Segmentation of patients with cataract according to their ocular biometry: the LAKE classification.

PURPOSE: To create a classification for cataractous eyes based on their ocular biometry and thus provide a standard to stratify eyes when evaluating intraocular lens (IOL) formulas performance in different groups.

SETTING

High-resolution cataract surgery unit.

DESIGN

Monocentric, observational, retrospective study.

METHODS

Data of 21 797 patients (both eyes explored) were acquired between 2016 and 2024. An exploratory analysis on the most relevant biometric variables determined the presence of uncommon eyes, justifying the need of a stratification. Eyes were classified in 27 subgroups based on axial length (AL), anterior chamber depth (ACD), and mean keratometry (Km). Subgroups were named with 3 letters corresponding to 3 levels-high (H), normal (N), or low (L)-for AL, ACD, and Km, respectively. Cutoffs were created on average ± SD. Linear mixed models were used to evaluate differences between groups, where α < 0.05 meant significant.

RESULTS

43 954 eyes were classified, and all the subgroups contained at least 1 eye. The commonest subgroup was NNN (18 297 eyes, 41.97%), including eyes with (22.13 mm ≤ AL < 24.84 mm) AND (2.64 mm ≤ ACD < 3.43 mm) AND (42.50 diopters [D] ≤ Km < 45.60 D). Rarest subgroups were HLH and LHL (5 eyes and 1, respectively). Subgroups' characteristics (age, lens thickness [LT], LT/ACD, etc) reflect patterns seen frequently in clinical practice, for example, older patients with delayed surgeries, high LT and narrow ACD.

CONCLUSIONS

There is no consensus on how to subdivide eyes when evaluating IOL power formulas performance. LAKE classification addresses this need by providing a systematic method for categorizing cataractous eyes based on their biometry.