Hello, my name is Anat Loewenstein from the Tel Aviv Medical Center in Tel Aviv, Israel. I'm thrilled to be here at the ASRS in Montreal in the year 2026. This meeting is very exciting and I got the opportunity to present in the AMD session, which was one of the first sessions of the meeting. In my presentation, I talked about the validation of evaluation of a long-term change in detection of fluid comparing retina physicians with AI. What we were able to show in the past is that when you compare the home OCT with AI assessed trajectories to the office OCT, we have a very good repeatability, both for the presence and absence of retinal fluid. We were also able to show in the past that we have good agreement between the AI and retina specialists. Actually, the agreement between the AI and retina specialists and graders is better than among the graders themselves.

The question that came up for this study was to see if the AI detects a change in retinal fluid over longitudinal follow-up, will that change be detected also by physicians? And for that, we adopted a method from laboratory medicine that looks at a signal to noise ratio, which differs between different trajectories. If the trajectory of a specific patient is particularly noisy, then the change would be detected at a higher level as compared to a very stable trajectory where the change is detected at a lower level. Using this method, we were able to increase the sensitivity of the detection to 89%, a very high sensitivity, which was actually better than the regular population-based approach. This was achieved while maintaining a good specificity of 84%. So altogether we've been able to show that the AI enhanced home OCT follow up of the patients can detect change in fluid in a matter which is sensitive and specific.