Abstract
Purpose :
Home-care OCT monitoring of AMD could improve early relapse recognition while cutting down the frequency of office appointments. This might improve treatment outcome, reduce disease burden and possibly cut down over-all disease management costs. However, so far no means has been found that can accomplish sufficient diagnostic accuracy at the cost limit of home-care solutions.
With Off-Axis Full-Field Time-Domain OCT (OA-FF-TD-OCT), our group has proposed an OCT technology that has the potential to be sold for less than 1.000 USD. However, in contrast to clinically used SD-OCT systems, the field of view is limited to about 4.5 x 1.5 mm and resolution and signal-to-noise ratio (SNR) are reduced.
With ever increasing quality of OCT imaging, the question which performance is actually needed for AMD diagnostic follow-up examinations is unanswered. Therefore, we performed detailed analysis about minimum requirements for reliable AMD monitoring. Furthermore, we analysed possibilities for automated analysis of the obtained imaging.
Methods :
First, we digitally degraded clinical high-quality OCT volume scans to be reduced in field of view, resolution and SNR to mimic the retinal images obtained by OA-FF-TD-OCT. Both original and degraded images were graded by a retina specialist and results were compared.
Furthermore, we retrospectively analysed OCT volume scans of AMD patients with respect to different biomarkers to determine their respective sensitivity and specificity in detecting disease activity.
Results :
In an analysis of 88 degraded OA-FF-TD-OCT mimicking volume scans, sensitivity for detecting disease activity was 98% and specificity was 100%, when compared to the original scan.
In an analysis of 764 volume scans, the following sensitivities/specificities for different biomarkers were found: Presence of subretinal fluid (SRF) 70%/100%; intraretinal fluid (IRF) 65%/82%; pigment epithelium detachment (PED) 95%/11%; either SRF or IRF 99%/82%. Changes in retinal volume could detect disease relapse with 90%/87%.
Conclusions :
OA-FF-TD-OCT allows cost-reduction while still providing image quality sufficient for AMD diagnostics. Biomarker and volume change detection are feasible approaches for automated evaluation. We are currently testing the device on AMD patients.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.