https://doi.org/10.31288/oftalmolzh20144913

Application of the method of multiscale textural gradient for automatization of diabetes retinopathy diagnosis via analysis of digital images of the eye fundus

Kresyun N. V.1, Tatarchuk T. V.2, Shakun K. S. 2, Godlevsky L. S. 1

1  Odessa National Medical University,

2  Odessa National Maritime Academy; Odessa (Ukraine)

Key words: diabetic retinopathy, mi-croaneurism, textural gradient, digital images.

Introduction. Ophthalmoscopic detection of microaneurism is important for the diagnosis of diabetic retinopathy severity.

Aim of the investigation. To improve information value of digital images of the eye fundus using textural gradient technology application.

Material and Methods. 17practically healthy volunteers (30.5+3.2 years old) and 52 insulin — dependent patients. (31.7+2.7 years old) have been observed. The method consisted in getting ophthalmoscopic digital images in RGB color scale with satisfactory characteristics of the image such as brightness, color balance and contrast. The next step was transfer from two- dimensional space of color vector characteristics to the space of the textural gradient and presentation offinal images for expert diagnosis of microaneurism.

Results. Expert diagnosis of microaneurism, which was made after developed technology application revealed increase of the sensitivity by 14.7 % (P>0.05), while specificity rose 2.22 times when compared with the traditional method of diagnosis (P<0.05).

Conclusions. The developed method permitted to improve diagnosis of microaneurism mainly via exclusion of false-positive diagnosis.

 

References

1.     Gonzalez R, Woods R, Eddins S. Digital Image Processing using MATLAB (translated from English). M.; 2006. 616 p.

2.     Chochia PA. Pyramidal algorithm for image segmentation. Informatsionnyie protsessy. 2010;10(1): 23- 35. Russian.

3.     Mansberger SL, Gleitsmann K, Gardiner S et al. Comparing the effectiveness of telemedicine and traditional surveillance in providing diabetic retinopathy screening examinations : a randomized controlled trial. Telemed. J. e Health. 2013;19(12):942-8.
Crossref

4.     Chakrabarti D. Application of mobile technology in ophthalmology to meet the demands of low-resource settings. J. Mobile Technol. in Medicine. 2012;1(4S):1-3.
Crossref

5.     Fadzil MHA, Izhar LI, Venkatachalam PA, Karunakar TVN. Extraction and reconstruction of retinal vasculature. J. Medical Engineer. Technol. 2007;31:435-42.
Crossref

6.     Hani AFM, Nugroho HA. Retinal vasculature enhance­ment using independent component analysis. J. Biomed. Sci. Engineer. 2009;2:543-9.
Crossref

7.     Parisi V, Uccioli L. Visual electrophysiological responses in persons with type 1 diabetes. Diabetes Metab. Res. Rev. 2011;17(1):12-8.
Crossref

8.     Chin EK, Ventura BV, See KY et al. Nonmydriatic fundus photography for teleophthalmology diabetic retinopathy screening in rural and urban clinics. Telemed. J. e-Health. 2014;1:102-8.
Crossref

9.     Wu D, Zhang M, Liu JC, Bauman W. On the adaptive de­tection of blood vessels in retinal images. Biomed. Engi­neer., IEEE Transactions. 2006; 53:341- 3.
Crossref