Open AccessAutomated summarisation of SDOCT volumes using deep learning: Transfer learning vs de novo trained networks
Author(s) -
Bhavna J. Antony,
Stefan Maetschke,
Rahil Garnavi
Publication year - 2019
Publication title -
plos one
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.99
H-Index - 332
ISSN - 1932-6203
DOI - 10.1371/journal.pone.0203726
Subject(s) - optical coherence tomography , artificial intelligence , transfer of learning , deep learning , computer science , test set , modality (human–computer interaction) , set (abstract data type) , pattern recognition (psychology) , volume (thermodynamics) , data set , relevance (law) , clinical practice , medicine , ophthalmology , physics , quantum mechanics , political science , law , programming language , family medicine
Spectral-domain optical coherence tomography (SDOCT) is a non-invasive imaging modality that generates high-resolution volumetric images. This modality finds widespread usage in ophthalmology for the diagnosis and management of various ocular conditions. The volumes generated can contain 200 or more B-scans. Manual inspection of such large quantity of scans is time consuming and error prone in most clinical settings. Here, we present a method for the generation of visual summaries of SDOCT volumes, wherein a small set of B-scans that highlight the most clinically relevant features in a volume are extracted. The method was trained and evaluated on data acquired from age-related macular degeneration patients, and “relevance” was defined as the presence of visibly discernible structural abnormalities. The summarisation system consists of a detection module, where relevant B-scans are extracted from the volume, and a set of rules that determines which B-scans are included in the visual summary. Two deep learning approaches are presented and compared for the classification of B-scans—transfer learning and de novo learning. Both approaches performed comparably with AUCs of 0.97 and 0.96, respectively, obtained on an independent test set. The de novo network, however, was 98% smaller than the transfer learning approach, and had a run-time that was also significantly shorter.