Investigation Of Corneal Biomechanics And Development Of Corresponding Software System Based On Corneal Visualization Scheimpfiug Technology

Due to its important role in the study of refractive surgery and ophthalmic diseases such as keratoconus and glaucoma,corneal biomechanical parameters research based on Corneal visualization scheimpfiug technology（Corvis ST）have gradually become a hot topic in the field of ophthalmology.However,existing corneal biomechanical research lacked a systematic analysis and ignored the changes in corneal thickness.Therefore,this paper focuses on the study of corneal thickness changes and systematically analyzes the application of existing corneal biomechanical parameters to establish a hybrid parameter group for classification in ophthalmic diseases.In this paper,firstly,with the help of image processing methods with higher axial resolution,it was found that there was a change in central corneal thickness（CCT）with Corvis ST.Therefore,we proposed two new quantitative parameters for CCT changes:relative change of CCT(RC_CCT)and rate of change in CCT(ROC_CCT),and statistically analyzed RC_CCT and ROC_CCT extracted from normal corneal and keratoconus data.It is found that in most cornea RC_CCT and ROC_CCTCT do change during air puff indentation.In other words,the CCT would decrease during the indentation.Furthermore,it’s found that RC_CCT of keratoconus is greater than the normal group,while ROC_CCTCT is smaller in keratoconus than in normal group.But for both two parameters,there are significant differences in the data from normal and keratoconus cornea groups,which suggests that RC_CCT and ROC_CCT can potentially be used in diagnosis of keratoconus disease.Secondly,29 representative parameters(including RC_CCT and ROC_CCT)were selected from the existing corneal biomechanical parameters and grouped as:shape,deformation,and kinematic groups.After significant difference tests,multicollinearity tests,and logistic regression analysis,a hybrid set of corneal biomechanical parameter is established(including ROC_CCT,the peak distance and other 6 parameters),and its accuracy in the classification of normal cornea and keratoconus reached 99.3%（AUC=0.997,Sensitivity=100%,Specificity=98.5%,Overall correct detection rate=99.3%）.This parameter group is expected to become an important tool for clinical keratoconus screening.Finally,a corneal bio-mechanical test GUI（Graphical User Interface Development Environment）is developed on MATLAB platform,which can realize automatic corneal boundary extraction,extraction of 50 corneal biomechanical parameters and parameters 3D display,etc.In summary,the main content of this paper can be divided into three parts:the first part is the proposal of two new parameters related to corneal thickness changes;the second part is the establishment of an hybrid parameter group for corneal status classification through statistical analysis;The third part is the design of a software system that can automatically extract corneal biomechanical parameters and 3D display.