An Analysis of the Areas Occupied by Vessels in the Ocular Surface of Diabetic Patients: An Application of a Nonparametric Tilted Additive Model
Abstract
:1. Introduction
2. Materials and Methods
2.1. Study Population
2.2. Photography
2.3. Statistical Method
3. Results
4. Discussion
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
Appendix A. R Code for Statistical Analysis
- #-----------------------------------------Data and bandwidths
- res<-read.csv("")
- var<-read.csv("")
- h_ll=c(2.845401,1.587784,9.152896,
- 108.1861,12.0652,0.5424099,0.5,0.2152706,8.7255)
- h_kt=c(5.5,1/0.4,50,1/0.01,1/0.15
- ,1/2,1/0.65,1/10,1/0.11)
- #-----------------------------------------Bootstrap function for CI
- library(boot)
- Qfun <- function(data, i){
- d <- data[i, ]
- return(quantile(d,probs = c(0.025, 0.975)))
- }
- #-----------------------------------------Calculating df
- library(psych)
- l_Tllp <- function(x,X,h,p)p*bp(x,h,X)/mean(bp(x,h,X))
- l_Tllp=Vectorize(l_Tllp,"x")
- #-----------------------------------------TAM fitting
- MSE=rep(NA,9)
- e=matrix(NA,length(res[,1]),9)
- df1=rep(NA,9) #df1=tr(t(S)*S)
- df2=rep(NA,9) #df2=tr(S)
- df3=rep(NA,9) #df3=tr(2*S-t(S)*S)
- micro=c("Age","BMI","HTN_,Dur","UPE","GFR",
- "A1C","BUN","S_cr","MAP")
- fmicro=c("f(Age)","f(BMI)","f(HTN_Dur)","f(UPE)",
- "f(GFR)","f(A1C)","f(BUN)","f(S_cr)","f(MAP)")
- par(mfrow=c(3,3))
- resid=matrix(NA,length(res[,1]),9)
- for (i in 1:9) {
- X=var[,i]
- Y=res[,i]
- n=length(Y)
- plot(X,Y,col=rgb(0.4,0.4,0.8,0.6),
- pch=16 , cex=1.3 ,ylab=fmicro[i],
- xlab =micro[i],ylim=c(-0.17,0.1))
- abline(lm(Y~X),col="darkgreen",lwd=3)
- resid[,i]=resid(lm(Y~X))
- curve(rn_ll(x,Y,X,h=h_ll[i]),min(X),max(X),
- add = T,col=2, lwd=3 ,ylim=c(-0.25,0.3))
- curve(rn_Iorder(x,h=h_kt[i],X,Y=Y),min(X),max(X)
- ,add=T,lwd=3,col="blue")
- theta_opt <- as.double(try(constrOptim(c(rep(1/n,k-1),h_ll[i]),
- targetfunc,gr=NULL,ui=ui,ci=ci,Y=Y,X=X,$par,silent=T))
- pk <- k/n-sum(theta_opt[-k])
- p <- rep(c(theta_opt[-k],pk),each=n/k)
- curve(rn_Tllp(x,Y,X,theta_opt[k],p),min(X),max(X),
- add=T,lwd=3,col="red",ylab=fmicro[i], xlab =micro[i])
- #rug(X, ticksize = 0.06, side = 1, lwd =1)
- df2[i]=tr(l_Tllp(X,X,h=theta_opt[k],p=p))
- e[,i]=Y-rn_Tllp(X,Y,X,theta_opt[k],p)
- MSE[i]=mean((e[,i])^2)
- eb=(e[,i]-mean(e[,i]))/sd(e[,i]) #standardized e for bootstrapping
- data <- data.frame(eb)
- bo <- boot(data[,"eb", drop = FALSE], statistic=Qfun, R=5000)
- lcb= rn_Tllp(X,Y,X,theta_opt[k],p)+mean(bo$t[,1])*sqrt(MSE[i])
- ucb= rn_Tllp(X,Y,X,theta_opt[k],p)+mean(bo$t[,2])*sqrt(MSE[i])
- myPredict <- cbind(rn_Tllp(X,Y,X,theta_opt[k],p),lcb,ucb)
- ix <- sort(X,index.return=T)$ix
- polygon(c(rev(X[ix]), X[ix]), c(rev(myPredict[ ix,3]),
- myPredict[ ix,2]), col = rgb(0.7,0.7,0.7,0.4) , border = NA)
- }
- #---------------------------------Ftest linearity vs. nonlinearity
- e2=e^2
- dev <- apply(e2, 2, sum)
- resid2=resid^2
- devlm<- apply(resid2, 2, sum)
- fstat=rep(NA,9)
- pvalue=rep(NA,9)
- for (i in 1:9) {
- fstat[i]=((devlm[i]-dev[i])*(168-df2[i]))/((dev[i])*(df2[i]-1))
- pvalue[i]=1-pf(abs(fstat[i]),df2[i]-1,168-df2[i])
- }
- RSSlm=devlm
- RSS=dev
- table=cbind(micro,RSSlm,RSS,df2,fstat,pvalue)
- #-----------------------------------------Residuals normality plot
- par(mfrow=c(1,2))
- hist(e[,5], xlab="residuals",main= "Histogram⊔of⊔Model⊔Residuals")
- qqnorm(e[,5], pch = 1,frame = FALSE)
- qqline(e[,5], col = "steelblue", lwd = 2)
- shapiro.test(e[,5])
References
- World Health Organization. Definition and Diagnosis of Diabetes Mellitus and Intermediate Hyperglycaemia: Report of a WHO/IDF Consultation; World Health Organization: Geneva, Switzerland, 2006. [Google Scholar]
- Saeedi, P.; Petersohn, I.; Salpea, P.; Malanda, B.; Karuranga, S.; Unwin, N.; Colagiuri, S.; Guariguata, L.; Motala, A.A.; Ogurtsova, K.; et al. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: Results from the International Diabetes Federation Diabetes Atlas. Diabetes Res. Clin. Pract. 2019, 157, 107843. [Google Scholar] [CrossRef] [Green Version]
- Fowler, M.J. Microvascular and macrovascular complications of diabetes. Clin. Diabetes 2008, 26, 77–82. [Google Scholar] [CrossRef] [Green Version]
- Mendes, A.L.; Miot, H.A.; Haddad Junior, V. Diabetes mellitus and the skin. An. Rasileiros Dermatol. 2017, 92, 8–20. [Google Scholar] [CrossRef] [Green Version]
- Kumar, M.; Mishra, L.; Mohanty, R.; Nayak, R. Diabetes and gum disease: The diabolic duo. Diabetes Metab. Syndr. Clin. Res. Rev. 2014, 8, 255–258. [Google Scholar] [CrossRef] [PubMed]
- Lee, W.B.; Mannis, M.J. Historical concepts of ocular surface disease. In Ocular Surface Disease: Cornea, Conjunctiva and Tear Film; Elsevier Inc.: Philadelphia, PA, USA, 2013; pp. 3–10. [Google Scholar]
- Alves, M.D.C.; Carvalheira, J.B.; Módulo, C.M.; Rocha, E.M. Tear Film and Ocular Surface Changes in Diabetes Mellitus. Arq. Bras. Oftalmol. 2008, 71, 96–103. [Google Scholar] [CrossRef] [Green Version]
- Ljubimov, A.V. Diabetic complications in the cornea. Vis. Res. 2017, 139, 138–152. [Google Scholar] [CrossRef]
- Owen, C.G.; Newsom, R.S.; Rudnicka, A.R.; Ellis, T.J.; Woodward, E.G. Vascular response of the bulbar conjunctiva to diabetes and elevated blood pressure. Ophthalmology 2005, 112, 1801–1808. [Google Scholar] [CrossRef]
- Cheung, A.; Hu, B.; Wong, S.; Chow, J.; Chan, M.; To, W.; Li, J.; Ramanujam, S.; Chen, P. Microvascular abnormalities in the bulbar conjunctiva of contact lens users. Clin. Hemorheol. Microcirc. 2012, 51, 77–86. [Google Scholar] [CrossRef]
- To, W.J.; Telander, D.G.; Lloyd, M.E.; Chen, P.C.; Cheung, A.T. Correlation of conjunctival microangiopathy with retinopathy in type-2 diabetes mellitus (T2DM) patients. Clin. Hemorheol. Microcirc. 2011, 47, 131–141. [Google Scholar] [CrossRef]
- Banaee, T.; Pourreza, H.; Doosti, H.; Abrishami, M.; Ehsaei, A.; Basiry, M.; Pourreza, R. Distribution of Different Sized Ocular Surface Vessels in Diabetics and Normal Individuals. J. Ophthalmic Vis. Res. 2017, 12, 361. [Google Scholar] [CrossRef]
- Jiang, H.; Zhong, J.; DeBuc, D.C.; Tao, A.; Xu, Z.; Lam, B.L.; Liu, C.; Wang, J. Functional slit lamp biomicroscopy for imaging bulbar conjunctival microvasculature in contact lens wearers. Microvasc. Res. 2014, 92, 62–71. [Google Scholar] [CrossRef] [Green Version]
- Khansari, M.M.; Wanek, J.; Tan, M.; Joslin, C.E.; Kresovich, J.K.; Camardo, N.; Blair, N.P.; Shahidi, M. Assessment of conjunctival microvascular hemodynamics in stages of diabetic microvasculopathy. Sci. Rep. 2017, 7, 45916. [Google Scholar] [CrossRef] [Green Version]
- Manchikanti, V.; Kasturi, N.; Rajappa, M.; Gochhait, D. Ocular surface disorder among adult patients with type II diabetes mellitus and its correlation with tear film markers: A pilot study. Taiwan J. Ophthalmol. 2020. [Google Scholar] [CrossRef]
- Di Zazzo, A.; Coassin, M.; Micera, A.; Mori, T.; De Piano, M.; Scartozzi, L.; Sgrulletta, R.; Bonini, S. Ocular surface diabetic disease: A neurogenic condition? Ocul. Surf. 2021, 19, 218–223. [Google Scholar] [CrossRef] [PubMed]
- Lindsey, J.K. Applying Generalized Linear Models; Springer Science & Business Media: New York, NY, USA, 2000. [Google Scholar]
- Nelder, J.A.; Wedderburn, R.W. Generalized Linear Models. J. R. Stat. Soc. Ser. A (Gen.) 1972, 135, 370–384. [Google Scholar] [CrossRef]
- Hastie, T.; Tibshirani, R. Generalized Additive Models for Medical Research. Stat. Methods Med Res. 1995, 4, 187–196. [Google Scholar] [CrossRef] [PubMed]
- Hastie, T.; Tibshirani, R. Generalized additive models: Some applications. J. Am. Stat. Assoc. 1987, 82, 371–386. [Google Scholar] [CrossRef]
- Hastie, T.J.; Tibshirani, R.J. Generalized Additive Models; CRC Press: Boca Raton, FL, USA, 1990; Volume 43. [Google Scholar]
- Boroumand, F.; Shakeri, M.T.; Kordzakhia, N.; Salehi, M.; Doosti, H. Tilted Nonparametric Regression Function Estimation. 2021. Available online: http://xxx.lanl.gov/abs/2102.02381 (accessed on 4 February 2021).
- Pourreza, R.; Banaee, T.; Pourreza, H.; Kakhki, R.D. A Radon transform based approach for extraction of blood vessels in conjunctival images. In Proceedings of the Mexican International Conference on Artificial Intelligence, Atizapan de Zaragoza, Mexico, 27–31 October 2008; Springer: Berlin/Heidelberg, Germany, 2008; pp. 948–956. [Google Scholar]
- Maurer, C.R.; Qi, R.; Raghavan, V. A linear time algorithm for computing exact Euclidean distance transforms of binary images in arbitrary dimensions. IEEE Trans. Pattern Anal. Mach. Intell. 2003, 25, 265–270. [Google Scholar] [CrossRef]
- Wasserman, L. All of Nonparametric Statistics; Springer Science & Business Media: New York, NY, USA, 2006. [Google Scholar]
- Buja, A.; Hastie, T.; Tibshirani, R. Linear Smoothers and Additive Models. Ann. Stat. 1989, 17, 453–510. [Google Scholar] [CrossRef]
- Samson, S.L.; Garber, A.J. Metabolic Syndrome. Endocrinol. Metab. Clin. 2014, 43, 1–23. [Google Scholar] [CrossRef]
- Hosojima, H.; Miyauchi, E.; Morimoto, S. Urinary Excretion of Angiotensin Converting Enzyme in NIDDM Patients with Nephropathy. Diabetes Care 1989, 12, 580–582. [Google Scholar] [CrossRef]
- Herman, A.A.; Hastie, T.J. An Analysis of Gestational Age, Neonatal Size and Neonatal Death Using Nonparametric Logistic Regression. J. Clin. Epidemiol. 1990, 43, 1179–1190. [Google Scholar] [CrossRef]
- Charytanowicz, M.; Kulczycki, P. Nonparametric Regression for Analyzing Correlation between Medical Parameters. In Information Technologies in Biomedicine; Springer: Berlin/Heidelberg, Germany, 2008; pp. 437–444. [Google Scholar]
- Seposo, X.T.; Dang, T.N.; Honda, Y. How does ambient air temperature affect diabetes mortality in tropical cities? Int. J. Environ. Res. Public Health 2017, 14, 385. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Hall, P.; Presnell, B. Intentionally biased bootstrap methods. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 1999, 61, 143–158. [Google Scholar] [CrossRef]
- Hall, P.; Huang, L.S. Nonparametric kernel regression subject to monotonicity constraints. Ann. Stat. 2001, 29, 624–647. [Google Scholar] [CrossRef]
- Carroll, R.J.; Delaigle, A.; Hall, P. Testing and estimating shape-constrained nonparametric density and regression in the presence of measurement error. J. Am. Stat. Assoc. 2011, 106, 191–202. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Doosti, H.; Hall, P. Making a non-parametric density estimator more attractive, and more accurate, by data perturbation. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 2016, 78, 445–462. [Google Scholar] [CrossRef]
- Doosti, H.; Hall, P.; Mateu, J. Nonparametric tilted density function estimation: A cross-validation criterion. J. Stat. Plan. Inference 2018, 197, 51–68. [Google Scholar] [CrossRef]
Predictor | Mean | Standard Deviation |
---|---|---|
Age | 47.79 | 3.70 |
Body Mass Index (BMI) (kg/m) | 28.66 | 4.81 |
Hypertension Duration (HTN-Dur) | 23.98 | 29.56 |
daily Urinary Protein Excretion (UPE) | 433.3 | 310.70 |
Glomerular Filtration Rate (GFR) | 80.14 | 25.76 |
Hemoglobin A1C | 8.824 | 1.85 |
Blood Urea Nitrogen (BUN) | 13.55 | 4.44 |
Fasting Blood Sugar (FBS)(mg/dL) | 166.20 | 54.76 |
Serum Creatinine (S-cr) (mg/dL) | 1.11 | 0.20 |
Mean Arterial Pressure (MAP) (mg/dL) | 100.06 | 15.43 |
Predictor | RSS L | RSS N | df N | F Value | p-Value |
---|---|---|---|---|---|
Age | 0.091 | 0.094 | 3.015 | −3.181 | 0.0437 * |
Body Mass Index (BMI) | 0.143 | 0.122 | 3.880 | 9.397 | <0.001 ** |
Hypertension Duration (HTN-Dur) | 0.091 | 0.095 | 4.821 | −2.023 | 0.097 |
Daily Urinary Protein Excretion (UPE) | 0.094 | 0.091 | 3.093 | 3.264 | 0.038 * |
Glomerular Filtration Rate (GFR) | 0.096 | 0.090 | 8.913 | 1.293 | 0.251 |
Hemoglobin A1C | 0.107 | 0.090 | 7.759 | 4.275 | <0.001 ** |
Blood Urea Nitrogen (BUN) | 0.181 | 0.099 | 8.736 | 17.085 | <0.001 ** |
Serum Creatinine (S-cr) | 0.093 | 0.093 | 8.227 | −0.146 | 0.995 |
Mean Arterial Pressure (MAP) | 0.114 | 0.116 | 6.770 | −0.364 | 0.895 |
Predictor | Estimate | Std. Error | t Value | p-Value |
---|---|---|---|---|
Intercept | 0.0654 | 0.0891 | 0.733 | 0.4646 |
Age | −0.002 | 0.001 | −2.373 | 0.019 * |
Gender (Female/male) | −0.003 | 0.007 | −0.365 | 0.716 |
Hypertension Duration (HTN-Dur) | 0.0002 | 0.0001 | 1.436 | 0.153 |
Glomerular Filtration Rate (GFR) | −1.232 × 10−6 | 1.791 × 10−4 | −0.007 | 0.994 |
Serum Creatinine (S-cr) | −0.036 | 0.0348 | −1.036 | 0.302 |
Mean Arterial Pressure (MAP) | 0.0005 | 0.0002 | 2.728 | 0.007 ** |
Area (2/1) | −0.0013 | 0.0057 | −0.231 | 0.818 |
Area (3/1) | 0.0123 | 0.0057 | 2.142 | 0.034 * |
Area (4/1) | 0.0075 | 0.0057 | 1.315 | 0.191 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Boroumand, F.; Shakeri, M.T.; Banaee, T.; Pourreza, H.; Doosti, H. An Analysis of the Areas Occupied by Vessels in the Ocular Surface of Diabetic Patients: An Application of a Nonparametric Tilted Additive Model. Int. J. Environ. Res. Public Health 2021, 18, 3735. https://doi.org/10.3390/ijerph18073735
Boroumand F, Shakeri MT, Banaee T, Pourreza H, Doosti H. An Analysis of the Areas Occupied by Vessels in the Ocular Surface of Diabetic Patients: An Application of a Nonparametric Tilted Additive Model. International Journal of Environmental Research and Public Health. 2021; 18(7):3735. https://doi.org/10.3390/ijerph18073735
Chicago/Turabian StyleBoroumand, Farzaneh, Mohammad Taghi Shakeri, Touka Banaee, Hamidreza Pourreza, and Hassan Doosti. 2021. "An Analysis of the Areas Occupied by Vessels in the Ocular Surface of Diabetic Patients: An Application of a Nonparametric Tilted Additive Model" International Journal of Environmental Research and Public Health 18, no. 7: 3735. https://doi.org/10.3390/ijerph18073735