Research On Tongue Features Of Diabetic Kidney Disease And The Correlation With Physical And Chemical Indexes

Objective:A parameterized tongue color determination strategy was established based on LAB color space to assist in the observation of diabetic kidney disease tongue features.Expatiated pathogenesis and evolution of diabetic kidney disease objectively from the tongue image.Then analyzed the relationship between tongue and physical and chemical indexes,to explore the internal physical and chemical changes corresponding to different tongue manifestations.To provide a theoretical basis for the unification of the syndrome differentiation and classification of diabetic kidney disease,and expand the application of tongue diagnosis in the diagnosis and treatment of diabetic kidney disease.Methods:1.Set up standard lighting environment: Adjust the color temperature of the filling light source to 5600 K,color rendering > 95,adjustable brightness,shooting Angle 45°/0°(angle of the light source/shooting Angle,relative to the horizontal line),set the camera to manual white balance(5600K),taken 100 tongue images under the ambient light source covered by the bed curtain.Selected the top 50 tongue graphs according to the shooting order,each tongue image was judged by two experienced physicians,and extracted the L,A and B values of tongue graphs.And then,observed the value range and characteristics of different tongue colors in L,A and B,set up the boundary value to establish the parameter determination strategy.The last 50 cases were used to compare the similarity between parametric decision and subjective decision,to verify its feasibility and accuracy.2.Screened eligible patients with diabetic renal disease according to strict diagnosis and discharge criteria,divided the patients into three stages: early,middle and late according to the Mogensen stage.Collected tongue images in a relatively standard light environment,and determined the tongue image by two experienced TCM physicians combined with the results of parameter determination,collected the general data and physical and chemical indexes of the patients.Chi-square test,one-way analysis of variance,nonparametric rank sum test and other statistical methods were used to analyze the data.Results:1.The value range and characteristics of different tongue colors in Lab color space: the L values of different tongue colors ranged from 39 to 75,a value ranges from 17 to 40;b values range from-10 to 18.The specific values of L,a and b of each tongue color are as follows:(1)The value of B of purple tongue is all negative,which is obviously different from that of other tongue colors.(2)There is a difference in the range of a value between the pale white tongue and the pale red tongue.The maximum a value of the pale white tongue group is less than the minimum a value of the pale red tongue group.(3)The values of a and L are obviously different between pale white tongue and pale white tongue: the minimum L value of pale white tongue is greater than that of pale white tongue;The maximum A value of pale tongue is less than the minimum A value of pale tongue.(4)There were statistical differences in the distribution of L value and A value between Pale and Red tongues.2.Lab value based tongue color determination and artificial determination of similarity comparison: in 50 cases of tongue images,46 cases of the same determination results,the same results accounted for 92% of the total cases.The difference between the two methods was found near the boundary value of red tongue,light red tongue and purple tongue images.The accuracy of tongue color discrimination near the boundary value exceeded the sensitivity of human eyes to color,resulting in the difference.3.The distribution characteristics of tongue color were as follows:(1)Of the 109 cases of tongue color,there were 40 cases of pale red tongue,40 cases of red tongue,22 cases of purple tongue,5 cases of pale white tongue and 2 cases of pale pale tongue.There was no significant difference in tongue color distribution between three stages(P>0.05).(2)The distribution characteristics of tongue shape: cracked tongue accounted for 45%,punctured tongue accounted for 49%,tooth mark tongue accounted for 14%,tongue petechiae petechiae accounted for 11%.There was statistical significance in the distribution difference of puncturing tongue in the third stage,the proportion of puncturing tongue in the middle and late stage was less than that in the early stage,and there was statistical difference in the comparison between the middle and early stage and the late and early stage(P<0.05).(3)Distribution characteristics of tongue coating color: 2 cases had no tongue coating,and 107 cases had tongue coating.In the tongue map,white tongue coating accounted for 76%,yellow tongue coating accounted for 22%,and other tongue coating colors,such as gray and black,were not found.There was no significant difference in tongue coating color between different stages(P>0.05).(4)There was no statistical significance in the distribution of liverwort at three stages,including thin-thickness,moisture-dryness,greasy and rot(P>0.05).(5)The distribution characteristics of sublingual colladerals: there was no statistical significance in the distribution of sublingual colladerals varicose or more than 2 branches(P>0.05).However,there were statistical differences in the distribution of sublingual bleeding points in the third stage,and the proportion of sublingual bleeding points in the late stage was higher than that in the early stage,and the difference was statistically significant(P<0.05).4.The general data and physical and chemical indexes of different tongue images were compared as follows:(1)There was no statistical significance between the general data of different tongue colors(P>0.05);(2)Urinary albumin excretion rate of purple tongue group was significantly higher than that of light red tongue group and red tongue group(P<0.05);The proportions of fasting blood glucose and dry or sticky stool in red tongue group were significantly higher than those in light red tongue and purple tongue group(P<0.05).(3)There was no statistical significance in comparison of general data of puncturing tongue with or without prickle tongue(P>0.05);(4)The levels of glomerular filtration rate,erythrocyte and hemoglobin in patients with a little prickle were higher than those without prickle.osmotic Plasma pressure,serum creatinine and urea levels were lower than those in non-prick tongue group,and the differences were statistically significant(P<0.05).(5)The erythrocyte and hemoglobin counts in the group with tooth marks were significantly lower than those in the group without tooth marks.(6)The age of the tongue group with ecchymosis was higher than that without ecchymosis,and the difference was statistically significant(P<0.05).(7)The white blood cell count of thick tongue coating was significantly higher than that of thin tongue coating(P<0.05).(8)The erythrocyte and hemoglobin counts in the group with sublingual bleeding spots were significantly lower than those in the group without.Conclusion:1.The tongue image determination strategy based on LAB color space has a higher similarity rate with physicians’ subjective judgment,verified the feasibility of the judgment idea,increased the transformation of tongue image from parameter descriptive statistics to qualitative descriptive statistics,it improves the accuracy and objectivity of qualitative description of tongue diagnosis research,it is conducive to the development of modern research on tongue diagnosis based on big data in the future.2.Tongue manifestations of diabetic kidney disease patients suggest that the pathogenesis includes excessive heart fire,blood stasis,spleen deficiency,dampness and turbidity,Yin deficiency,overheating of Yangming,and deficiency of qi and blood,etc.There are more patients with excessive heart fire in the early stage than in the middle and late stages,and more patients with insufficient Qi and blood in the late stage,from the perspective of tongue picture,it provides a basis for the unified classification of diabetic kidney disease in the future.3.In the early diagnosis of diabetic kidney disease,the tongue whether had prick should be observed.Its internal physical and chemical manifestations may be related to the compensatory mechanisms of increasing glomerular filtration rate,increasing creatinine and urea excretion,and increasing the contents of red blood cells and hemoglobin.Although it increased the excretion of waste,it also increased the burden of kidney.it provided more theoretical basis for the application of the method of clearing away the heart-fire in the treatment of early diabetic kidney disease from the perspective of tongue image.4.In traditional tongue diagnosis,the presence or absence of sublingual bleeding points has received less attention,and sublingual bleeding points are associated with lower hemoglobin and red blood cell count;Therefore,in the tongue diagnosis of diabetic kidney disease,especially advanced patients,attention should be paid to observe whether there are bleeding points under the tongue as one of the signs to assess the patient’s Qi and blood deficiency.5.The pathogenesis of blood stasis represented by purple tongue is related to high urinary albumin excretion rate,it reflects the influence of blood stasis pathogenesis on the progression of diabetic kidney disease;From the Angle of tongue image,it provides a theoretical basis for delaying the progress of glycokidney by promoting blood circulation and removing blood stasis.6.Red tongue is associated with elevated fasting blood glucose and dry or sticky stools,it provides the basis for treating high fasting blood glucose and dry or sticky stool by Qingyang Ming-heat from the Angle of tongue image.7 Thick tongue coating was correlated with the increase of white blood cell count,which confirmed that the thickness of tongue coating was one of the bases to reflect the advance and decline of pathogenic factors.The essential connotation of the pathogenesis of spleen deficiency represented by tooth marks and tongue may include the decrease of RBC and Hb in blood.