ABSTRACT
Purpose: To explore and validate the utility of machine learning (ML) methods using a limited sample size to predict changes in visual acuity and keratometry 2 years following corneal crosslinking (CXL) for progressive keratoconus. Methods: The study included all consecutive patients with progressive keratoconus who underwent CXL from July 2014 to December 2020, with a 2 year follow-up period before July 2022 to develop the model. Variables collected included patient demographics, visual acuity, spherical equivalence, and Pentacam parameters. Available case data were divided into training and testing data sets. Three ML models were evaluated based on their performance in predicting case corrected distance visual acuity (CDVA) and maximum keratometry (Kmax) changes compared to actual values, as indicated by average root mean squared error (RMSE) and R-squared (R2) values. Patients followed from July 2022 to December 2022 were included in the validation set. Results: A total of 277 eyes from 195 patients were included in training and testing sets and 43 eyes from 35 patients were included in the validation set. The baseline CDVA (26.7%) and the ratio of steep keratometry to flat keratometry (K2/K1; 13.8%) were closely associated with case CDVA changes. The baseline ratio of Kmax to mean keratometry (Kmax/Kmean; 20.9%) was closely associated with case Kmax changes. Using these metrics, the best-performing ML model was XGBoost, which produced predicted values closest to the actual values for both CDVA and Kmax changes in testing set (R2 = 0.9993 and 0.9888) and validation set (R2 = 0.8956 and 0.8382). Conclusion: Application of a ML approach using XGBoost, and incorporation of identifiable parameters, considerably improved variation prediction accuracy of both CDVA and Kmax 2 years after CXL for treatment of progressive keratoconus.
ABSTRACT
AIM: To observe the characteristics of iridociliary cysts in myopic patients and evaluate the influences on the position and safety of implantable collamer lens (ICL) after surgery. METHODS: Totally 270 eyes of 135 patients who underwent ICL surgery for the corrections of myopia were included in this study. Preoperative and postoperative morphology of iridociliary cysts were observed in ultrasonic biomicroscopy (UBM) image. RESULTS: A total of 138 iridociliary cysts were found in 88 eyes of 50 patients among 270 eyes of 135 patients before surgery (37%). Twenty-five patients had cysts in one eye (50%) and 25 had cysts in both eyes (50%). The prevalence of iridociliary cysts was negatively correlated with age, but no gender difference (P>0.05). The incidence of iridociliary cysts was much less in eyes with myopia greater than -9.00 D (P<0.05). The diameter of the largest cyst was 1.96 mm and the smallest cyst was 0.24 mm, with a majority within the range of 0.5 to 1.0 mm. Most of the cysts were located in the inferior temporal quadrant. One year after ICL implantation, 51 iridociliary cysts (37%) remained unchanged, 47 cysts (34%) decreased in size, and 40 cysts (29%) disappeared. Most of cysts that changed after surgery were smaller than 1.0 mm (P<0.05) and located in the nasal and temporal sides around the haptics of implantable lens. All the ICL were in their original position. CONCLUSION: Iridociliary cysts are commonly seen in myopic eyes. The cysts have no impact on the safety of ICL surgery. Some cysts may decrease in size or disappear after ICL implantation.
ABSTRACT
PURPOSE: To identify associations between crystalline lens-related factors and central vault after Implantable Collamer Lens (ICL) (Staar Surgical, Monrovia, CA) implantation. METHODS: This retrospective clinical study included 320 eyes from 186 patients who underwent ICL implantation surgery. At 1 year after surgery, the central vault was measured using anterior segment optical coherence tomography. Preoperative anterior chamber depth, lens thickness, lens position (lens position = anterior chamber depth + 1/2 lens thickness), and vault were analyzed to investigate the effects of lens-related factors on postoperative vault. RESULTS: The mean vault was 513 ± 215 µm at 1 year after surgery. Vault was positively correlated with preoperative anterior chamber depth (r = 0.495, P < .001) and lens position (r = 0.371, P < .001), but negatively correlated with lens thickness (r = -0.262, P < .001). Eyes with vaults of less than 250 µm had shallower anterior chambers, thicker lenses, and smaller lens position than eyes in the other two vault groups (which had vaults ≥ 250 µm) (P < .001). Eyes with both anterior chamber depth less than 3.1 mm and lens position less than 5.1 mm had greatly reduced vaults (P < .001). CONCLUSIONS: The crystalline lens could have an important influence on postoperative vault. Eyes with a shallower anterior chamber and a forward lens position will have lower vaults. [J Refract Surg. 2017;33(8):519-523.].