Abstract
Purpose To determine real life quantitative changes in OCT biomarkers in a large set of treatment naive patients undergoing anti-VEGF therapy. For this purpose, we devised a novel deep learning based semantic segmentation algorithm providing, to the best of our knowledge, the first benchmark results for automatic segmentation of 11 OCT features including biomarkers that are in line with the latest consensus nomenclature of the AAO for age-related macular degeneration (AMD).
Design Retrospective study.
Participants Segmentation algorithm training set of 458 volume scans as well as single scans from 363 treatment naive patients for the analysis.
Methods Training of a Deep U-net based semantic segmentation ensemble algorithm leveraging multiple deep convolutional neural networks for state of the art semantic segmentation performance as well as analyzing OCT features prior to, after 3 and 12 months of anti-VEGF therapy.
Main outcome measures F1 score for the segmentation efficiency and the quantified volumes of 11 OCT features.
Results The segmentation algorithm achieved high F1 scores of almost 1.0 for neurosensory retina and subretinal fluid on a separate hold out test set with unseen patients. The algorithm performed worse for subretinal hyperreflective material and fibrovascular PED, on par with drusenoid PED and better in segmenting fibrosis. In the evaluation of treatment naive OCT scans, significant changes occurred for intraretinal fluid (mean: 0.03µm3 to 0.01µm3, p<0.001), subretinal fluid (0.08µm3 to 0.01µm3, p<0.001), subretinal hyperreflective material (0.02µm3 to 0.01µm3, p<0.001), fibrovascular PED (0.12µm3 to 0.09µm3, p=0.02) and central retinal thickness C0 (225.78µm3 to 169.40µm3).The amounts of intraretinal fluid, fibrovascular PED and ERM were predictive of poor outcome.
Conclusions The segmentation algorithm allows efficient volumetric analysis of OCT scans. Anti-VEGF therapy provokes most potent changes in the first 3 months and afterwards only acts as a stabilizing agent. Furthermore, a gradual loss of RPE hints at a progressing decline of visual acuity even beyond month 12. Additional research is required to understand how these accurate OCT predictions can be leveraged for a personalized therapy regimen.
Précis Novel high performance segmentation algorithm shows most volumetric changes under anti-VEGF therapy in oct biomarkers occur in the first 3 months. Afterwards the injections seem only to serve as a stabilizing agent.
Competing Interest Statement
Ben Asani has received speaker fees from Novartis Pharma GmbH. Olle Holmberg has ownership in Helsing GmbH. Christoph Kern has recieved travel fees from Bayer AG and Novartis GmbH. Consultant: Bayer AG. Karsten Kortuem has recieved speaker fees from Novartis GmbH, Bayer AG and Allergan. Jakob Siedlecki receives speaker honoraria and travel reimbursement from Carl Zeiss Meditec AG, Novartis Pharma GmbH, Bayer AG, Pharm-Allergan GmbH, Oculentis OSD Medical GmbH. Consultant: Bayer AG, Novartis Pharma GmbH, Pharm-Allergan GmbH. Travel reimbursement: Oertli AG. Fabian Theis consults for Immunai Inc., Singularity Bio B.V., CytoReason Ltd, and Omniscope Ltd, and has ownership interest in Dermagnostix GmbH and Cellarity. Siegfried Priglinger receives speaker and consultant honoraria from and has served on advisory boards for Abbott, Alcon, Geuder, Oculus, Schwind, STAAR, TearLab, Thieme Compliance, Ziemer, Zeiss and research funding from Abbott, Alcon, Hoya, Oculentis, Oculus, Schwind and Zeiss.
Funding Statement
This study was funded by a research grant from the German Federal Ministry of Education and Research (BMBF grant# 031L0210A) and by the Helmholtz Association Initiative and Networking Fund through Helmholtz AI [grant number: ZT-I-PF-5-01].
Author Declarations
I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
Yes
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
Ethics committee of the medical faculty of the Ludwig-Maximilians-University of Munich gave ethical approval for this work.
I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.
Yes
Footnotes
Financial support: This study was funded by a research grant from the German Federal Ministry of Education and Research (BMBF grant# 031L0210A) and by the Helmholtz Association Initiative and Networking Fund through Helmholtz AI [grant number: ZT-I-PF-5-01].
Conflict of interest: Ben Asani has received speaker fees from Novartis Pharma GmbH. Olle Holmberg has ownership in Helsing GmbH. Christoph Kern has recieved travel fees from Bayer AG and Novartis GmbH. Consultant: Bayer AG. Karsten Kortuem has recieved speaker fees from Novartis GmbH, Bayer AG and Allergan. Jakob Siedlecki receives speaker honoraria and travel reimbursement from Carl Zeiss Meditec AG, Novartis Pharma GmbH, Bayer AG, Pharm-Allergan GmbH, Oculentis OSD Medical GmbH. Consultant: Bayer AG, Novartis Pharma GmbH, Pharm-Allergan GmbH. Travel reimbursement: Oertli AG.
Fabian Theis consults for Immunai Inc., Singularity Bio B.V., CytoReason Ltd, and Omniscope Ltd, and has ownership interest in Dermagnostix GmbH and Cellarity. Siegfried Priglinger receives speaker and consultant honoraria from and has served on advisory boards for Abbott, Alcon, Geuder, Oculus, Schwind, STAAR, TearLab, Thieme Compliance, Ziemer, Zeiss and research funding from Abbott, Alcon, Hoya, Oculentis, Oculus, Schwind and Zeiss.
Data Availability
The code for the models and training procedures as well as result analysis will be made available through the public Github repository upon publication. All data produced in the present study are available upon reasonable request to the authors.