Deep learning based retinal OCT segmentation

M Pekala; N Joshi; T Y Alvin Liu; N M Bressler; D Cabrera DeBuc; P Burlina

doi:10.1016/j.compbiomed.2019.103445

Deep learning based retinal OCT segmentation

Comput Biol Med. 2019 Nov:114:103445. doi: 10.1016/j.compbiomed.2019.103445. Epub 2019 Sep 17.

Authors

M Pekala¹, N Joshi¹, T Y Alvin Liu², N M Bressler², D Cabrera DeBuc³, P Burlina⁴

Affiliations

¹ Johns Hopkins University Applied Physics Laboratory, Laurel, MD, USA.
² Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore, MD, USA.
³ Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA.
⁴ Johns Hopkins University Applied Physics Laboratory, Laurel, MD, USA; Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore, MD, USA. Electronic address: pburlin2@jh.edu.

PMID: 31561100
DOI: 10.1016/j.compbiomed.2019.103445

Abstract

We look at the recent application of deep learning (DL) methods in automated fine-grained segmentation of spectral domain optical coherence tomography (OCT) images of the retina. We describe a new method combining fully convolutional networks (FCN) with Gaussian Processes for post processing. We report performance comparisons between the proposed approach, human clinicians, and other machine learning (ML) such as graph based approaches. The approach is demonstrated on an OCT dataset consisting of mild non-proliferative diabetic retinopathy from the University of Miami. The method is shown to have performance on par with humans, also compares favorably with the other ML methods, and appears to have as small or smaller mean unsigned error (equal to 1.06), versus errors ranging from 1.17 to 1.81 for other methods, and compared with human error of 1.10.

Keywords: Fully convolutional networks; Gaussian process regression; Neurodegenerative; OCT segmentation; Retinal and vascular diseases.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Algorithms
Deep Learning*
Diabetic Retinopathy / diagnostic imaging
Humans
Image Interpretation, Computer-Assisted / methods*
Retina / diagnostic imaging*
Tomography, Optical Coherence / methods*