A multimodal paired slit-lamp images and anterior phase optical coherence tomography dataset for anterior chamber biometry and cross-modality modelling

We describe a multimodal ophthalmic dataset comprising paired slit-lamp pictures, anterior phase optical coherence tomography (AS-OCT) pictures, and linked eye-level medical and biometric measurements collected at a single tertiary eye hospital. The launch incorporates 286 de-identified individuals and 562 eye-level information, together with 15,582 slit-lamp pictures and 9,092 AS-OCT pictures. Openly downloadable sources embrace de-identified tabular knowledge, a knowledge dictionary, and documentation, whereas imaging recordsdata are distributed by a separate controlled-access Zenodo file as a result of ocular pictures could retain residual re-identification danger. Records are linked throughout modalities utilizing non-meaningful patient_id and eye identifiers. We present dataset-level descriptive statistics, repository construction, de-identification procedures, and quality-control checks, along with a compact baseline experiment displaying that paired slit-lamp and AS-OCT information include usable cross-modal sign for anterior chamber depth prediction. This useful resource is meant to help reproducible benchmarking, cross-modality modelling, and secondary analysis use underneath clear entry governance and versioned launch practices.