Digital photos could help diagnose cancer in babies

Parents are known for filling up their phones with photo after photo of the best thing that’s ever happened to them/the reason they’ll never sleep again. But it turns out all those pictures of your new baby aren’t just a side effect of being obsessed with them, but also a potential tool for diagnosing cancer.

Bryan Shaw is a professor of chemistry at Baylor University in Texas. When his son Noah was four months old, he was diagnosed with retinoblastoma, a rare type of eye cancer that affects children younger than five. When the condition is diagnosed too late, it’s fatal: of 8,000 cases a year, around half don’t make it to their fifth birthday. Noah survived, but had to have an eye removed to ensure the tumours were all gone.

Bryan’s wife had previously expressed concern about Noah’s pupil looking white, and this symptom, known as leukocoria, turned out to be a tell-tale sign of retinoblastoma. When Bryan looked back at photos of his son, he saw that this was visible from when his son was 12 days old. And he realised that photos might help other parents spot the problem sooner, if they knew what to look for.

So, with the help of developers, he began working on software that will detect leukocoria. He’s now making an iPad app that can use the device’s in-built iris recognition to check a child’s eyes against a database of cases, and an iPhone version that automatically checks photos as they’re taken. They will be available later this summer, free to download, and he also has plans for a future desktop version that could be embedded on Facebook or Flickr.

Until now, the diagnostic tool for leukocoria has been a doctor quickly checking a baby’s eyes with a torch, but photos can show how a child looks consistently, in natural daylight, and the ability to check against other cases of the same symptom is invaluable. And not just for retinoblastoma: white pupils can be a sign of other illnesses, including cataracts and inflammatory conditions of the eye.

The challenge is that, as with other diagnosis apps, there is a chance of false positives. But if nothing else, it could help make parents aware that the problem might not be poor photography, but a poorly child – which will hopefully boost survival rates.

Image via Eric Fleming’s Flickr.

Diane Shipley