusion will cause the potential in this area to be lower than in other areas.The agar knee model was made according to the model size,and impedance measurements were performed for different levels of effusion.The experimental results show that the experimental measurement values are consistent with the impedance change trend in the simulation calculation.The rationality of the three-dimensional model of the knee joint and the effectiveness of the selection of the location of the detection points and the construction of the system were verified.Finally,a classification algorithm of knee effusion severity based on ensemble learning is proposed.Using the simulation data,using the impedance and conductivity as feature quantities,a knee joint effusion dataset was produced.By constructing a Stacking integration model with KNN,DT,SVM,DNN as the base classifier and Logistic regression algorithm as the secondary classifier,The data set training uses a 5-fold cross-validation method and calculates the prediction indicators of the five models,classify and analyze the data in the data set,and realize the graded prediction of the severity of the effusion.The results show that the Stacking model has the best prediction effect,the classification accuracy reaches 97.67%,the Kappa coefficient and AUC value reach 0.9650 and 0.9825,respectively,which proves that the proposed Stacking-based model is suitable for predicting the severity of knee joint effusion,and can accurately distinguish the diseases..