Abstract
Purpose :
To demonstrate connectivity between the Fight Retinal Blindness! (FRB!) registry’s high quality clinical data with patients' corresponding multimodal images to enhance its research capabilities. Utilizing demigraphic and clinical data, in addition to images, unlocks increasingly personalized artificial intelligence (AI) data-driven insights.
Methods :
Adhering to RANZCO’s ethics approval and with informed patient consent, 2 FRB! sites in Sydney piloted this project. The pseudonymized tabular information (demographic and clinical data) routinely collected in FRB!’s age related macular degeneration module was migrated to a secure Australian-based cloud destination from locally hosted servers. Multimodal images from Heidelberg Engineering® devices, including ocular coherence tomography (OCT), infrared and autofluorescence were exported from pilot clinics. Corresponding FRB! pseudonymized identifiers were computationally inserted enabling linkage. Other personal information was removed.
Results :
Image-linkage was achieved in over 100 patients and enrolment continues. Tabular data was successfully moved to the cloud, captured as standardized SNOMED-CT concepts and converted to FHIR® (Fast Healthcare Interoperability Resource) format facilitating multiple interoperable capabilities. Images were exported in both ‘e2e’ and DICOM (Digital Imaging and Communications in Medicine) formats from Heyex 1 and Heyex 2 platforms respectively. Using a cloud-based instance of Heyex 2, each macular OCT produced 2 DICOM files; an image and an ‘ePDF’, which contained 22 OCT computer-readable measurements. These were also converted to FHIR and linked to FRB! tabular data.
Conclusions :
This process demonstrated successful longitudinal image-linkage at the patient-visit level, enriching FRB! research structure-function relationships and its potential for increased clinical impact on visual outcomes. This offers unprecedented possibilities for AI prognostication derived from multimodal data inputs realizable both at the point of care and for population health. This is increasingly important in our current rapidly changing paradigms with emerging treatments for retinal disease.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.