Our Geographic Atrophy Intelligent Analytics (GAIA) algorithm is applied to Optical Coherence Tomography (OCT) b-scans acquired with the Heidelberg Spectralis from patients with Geographic Atrophy (Late Dry Age-related Macular Degeneration) exported from Heyex.
This is achieved by automated b-scan-level segmentation, quantification, for monitoring of progression and classification of Geographic Atrophy (GA) on Spectralis OCTs following the Classification of Atrophy Meeting Report consensus definitions, including:
- incomplete Retinal Pigment Epithelium (RPE) and Outer Retinal Atrophy (iRORA)
- complete RPE and outer retinal atrophy (cORA)
- incomplete (iORA) and complete (cORA) Outer Retinal Atrophy
As shown on the above figure and in our publication, GAIA can also segment GA into its constituent parts:
- Photoreceptor Degeneration (PRD)
- RPE loss
- Hyper transmission.
Deep learning model implemented in PyTorch trained on clinical trials data of Heidelberg scans from patients with Geographic Atrophy and validated in an independent, external, real-life dataset demonstrating human-level performance. The model runs on AWS lambda. It takes as input a macular OCT volume and produces segmentation maps.
A PDF report is returned which contains B-scans visualisation of the segmentation output (GA and its constituent features) along with summary metrics. Embedded in the PDF report are hyperlinks to the GAIA Visualisation Platform which allows interactive 3D visualisation of the segmented volume scan.