The problem came from Ichilov, Tel Aviv’s largest hospital, as one of the challenges at a Tel Aviv University hackathon: how do you monitor patients with age-related macular degeneration from a distance?
It matters because the disease can change form. When it moves from the dry form to the wet form it becomes far more serious, and it needs treatment quickly. But these patients are often elderly and dependent on others, on buses, without a car. Many of them do not even make it to the treatment they need.
I led a team, and I did not want to wait for the hackathon to start before understanding the problem.
Going before the start
I went to Ichilov with a teammate and found the phone number of a lead doctor in the clinic, cold. He was not available. A week later I went back alone and asked for him at reception. He was not there either, and they told me he would be back the following week.
So I used the time on the people the problem was about. I sat in the waiting room and talked to patients about what was hard for them, and started noticing problems in the place itself.
Filming the way there
A few days later I came back alone and filmed the route to the clinic. It was hard to find even for someone whose eyesight is fine. For someone losing the centre of their vision, it was a maze.
By luck I ran into an elderly veteran, a patient of the clinic, who remembered the way and showed it to me. I spent time with him and followed his day as he was moved from one doctor to the next, because every step there takes a long time.
Five hours later
In between I met a doctor from another department, who talked with me about the hospital's problems more broadly. I went to eat, came back, and started filming the way patients usually arrive, to show how hard it is just to get there.
I was not happy with the footage, so I filmed it one last time. And while the camera was running, I ran into the veteran again. It was five hours after I had met him in the morning. He was only just finishing, and he said so on camera.
Five hours, for one visit, for a patient who cannot drive and struggles to see the signs. That was the real problem behind the challenge. Remote monitoring is not a nicer version of the clinic. For these patients, it is the difference between being checked and not being checked at all.
What the hospital told me
The lead doctor I never managed to meet. Instead I cold-emailed a senior executive at Ichilov and got a meeting in person.
She told me the part that decides whether any of this works. Monitoring these patients is really hard, because they usually struggle to use ordinary apps at all. Very often it is a caregiver who handles their scheduling, not the patient. Whatever I built would have to get past the fact that most of the people it was for are not comfortable with technology.
There are many apps that put the standard home tests for macular degeneration on a screen. The hard question was never whether a test can run on a phone. It is whether the person it is for will ever open it.
What happened next
The project was never presented. My teammates and I disagreed about the direction, which I had built from the conversations with patients and clinicians, and I was removed from the team before the presentation.
I did not let it end there. Afterwards I put the app in front of about 40 patients with macular degeneration myself. It flagged the problems in their eyes, and most of them preferred the final version, the one built from what the waiting room had taught me, over the first. Some asked to keep using it, and I had no way to let them.
Every one of them already had the disease, so that shows the test responds to it, not that it can tell a healthy eye from a sick one. That needs a proper study, and this was one student testing alone.
Afterwards I tried to take the idea further on my own, and it stalled in the hospital’s bureaucracy. That taught me something too: if I wanted to change this, it would not happen inside one institution, and it would not happen quickly.
The work became Qeben, a home-monitoring research prototype for macular degeneration. It was never clinically validated and never deployed. What it kept from that day was the order of the questions: first, who gets to the test and who does not, and only then, what the test should measure.