Clinical Application Of Retinal Vessel Geometric Parameters In The Diagnosis And Screening Of DPN

Objective: To explore the feasibility of retinal vascular geometric parameters for the diagnosis and Screening of DPN.Methods: 242 patients with type 2 diabetes were divided into 0 DPN group(n = 110),1 DPN group(n = 56),2 DPN group(n = 45)and 3 DPN group(n = 32),according to the severity of diabetic peripheral neuropathy.The clinical data of the patients were collected,the VPT was measured,and the fundus photography was performed.The fundus of the retina was analyzed by retinal image automatic analysis system,and the retinal vascular geometry parameters were obtained.One-way multivariate analysis of variance was performed in 0 DPN group,1 DPN group,2 DPN group and 3 DPN group.Univariate logistic regression analysis was used to determine the relative risk factors of DPN.And then analyzed their correlation with DPN staging.The ROC curve was used to compare the efficiency of the retinal vascular geometric parameters and VPT diagnosis of DPN.The Kappa value was calculated to compare the consistency of the retinal vascular geometric parameters and VPT diagnostic DPN.By paired Chi-square test,Compared retinal vascular geometric parameters on DPN’s detection rate with VPT on DPN’s detection rate.Results: 1.132(54.5%)were clinically diagnosed with DPN.2.The duration of diabetes,VPT,branching coefficient arterioles and branching coefficient veins increased with DPN progression,and there was significant difference((49)<0.05)between DPN groups.Fractal dimension arterioles decreased with DPN progression.There was significant difference((49)<0.05)between DPN groups.The branching angle arterioles,CRAE decreased with DPN progression first,then increased,and there was significant difference((49)<0.05)between DPN groups.There was no significant difference in CRVE between 0 DPN group and 1DPN group((49)= 0.319),but significant difference between other groups((49)<0.05).The fractal dimension veins of 3 DPN group was smaller than that of 2 DPN group,but the difference was not significant((49)= 0.390).There was significant difference((49)<0.05)in the other groups.3.DPN staging was linearly correlated with VPT(r = 0.818),CRVE(r =0.716)and fractal dimension arterioles(r =-0.769).4.VPT,CRVE,fractal dimension arterioles,and fractal dimension veins for the diagnosis of DPN have good sensitivity(80% vs 55% vs 82% vs 67%)and specificity(92% vs 93% vs 82% vs 80%).5.CRVE,fractal dimension arterioles,fractal dimension veins,and combination of the three geometric parameters of retinal vessels diagnostic DPN and VPT diagnostic DPN consistency is better,their Kappa values were 0.470,0.547,0.435,and 0.435,respectively.6.The detection rate of VPT to DPN was 47.52%,and the detection rate of combination of the three geometric parameters of retinal vessels to DPN was 64.88%.The difference between the two detection rate was significant((49)<0.001).Conclusion: 1.Retinal vascular geometry changes with a significant correlation between severity of DPN.2.VPT,CRVE,fractal dimension arterioles,fractal dimension veins for diagnosis of DPN with high efficiency.CRVE,fractal dimension arterioles,fractal dimension veins,and combination of the three geometric parameters of retinal vessels have obvious consistency with VPT diagnosis of DPN.The detection rate of combination of the three geometric parameters of retinal vessels was greater than that of VPT.3.Detection of geometric parameters of retinal vessels can be used for clinical diagnostic screening DPN.