English

Inference of visual field test performance from OCT volumes using deep learning

Computer Vision and Pattern Recognition 2019-10-11 v3

Abstract

Visual field tests (VFT) are pivotal for glaucoma diagnosis and conducted regularly to monitor disease progression. Here we address the question to what degree aggregate VFT measurements such as Visual Field Index (VFI) and Mean Deviation (MD) can be inferred from Optical Coherence Tomography (OCT) scans of the Optic Nerve Head (ONH) or the macula. Accurate inference of VFT measurements from OCT could reduce examination time and cost. We propose a novel 3D Convolutional Neural Network (CNN) for this task and compare its accuracy with classical machine learning (ML) algorithms trained on common, segmentation-based OCT, features employed for glaucoma diagnostics. Peak accuracies were achieved on ONH scans when inferring VFI with a Pearson Correlation (PC) of 0.88±\pm0.035 for the CNN and a significantly lower (p << 0.01) PC of 0.74±\pm0.090 for the best performing, classical ML algorithm - a Random Forest regressor. Estimation of MD was equally accurate with a PC of 0.88±\pm0.023 on ONH scans for the CNN.

Keywords

Cite

@article{arxiv.1908.01428,
  title  = {Inference of visual field test performance from OCT volumes using deep learning},
  author = {Stefan Maetschke and Bhavna Antony and Hiroshi Ishikawa and Gadi Wollstein and Joel Schuman and Rahil Garnavi},
  journal= {arXiv preprint arXiv:1908.01428},
  year   = {2019}
}

Comments

12 pages, 3 figures

R2 v1 2026-06-23T10:39:24.373Z