Image phenotyping

Deriving detailed and reproducible phenotypes from ophthalmic images

Ophthalmic images contain rich information about anatomy, disease, and variation between individuals. I develop and apply computational methods to derive detailed, reproducible phenotypes from these images at scale.

These phenotypes can support studies of disease mechanisms, improve the measurement of clinically important traits, and provide new ways to investigate relationships between the eye and the rest of the body.

This work includes the study of retinal layer thickness, foveal morphology, and deep learning-derived imaging features(Julian et al., 2025; Green et al., 2025; Julian et al., 2026).

Ophthalmic imaging provides complementary views of anatomy and vascular structure, including optical coherence tomography, colour fundus photography, and optical coherence tomography angiography.

References

2026

  1. Journal article
    Multi-omic analysis of deep learning-derived phenotypes links ophthalmic imaging to cardiovascular and neurological traits
    Thomas H Julian, Haoran Dou, Jinming Duan, and 15 more authors
    Nature Cardiovascular Research, 2026
    DOI

2025

  1. Journal article
    Pigmentation and retinal pigment epithelium thickness: a study of the phenotypic and genotypic relationships between ocular and extraocular pigmented tissues
    Thomas H Julian, Tomas Fitzgerald, UK Biobank Eye and Vision Consortium, and 2 more authors
    Pigment Cell & Melanoma Research, 2025
  2. Preprint
    Genetic insights into foveal morphology and its associations with pigmentation and age-related macular degeneration
    David J Green, David Romero-Bascones, Thomas H Julian, and 8 more authors
    medRxiv, 2025
    Preprint