ABSTRACT
Objective: The objective of this retrospective study was to employ machine learning techniques to examine age-related traits of ocular aberrations in a substantial population with myopia and myopic astigmatism. Methods: This was a cross-sectional study. Data from a population of myopic and myopic astigmatism patients who underwent wavefront aberration examinations at the Refractive Surgery Center of Tianjin Eye Hospital in Tianjin, China, were collected continuously from January 2013 to July 2017. The data from the right eye of each individual were collected for analysis. Each eye had 32 outcome data points, including age, best-corrected visual acuity (BCVA), lower-order aberrations (spherical diopter, cylindrical diopter, and astigmatic axis), and higher-order aberrations [Zernike coefficients and root mean square (RMS) of the third to sixth order aberrations] were analyzed. Higher-order aberrations were measured by Hartmann-Shack aberrometer. Results: The study included 1 507 subjects (1 507 eyes), comprising of 694 males and 813 females, with a mean age of (23.28±5.45) years. The findings demonstrated a decrease followed by an increase in most of the higher-order aberrations with age between 15-40 years. The minimum value points were observed in the age group of 25-30 years for RMS of total higher-order aberrations, 3rd RMS (with a confidence of 47.74% for range 1), 4th RMS (with a confidence of 86.01% for range 1), and trefoil aberrations (with a confidence of 56.38% for C33 and 73.25% for C3-3). The minimum value points were also observed in the age group of 30-35 years for primary spherical aberration (with a confidence of 56.10% for C40) and vertical coma-like aberration (with a confidence of 56.91% for C3-1). In contrast, astigmatism with the rule tended to decrease with age, while astigmatism against the rule and oblique astigmatism tended to increase (with a confidence of 88.66%, 84.71%, 81.07%, 79.67%, and 66.35% for astigmatism with the rule in different age groups). Conclusions: As age increases, the population with with-the-rule astigmatism decreases while the population with against-the-rule astigmatism increases. The high-order aberrations are the lowest in the 25-35 age group.
Subject(s)
Astigmatism , Myopia , Male , Female , Humans , Adult , Adolescent , Young Adult , Astigmatism/surgery , Visual Acuity , Retrospective Studies , Cross-Sectional Studies , Myopia/surgery , Cornea/surgery , Refraction, Ocular , Corneal TopographyABSTRACT
Objective: To investigate the diagnosis of normal cornea, subclinical keratoconus and keratoconus by artifical intelligence. Methods: Diagnostic study. From January 2016 to January 2019, who admitted to Tianjin Eye Hospital from 18 to 48 years old, with an average of (28.4±8.2) years of myopia patients in 2 018 cases. Two experienced ophthalmologists labeled keratoconus, subclinical keratconus and nomal cornea based on the topography. The data of 80% (1 615 cases) patients were randomly selected as the training set by computer random sampling method, and the data of 20% (403 cases) patients were used as the verification set. Using the Gradient Boosting Decision Tree (GBDT) algorithm to extract 28 corneal parameters, and establish an algorithm model to diagnose the corneal condition of the patient, verify the diagnostic accuracy of the model by using the 10-fold cross-validation method, and evaluate the model using the receiver operating characteristic curve. Sensitivity and specificity with the original labeling and ophthalmic resident labeling. Results: The diagnostic accuracy of the model was 95.53%. The area under the receiver operating characteristic curve (AUC) of the validation set was 0.996 6. The accuracy of the model for diagnosis of subclinical keratoconus and normal cornea was 96.67%, the AUC of the validation set was 0.993 6; the accuracy of diagnosis of keratoconus and normal cornea was 98.91%, and the AUC of the validation set was 0.998 2. The diagnostic accuracy of the model is 95.53%, which is significantly better than the resident's 93.55%. Conclusion: The model established by artifical intelligence can diagnose the subclinical keratoconus with high accuracy, which can greatly improve the clinical diagnosis efficiency and accuracy of young and primary ophthalmologists. (Chin J Ophthalmol, 2019, 55: 911-915).