June 2021
Volume 62, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2021
Large Scale Cross-Vendor Linkage of Ophthalmic Imaging and American Academy of Ophthalmology IRIS® Registry Clinical Data
Author Affiliations & Notes
  • Michael Mbagwu
    Ophthalmology, Stanford University, Stanford, California, United States
    United States Department of Veterans Affairs, Livermore, California, United States
  • Theordore Leng
    Ophthalmology, Stanford University, Stanford, California, United States
  • Anil Luthra
    Verana Health, California, United States
  • Durga Borkar
    Duke University Department of Ophthalmology, Durham, North Carolina, United States
  • Ashley Kras
    Verana Health, California, United States
  • Footnotes
    Commercial Relationships   Michael Mbagwu, Verana Health (C); Theordore Leng, Genentech (C), Kodiak (F), Regeneron (C), Targeted Therapy Technologies (F), Verana Health (C); Anil Luthra, Verana Health (E); Durga Borkar, Verna Health (C); Ashley Kras, Verana Health (E)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2021, Vol.62, 2289. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Michael Mbagwu, Theordore Leng, Anil Luthra, Durga Borkar, Ashley Kras; Large Scale Cross-Vendor Linkage of Ophthalmic Imaging and American Academy of Ophthalmology IRIS® Registry Clinical Data. Invest. Ophthalmol. Vis. Sci. 2021;62(8):2289.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : Ophthalmic diagnostic imaging is indispensable to clinical practice. Historically, metadata harmonization between different device manufacturers and modalities has been challenging due to insubstantial compliance with the universal Digital Imaging and Communications in Medicine (DICOM) standards, which presents hurdles to patient clinical record linkage. Here, we retrieved and integrated patient-level DICOM file meta-data from two major ophthalmic imaging manufacturers and linked them with corresponding clinical and demographic data in the American Academy of Ophthalmology IRIS® Registry (Intelligent Research in Sight).

Methods : DICOM files of six different posterior segment imaging modalities were acquired from multiple ophthalmic practices. Extracted metadata revealed both ‘standards conformant’ and ‘private manufacturer’ metadata tags. Following manual tag inspection, we implemented a programmatically scalable metadata harmonization process of consolidating the tags. This conformed to DICOM standards where possible. Adhering to HITRUST guidelines, proprietary heuristic patient matching algorithms linked these DICOM files with patient records in the IRIS® Registry.

Results : 1,842,633 DICOM files from 58,517 unique patients were available. 1,512,373 images from 48,443 unique patients were successfully linked to existing IRIS ® patients. The patient match rate was 82.78%; this likely resulted from a combination of image ‘test’ patient exclusion, patient detail mismatch between electronic systems, and possible algorithmic parameter aberrations. Each manufacturer's DICOM files contained 127 and 77 metadata tags respectively, both consisting of standard and private tags. The resulting consolidated list consisted of 170 metadata tags.

Conclusions : Using demographic identifiers from DICOM metadata, we created an automated pipeline to comprehensively and accurately connect longitudinal real-world clinical data to various image modalities from multiple manufacturers at the patient and visit level. This curation process has produced a large, enriched and multimodal IRIS® Registry, enabling enhanced research and advanced analytics. As imaging dependence and digital data capture grows, standards compliance will be critical to advancing the virtuous cycle of data-driven clinical insights to improve quality of care and enable novel ophthalmic drug and device development approaches.

This is a 2021 ARVO Annual Meeting abstract.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×