Drawing And Diagnosis Of Female Basic Body Temperature Curve Based On MQ Kernel Regression

The changes in female physiological cycle are closely related to ovarian cycle activity,it is often accompanied by changes in body temperature before and after ovulation.The basal body temperature curve is obtained by measuring the basal body temperature,and then the changes can be observed to determine whether the woman has ovulation,the estimated ovulation date,and the analysis of the corpus luteum function.Therefore,determining the basal body temperature curve of women can help doctors diagnose and treat certain gynecological diseases.However,due to the dynamic and static characteristics of the human body,the acquisition of body temperature data and further analysis will be affected to some extent.Therefore,how to draw a curve that accurately reflects the changes in women's physiological cycle is particularly important.In this paper,we collect the body temperature data of the minute frequency,use the processing and analysis method of big data,combine the statistical modeling and medical mechanism to fit the basal body temperature curve of a woman in a cycle,and then carry out corresponding analysis,which can help doctors.More accurate judgments and predictions of women's physiological cycle and ovarian function are made clinically.The specific content of this paper is as follows:Firstly,the collected body temperature data is analyzed,and the intra-day basal body temperature extraction algorithm is constructed to extract the daily representative body temperature point in combination with the characteristics of human body temperature.Due to factors such as measurement instrument error,human factors and environment during the measurement process,some of the missing data will result in poor performance of the body temperature curve,which will affect our judgment.Therefore,after we get the daily representative body temperature point,we use the multiple interpolation method to supplement the missing data.Since the body temperature data after interpolation still belongs to the non-stationary nonlinear body temperature signal,it is not obvious enough to judge the period through the graph.Therefore,we use the EMD algorithm to perform deep-level filtering and gradually decompose a series of high-frequency to low-frequency A more adaptive basis function,then select the bandpass filter of the appropriate frequency to reconstruct the signal,find the inflection point of high and low temperature,and then find the menstrual cycle of women.Next,we combine multiple cycles into one cycle to plot the average body temperature data over a period.After observation,the temperature curve drawn still has noise phenomenon.Based on the superiority of non-parametric regression,we use the kernel regression to adjust the bandwidth of the MQ trigonometric function to construct the MQ triangle spline with smoothness,stability and consistency.The interpolation regression model finally uses the MQ trigonometric quasi-interpolation to return the basal body temperature biphasic curve.Finally,the clinical medical knowledge and the basal body temperature curve drawn are used to predict and diagnose the female's physiological cycle and ovarian function.The specific process is:by observing the characteristics of the speed,amplitude and duration of the high and low temperature phase change time in the basal body temperature curve,the physiological cycle and ovarian function of the women who come to the clinic are predicted,and they can be divided into normal ovulation according to the predicted results.And abnormal ovulation in two major categories.Among them,abnormal ovulation can be divided into early pregnancy,luteal dysfunction and so on.Through the research in this paper,women with gynecological diseases can be treated in time for medical treatment,including giving advice on pregnancy during a reasonable period of time,which can improve the success rate of conception and provide a new method for clinical medical diagnosis technology.