Research On Human Body Timing Signal Analysis Method Based On Wavelet Modulus Maximum Multi-fractal

The rapid development of information processing technology has made medical detection technology widely used.At the same time,with the continuous improvement of multi-fractal theory,the singularity detection of human body time-series has made important progress.In view of the fact that thyroid malignant tumors and heart diseases are currently high-risk diseases that seriously endanger human health,how to scientifically and effectively detect and evaluate the health status of human organs has become the focus of current research.Therefore,it is of great significance to carry out wavelet modulus maximal multi-fractal analysis of human thyroid temperature time-series and heart rhythm time-series.In this paper,the pretreatment of human thyroid temperature time-series and heart rate time-series were taken as the starting point,and the statistical distribution of the multi-fractal spectrum characteristic parameters of these were obtained respectively.Finally,the multi-fractal characteristics of thyroid temperature time-series between healthy individuals were respectively carried out.Difference test,multi-fractal characteristics difference test between different types of heart rhythm time-series,the specific research work were as followed:1.Wavelet modulus maximal multi-fractal analysis of thyroid temperature time-seriesFirstly,the image acquisition environment was set up,the dynamic infrared image of thyroid in healthy subjects was collected,image registration and meshing were performed,and the thyroid temperature time-series was obtained.Secondly,the optimal length of multi-fractal analysis of thyroid temperature time-series signal is discussed.The wavelet function and the transform scale factor were set,and the hard thresholds under different scale transformations were set,and then the multi-scale wavelet modulus maximal sequence was obtained.Then,the range of statistical moment orders was determined,and the multi-fractal spectrum of all grid regions were obtained.Statistical distribution of spectral lines;Finally,the SNK-q test was used to verify the difference in multi-fractal characteristics of thyroid temperature time-series between healthy individuals.Studies show that when the sampling point of the thyroid temperature time-series of healthy individuals is greater than 12000,the half-width of the multi-fractal begins to converge,and the fluctuation is gentle in the interval of 0.12-0.14;the fractal dimension of the multi-fractal obtains the extreme value,corresponding to the singularity index(8₁distribution concentrates in the range of 1.1-1.3,the gap coefficient(8₂ distributes in the range of 0.002-0.005,and the test levelαof individual difference of the mean of these factors is 0.01;the mean of the half-width of multi-fractal concentrates 0.165,the test levelαof no individual difference is 0.05.2.Wavelet modulus maximal multi-fractal analysis of heart rate time-seriesFirstly,seven kinds of heart rhythm time-series were downloaded in the MIT-BIH database,and the module wavelet bior5.5 was selected to perform multi-scale wavelet transform to give the value of the statistical moment order suitable for multi-fractal analysis of human heart rate time-series.Then,the fitting of the quality index was discussed,and the multi-fractal spectrum of different types of heart rhythm time-series was solved.Finally,the statistical distribution of the multi-fractal characteristic parameters(8₁ and(8₂ were analyzed,and the medical statistical test theory was used to verify the diseased heart rate time-series.The difference in half-width of multi-fractal spectrum between the signals with healthy heart rhythms and different types of disease signals was significant or not.The research shows that when the sampling point of heart rate timing signal is greater than8000,the variation trend of the half-width of the multi-fractal spectrum of different types of heart rhythm time-series is generally consistent,and the fluctuation is gentle in the corresponding convergence interval;the disease signals can be distinguished from the health signals,and the test levelαis 0.05;except for the arrhythmia and congestive heart failure,malignant ventricular ectopic and sudden cardiac death,atrial fibrillation and malignant ventricular ectopic are not recognized,the other diseases and diseases can be distinguished from each other,and the difference statistically significant at the test levelαis 0.05.