ESCRS - PP03.04 - Demographics Prediction And Heatmap Generation From Oct Images Of Anterior Segment Of The Eye: A Vision Transformer Model Study

Demographics Prediction And Heatmap Generation From Oct Images Of Anterior Segment Of The Eye: A Vision Transformer Model Study

Published 2022 - 40th Congress of the ESCRS

Reference: PP03.04 | Type: ESCRS 2022 - Posters | DOI: 10.82333/w10t-na04

Authors: Yun Jeong Lee* 1 , Sukkyu Sun 2 , Young Kook Kim 1

1Department of Ophthalmology,Seoul National University Hospital, Seoul National University College of Medicine,Seoul,Korea, Republic Of, 2Biomedical Research Institute, Seoul National University Hospital,Seoul,Korea, Republic Of

Purpose

To predict demographic characteristics from anterior segment optical coherence tomography (AS-OCT) images of eyes using a Vision Transformer (ViT) model.

Setting

Retrospective, cross-sectional study

Methods

Eligible participants (aged ≥40 years) were enrolled from a clinical data warehouse of a national tertiary hospital in South Korea. A total of 2,970 AS-OCT images were used to train, validate and test a ViT model to predict age and sex, and 2,616 images likewise were used for height, weight and body mass index (BMI). The main outcome measure was the area under the receiver operating characteristic (ROC) curve (AUC) of the ViT model.

Results

The ViT model achieved the largest AUC (0.910) for differentiating age ≤75 vs. >75 years, followed by age ≤60 vs. 60–75 vs. ≥75 years (AUC, 0.844), and for discriminating sex (AUC, 0.665). The prediction abilities for the other demographic characteristics were lower: an AUC of 0.521 for classifying height ≤170 vs. >170 cm in males and 0.479 for ≤155 vs. >155 cm in females; 0.522 for weight <70 vs. ≥70 kg in males and 0.503 for <55 vs. ≥55 kg in females, and 0.517 for BMI <23 vs. 23–25 vs. ≥25 kg/cm2. Heatmaps highlighted the area of the iridocorneal angle for its contribution to the prediction of age ≤75 vs. >75 years.

Conclusions

Although the ViT model demonstrated a good utility to classify age from AS-OCT images, it performed poorly for sex, height, weight and BMI. The heatmap obtained of the prediction will provide further clues to the understanding of age-related anterior segment changes in eyes.