Anatomical Determinants of Postoperative In-the-Bag IOL Position: Introducing a Novel Biometric Index.

Abstract

To investigate whether a novel preoperative biometric index that integrates lens geometry and the relative position of the capsular bag can predict the postoperative in-the-bag intraocular lens (IOL) position proportion coefficient and subsequently the final IOL position. Retrospective study. This study included 322 eyes of 322 patients undergoing uneventful cataract surgery with in-the-bag implantation of an aspheric Tecnis IOL (Johnson & Johnson Vision, Santa Ana, CA). Preoperative biometric assessments were performed in cataract surgery patients using swept-source anterior segment optical coherence tomography, encompassing conventional parameters (e.g., anterior chamber depth, lens thickness [LT], crystalline lens rise [CLR]) and composite indices such as the CLR to LT ratio (CLR/LT). Axial length (AL) and mean keratometry were measured using the IOLMaster 700. All parameters were analyzed for correlation with the in-the-bag IOL position proportion coefficient. The most predictive variables were incorporated into multiple linear and machine learning regression models to estimate this coefficient. The predicted coefficient was then used to calculate in-the-bag IOL position, and the accuracy was compared with conventional vergence formulas. Correlation between preoperative biometric parameters and the postoperative in-the-bag IOL position proportion coefficient; accuracy of in-the-bag IOL position estimation based on the predicted IOL position proportion coefficient, compared with actual postoperative IOL positions and with IOL positions derived from conventional vergence-based IOL power formulas across AL subgroups. CLR/LT demonstrated a strong association with postoperative IOL position proportion coefficient. A support vector machine model incorporating CLR/LT significantly improved the accuracy of IOL position proportion coefficient prediction. Compared with standard formulas, calculating IOL position via predicted IOL position proportion coefficient markedly enhanced prediction precision, particularly in eyes with atypical AL. CLR/LT is a clinically meaningful preoperative index for predicting in-the-bag IOL position. Its use enables more individualized IOL power selection and has the potential to improve refractive accuracy, especially in eyes with unusual biometric profiles.