A non-linear multivariable regression method for the investigation of the correlation between central corneal thickness and HRTII optic nerve head topographic measurements in glaucoma patients

  • Authors:

  • D. Kourkoutas;G. J. Tsekouras;I. S. Karanasiou;N. E. Mastorakis;M. Moschos;E. Iliakis;G. Georgopoulos

  • Affiliations:

  • Department of Ophthalmology, Glaucoma Unit, Medical School, University of Athens, Goudi, Athens, Greece and Department of Ophthalmology, General Hospital, Athens, Greece;School of Electrical and Computer Engineering, National Technical University of Athens, Zografou, Athens, Greece and Department of Electrical Engineering & Computer Science, Hellenic Naval Aca ...;School of Electrical and Computer Engineering, National Technical University of Athens, Zografou, Athens, Greece;Department of Electrical Engineering & Computer Science, Hellenic Naval Academy, Piraeus, Greece;Department of Ophthalmology, Glaucoma Unit, Medical School, University of Athens, Goudi, Athens, Greece;Department of Ophthalmology, Glaucoma Unit, Medical School, University of Athens, Goudi, Athens, Greece;Department of Ophthalmology, Glaucoma Unit, Medical School, University of Athens, Goudi, Athens, Greece

  • Venue:
  • BEBI'08 Proceedings of the 1st WSEAS international conference on Biomedical electronics and biomedical informatics
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper a new non-linear multivariable regression method is presented to investigate the existence of correlation between the central corneal thickness (CCT) and the optic nerve head (ONH) topographic measurements in patients with glaucoma. Specifically, the model is supplied with Heidelberg Retina Tomograph II data including optic nerve head topographic parameters (i.e. disc area, etc) as well as independent parameters (i.e. patient age, refraction, etc), constituting the proposed algorithm's input variables. Each input variable is transformed using non-linear functions, such as xa, I/x, In (x), e-x, and new are created. The algorithm performs an extensive search in order to select the appropriate transformation functions of input variables to be used, by taking into consideration the correlation analysis of the transformed input variables. Following all possible non-linear multivariable models are tested and the best is chosen according to the correlation index R2 between the experimental and predicted values of the central corneal thickness satisfying the F-test and t-tests criterions synchronously.The results from the application of the described method are presented for ninety three eyes with open angle glaucoma and they are also compared to those obtained from the application of standard regression methods.