ESCRS - PP03.12 - Prospective Evaluation Of Postoperative Vault Following Icl (Implantable Collamer Lens) Implantation Based On A New Deep Learning-Based Algorithm For Predicting Icl Size.

Prospective Evaluation Of Postoperative Vault Following Icl (Implantable Collamer Lens) Implantation Based On A New Deep Learning-Based Algorithm For Predicting Icl Size.

Published 2025 - 43rd Congress of the ESCRS

Reference: PP03.12 | Type: Free paper | DOI: 10.82333/4fs1-gr14

Authors: Peter Fedor* 1

1Great Lakes Eye Consultants,Traverse City,United States

Purpose

  • To compare the size of the ICL suggested by OCOS to the size of the ICL suggested by our algorithm.
  • To define the predictability of postoperative vault using our developed algorithm.
  • To define the incidence of low vault (<250 microns) and high vault (>750 microns) using our developed algorithm.

Setting

Rothschild Foundation Hospital, Paris, France

Methods

Eight eyes from forty patients eligible for ICL implantation were included in this study. The implanted ICL size was chosen based on the size suggested by the algorithm. Operated patients were followed at day 1, day 7, 1 month, 3 months, and up to 6 months post-operatively. The ICL size suggested by the algorithm was compared to the size suggested by the OCOS platform. Postoperative vaults were recorded at each visit.

Results

Among the 80 included eyes, the algorithm suggested a different ICL size than the OCOS platform in 42 eyes (70% of cases). The average postoperative vault obtained was 315 microns at day 1, 479 microns at day 7, 464 microns at one month, 414 microns at 3 months, and 407 microns at 6 months post-operatively. No cases of low vault (<250 microns) or high vault (>750 microns) were found. No cases of ICL exchange were reported.

Conclusions

The developed algorithm allows for the prediction of ICL size and postoperative vault in a predictable, reproducible, and highly safe manner.