This page is part of the HL7 FHIR® Implementation Guide: Ophthalmology Retinal, Release 1 (v0.1.0: STU1 Ballot 1) based on FHIR R4. . For a full list of available versions, see the Directory of published versions

Home Page

Introduction

“Eyes on FHIR ®” background and introduction Heading

Purpose

Despite FHIR’s® popularity and maturity accelerating as a community driven and implementer-focused modern flagship of healthcare standards to support interoperable healthcare data exchange, its application to medical specialties such as ophthalmology remains limited.

This is due to a core problem – there is no universal implementation guidance (IG), which is required to enable truly systemic interoperability that holds promise to advance clinical care and catalyze biomedical research. Hence, “Eyes on FHIR” initiative was formed with the overarching mission to develop the international standards that can enable FHIR ® to enhance the quality and outcomes of ophthalmic healthcare delivery.

To do this, “Eyes on FHIR” aims to represent the comprehensive ophthalmic clinical lexicon in FHIR ® format and through a series of compiled real world use cases, develop this IG to address the data exchange problems through semantically interoperable .

Please see the HL7 confluence page for more details and context here.

A notable achievement was demonstrated in May ‘21 connectathon, which showcased bidirectional real-time data exchange between various real world electronic medical record (EHR) vendors and fundamental ophthalmic diagnostic device and PACS manufacturers. This was a world first proof of concept, strongly endorsed by the NIH - please see the output connectathon report here for more details (link to PDF report) As a multimodal imaging-dependant specialty, there have already been significant artificial intelligence (AI) publications and implementations for automating the diagnosis and triaging of eye disease. A few cardinal publications are referenced here:

• Gulshan V, Peng L, Coram M, et al. Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs. JAMA. 2016;316(22):2402-2410. doi:10.1001/jama.2016.17216
• Poplin R, Varadarajan AV, Blumer K, Liu Y, McConnell MV, Corrado GS, et al. Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nature Biomedical Engineering. 2018;2: 158–164.
• De Fauw J, Ledsam JR, Romera-Paredes B, Nikolov S, Tomasev N, Blackwell S, et al. Clinically applicable deep learning for diagnosis and referral in retinal disease. Nat Med. 2018;24: 1342–1350.
• Yim J, Chopra R, Spitz T, Winkens J, Obika A, Kelly C, et al. Predicting conversion to wet age-related macular degeneration using deep learning. Nat Med. 2020;26: 892–899.

In addition to the outlined use cases, please see a proposed diagram here for how FHIR and DICOM compliment the workflow’s of clinical and imaging data respectively to optimally enable AI deployment. Use case document here

Authors

• Warren Oliver
• Ashley Kras
• Brett Esler