ABSTRACT
Purpose: The purpose of this study was to identify and analyze the clinical and ocular surface risk factors influencing the progression of keratoconus (KC) using an artificial intelligence (AI) model. Methods: This was a prospective analysis in which 450 KC patients were included. We used the random forest (RF) classifier model from our previous study (which evaluated longitudinal changes in tomographic parameters to predict “progression” and “no progression”) to classify these patients. Clinical and ocular surface risk factors were determined through a questionnaire, which included presence of eye rubbing, duration of indoor activity, usage of lubricants and immunomodulator topical medications, duration of computer use, hormonal disturbances, use of hand sanitizers, immunoglobulin E (IgE), and vitamins D and B12 from blood investigations. An AI model was then built to assess whether these risk factors were linked to the future progression versus no progression of KC. The area under the curve (AUC) and other metrics were evaluated. Results: The tomographic AI model classified 322 eyes as progression and 128 eyes as no progression. Also, 76% of the cases that were classified as progression (from tomographic changes) were correctly predicted as progression and 67% of cases that were classified as no progression were predicted as no progression based on clinical risk factors at the first visit. IgE had the highest information gain, followed by presence of systemic allergies, vitamin D, and eye rubbing. The clinical risk factors AI model achieved an AUC of 0.812. Conclusion: This study demonstrated the importance of using AI for risk stratification and profiling of patients based on clinical risk factors, which could impact the progression in KC eyes and help manage them better
ABSTRACT
Purpose: To develop a nomogram in cases with mismatch between subjective and Topolyzer cylinder, and based on the magnitude of the mismatch, customize a treatment plan to attain good visual outcomes post?laser?assisted in situ keratomileusis (LASIK) surgery. Methods: The patients were evaluated preoperatively using corneal tomography with Pentacam. Five optimal corneal topography scans were obtained from the Topolyzer Vario were used for planning the LASIK treatment. For the nomogram purpose, the patients were divided into three categories based on the difference between the subjective cylinder and Topolyzer (corneal) cylinder. The first group (group 1) consisted of eyes of patients, where the difference was less than or equal to 0.4 D. The second group (group 2) consisted of eyes, where the difference was more than 0.4 D and the subjective cylinder was lesser than the Topolyzer cylinder. The third group (group 3) included eyes where the difference was more than 0.4 D but the subjective cylinder was greater than the Topolyzer cylinder. LASIK was performed with the WaveLight FS 200 femtosecond laser and WaveLight EX500 excimer laser. Assessment of astigmatism correction for the three groups was done using Aplins vector analysis. For comparison of proportions, Chi?square test was used. A P value less than 0.05 was considered statistically significant. Results: The UDVA was statistically significantly different when compared between groups 1 and 2 (P = 0.02). However, the corrected distance visual acuity (CDVA) was similar among all the three groups (P = 0.1). Group 3 showed an increase of residual cylinder by ?0.25 D, which was significant at intermediate and near reading distances (P < 0.05). Group 3 showed significantly higher target?induced astigmatism (TIA) compared to groups 1 and 2 (P = 0.01). The mean surgically induced astigmatism (SIA) was the least in group 2, which was statistically significant (P < 0.01). Conclusion: The outcomes for distance vision using our nomogram postoperatively were excellent, but further refinement for improving the near vision outcomes is required