Design And Construction Of A Medical Efficacy Prediction Model Based On Gray Markov's Qi And Yin Deficiency Type Chronic Kidney Disease

Chronic kidney disease(CKD)refers to the dysfunction of kidney caused by various causes,and the dysfunction is more than 3 months.There are two judgments of chronic kidney disease.Firstly,the dysfunction of kidney lasts for more than three months,with abnormal pathological section or abnormal renal index,with or without the falling glomerular filtration rate(GFR).Secondly,when GFR<60ml/min,and the GFR lasts for more than three months,with or without the dysfunction of kidney.Although there is no chronic kidney disease in Traditional Chinese Medicine(TCM)theory,according to its clinical manifestations and characteristics,it can take full advantage of syndrome differentiation and treatment in TCM.It can prevent the GFR from cutting down,and even restore the kidney function.In the syndrome differentiation and treatment of CKD,TCM divides syndrome into the deficiency syndrome(spleen-kidney qi deficiency syndrome,spleen-kidney yang deficiency syndrome,Qi-yin deficiency syndrome,liver-kidney yin deficiency syndrome,yin-yang deficiency syndrome)and Empirical evidence(moist and damp syndrome,damp heat syndrome,stasis blood syndrome,water and gas syndrome,and muddy poison syndrome),in which Qi and Yin deficiency are most common in CKD.In the treatment of CKD with deficiency of qi and yin,it is very important to predict the efficacy of the medicine.It can not only predict the later patient’s recovery,but also modify the treatment plan in time.This study uses the Markov model to predict the clinical symptoms of CKD,uses the gray model(GM)to predict the laboratory data such as GFR,and combines the Markov model with the GM to make the Gray Markov model After correcting the gray prediction model whose accuracy cannot meet the requirements,the three models are applied to the prediction of clinical symptoms and laboratory data of CKD,comprehensive prediction and analysis of the efficacy of patients after medication.Objective To build a Markov model,a GM,and a Gray Markov model for patients with CKD with deficiency of both qi and yin,based on the data of a single patient that meets the predicted conditions,and to explore three models in TCM.The predictive model construction method for the effect prediction.Methods Based on the Markov model making method and its main points,the clinical symptoms of patients in accordance with the prediction conditions of Markov model were screened,and the Markov model was constructed based on the time sequence of the clinical symptoms of the patients and the state transition.The accuracy of the test,the state of changes in the late patients to predict the status of the law of the transfer of the analysis.Based on the GM modeling methods and the main points,according to the GM to predict the patient’s laboratory data,based on the data time series,build a GM,test the precision of the model and predict the specific numerical value of the laboratory value of the patient in the future with excellent prediction accuracy.Based on the Gray Markov model modeling method and the main points,the GM whose accuracy is low modified to use the GM constructed from laboratory data to predict interval values and numerical trends with low precision.Results A Markov model was constructed based on a total of 166 symptoms in 66 patients who met the prediction criteria.The predicted situation was compared with the actual situation,there were 121 cases which treats excellent,accounting for 72.90%,and the probability was consistent with the actual situation,with reliable prediction accuracy.After predicting 166 clinical symptoms by the Markov prediction model,it can be predicted that the improvement of the clinical symptoms of qi deficiency and yin deficiency type chronic kidney disease of this Chinese medicine reagent is as follows:backache weakness(44%),shortness of breath(50%),cold(55%),spontaneous sweating(78%),throat and throat(58%),hand,foot and heart heat(72%),dizziness and tinnitus(71%),dry eyes(50%),nocturia(50%).A GM was constructed based on the GFR values of the 76 patients that met the prediction criteria.The GM constructed based on the time series data of 11 patients GFR met the accuracy requirements,the accuracy requirements for the prediction of specific numerical values.After predicting the GFR value of 11 patients by the GM,it can be concluded that after taking the Chinese medicine treatment,the patients’ GFR of 10 patients is increasing,whose kidney function is recovered,and the only 1 patient’s GFR is decreased.In addition to the gray prediction model with excellent accuracy for 11 patients,a gray prediction model was established based on the GFR of 65 patients with low prediction accuracy,and a Markov model was used to correct it to construct a Gray Markov model.After predicting the GFR of 65 patients using the Gray Markov model,the GFR of 35 patients shows an upward trend,the GFR of 15 patients remains stable,and the GFR of 10 patients shows a decreasing trend.Conclusions In this study,the Markov model can predict the state transition and the transition clinical symptoms of patients.The GM can predict the specific laboratory values in later time,and the Gray Markov prediction model can be revised.The GM,which does not meet the required accuracy,predicts the numerical trend and numerical interval of the laboratory numerical class in the later period.When the three models are applied to TCM for CKD with deficiency of both qi and yin,the prediction results were satisfactory.At the same time,the three models has an ideal predicting capability,which is feasible,adaptable and accurate,and in well accordance with the developing rules of current medicine.Therefore,it can be used in traditional Chinese medicinal prediction.