Study On Microwave Non-invasive Blood Glucose Based On DLA Fractal Earlobe Dispersion Model

Diabetes is a major threat to human health.In contemporary society,the incidence of diabetes is increasing,which causes huge economic burden and social problems.At present,there is no radical cure for diabetes.Generally,blood is frequently obtained by means of needles to achieve blood glucose level monitoring,and then blood glucose level is effectively controlled.Compared with invasive or minimally invasive methods,microwave non-invasive blood glucose monitoring technology has the advantages of non-invasive,painless,non-infectious,safe and radiation-free.This paper takes the human earlobe as the research object.The earlobe area has many advantages of relatively simple biological structure,sufficient blood content,relatively flat surface,which makes the selected area easy to be measured and suitable to be monitored for long-term.Microwave non-invasive blood glucose detection is used to analyze and evaluate the blood glucose level according to the difference of microwave signals.These signals will have different characteristics after passed through earlobe areas with different blood glucose concentrations.This technology has a good feasibility and will make sense in the future.The open-ended coaxial probe method is used to measure the dielectric properties of different concentrations of glucose solution exposed to microwaves in the 200 MHz to 10 GHz range.In this paper,a set of one-order Debye parameters can be determined by analyzing the experimental data,which makes the deviation value between predicted data and experimental data is less than 0.32.An equation consisted by the set of Debye parameters is established to evaluate the practical dielectric properties for any given glucose solutions.In this paper,the two-dimensional and three-dimensional electromagnetic models of earlobe and antenna transceiver structure are established based on the finite difference time domain algorithm.Based on three-dimensional layered earlobe modeling and frequency-domain signal analysis,this paper studies the feasibility of microwave method for detecting blood glucose levels,the influence of dielectric properties on the spectral results,and the influence of blood volume and blood distribution position on the received microwave signal.Furthermore,this paper proposes a method for constructing the blood structure in the earlobe model using diffusion limited aggregation fractal technique,which can simulate the continuity,randomness and uniformity of blood distribution better than the traditional model,and provide a more accurate and complex simulation environment for subsequent studies.The blood structure constructed by using diffusion limited aggregation fractal method accounts for 50.14%of the fat space,and the error is only 0.14%compared to the expected 50%.Considering the dispersion characteristics of biological tissue,the auxiliary differential equation based on Debye biological model and the finite difference time domain are combined to deal with the dispersion characteristics of the biological tissues,and which will make simulation results more accuracy.Based on the earlobe dispersion model and the broadband signal transceiver simulation,it is found that there is a good linear dependence between the phase of S₂₁ and the glucose concentration in the frequency range of 4-6 GHz.Furthermore,the method of estimating glucose level based on S₂₁ phase is proposed,and its feasibility and superiority are analyzed.