ESCRS - MP21 - INTRODUCING THE LENS DIAMETER TO IOL ACD PREDICTION (PART OF THE 1000 EYES LENS POWER TRIAL)

INTRODUCING THE LENS DIAMETER TO IOL ACD PREDICTION (PART OF THE 1000 EYES LENS POWER TRIAL)

Published 2024 - 28th ESCRS Winter Meeting

Reference: MP21 | Type: Presented Poster & Poster | DOI: 10.82333/mfcb-qk41

Authors: Leon Pomberger* 1 , Lisa Tasch 1 , Haidar Khalil 1 , Klemens Waser 1 , Nino Hirnschall 1 , Matthias Bolz 1

1Ophthalmology,Kepler University Clinic,Linz,Austria;Ophthalmology,Johannes Kepler University,Linz,Austria

Purpose

To develop an artificial intelligence-based method to predict the postoperative anterior chamber depth, leveraging preoperative lens diameter as a predictive feature, and to investigate the impact of preoperative lens diameter on postoperative refraction.

Setting

Department of Ophthalmology and Optometry, Johannes Kepler University, Linz, Austria; Department of Ophthalmology, Medical University Graz, Austria; Department of Ophthalmology, Medical University Sankt Poelten, Austria

Methods

This multi-center study included 1000 patients that underwent cataract surgery between July 2021 and September 2023. All eyes received a swept-source OCT biometry (IOL Master 700, Carl Zeiss Meditec AG, Germany and CASIA2, Tomey, Japan) followed by implantation of a monofocal hydrophobic open loop haptic IOL (ZCB00, or ICB00, Johnson&Johnson, USA). Study examination was performed minimum 4 weeks and maximum 24 months after surgery and included optical biometry, autorefraction as well as subjective refraction. The primary objective of this investigation was to assess the prognostic significance of the preoperative lens diameter in relation to the postoperative anterior chamber depth. This assessment was conducted utilizing an artificial intelligence-based model where Partial least square regression and a random forest plot were used.

Results

In total, 1000 eyes of 1000 patients were included at three recruitment centers. Preliminary data of 143 patients were used for this abstract. Partial least squares regression identified preoperative anterior chamber depth (mean 3.07mm; SD 0.40mm), lens diameter (mean 10.18; SD 0.36mm), and eye length (mean 23.53mm; SD 1.41mm) as the most important predictors of postoperative anterior chamber depth. These three parameters were used to train a machine learning model known as a Random Forest plot (OOB 0.040).The ACD values predicted by the machine learning method showed a significant correlation with the ACD measured postoperatively. The study demonstrated a significant relationship between preoperative lens diameter and postoperative anterior chamber depth (ACD).

Conclusion

The recently developed artificial intelligence-based model can predict the postoperative ACD with greater precision compared to previous approaches. As a result, this model could potentially lead to a reduction in the incidence of refractive surprises following cataract surgery.