Skip to main content

The world’s first poop database needs your help

The world’s first poop database needs your help

/

Your poop can train artificial intelligence models

Share this story

World Toilet Day
Photo by Ralf Hirschberger / picture alliance via Getty Images

Artificial intelligence will soon be able to decode your poop. That’s the ultimate goal of a campaign to collect 100,000 fecal photos to build what developers say is the world’s first poop image database. The campaign, launched by microbial health company Seed, dares you to “give a shit” for science by uploading photos of your feces so that scientists can use it to train an AI platform launched out of MIT.

In this case, offering up your stool for science could potentially help the approximately 1 in 5 people in the US who have chronic gut conditions like irritable bowel syndrome, developers behind the database say. They’re hoping it will give doctors an extra diagnostic tool and allow patients to better manage their health.

offering up your stool for science could potentially help 1 in 5 people

“We don’t always think about stool as like a daily data — I’m putting air quotes around “dump” — but really, as a direct output of our gut health,” Ara Katz, co-founder and co-CEO of Seed Health, tells The Verge

Here’s how citizen scientists can contribute to the cause. To participate, go to seed.com/poop on your phone (because taking your laptop to the loo is weird, and the page doesn’t allow you to submit a photo unless you’re using your phone). Click on the big purple button that says “#GIVEaSHIT.” You’ll be prompted to enter your email address and whether you’re on a morning, afternoon, or evening poop schedule. Then, if you’ve already dropped a deuce, you can take or upload your photo or you can ask for an email reminder to be sent to you according to the time you indicated. After you’ve submitted your stool for posterity, the image is separated from the metadata (your email address and other potentially identifying information) so that your donation can remain anonymous and HIPAA compliant. 

doctors’ insights into your poop will help train artificial intelligence models

A team of doctors will diligently look through every image received. (Yes, that is a real job for seven gastroenterologists who take notes on what they see in the pictures.) Poop can fall into seven categories identified along the Bristol stool scale, which can tell you and your doctor whether you’re constipated, lacking fiber, have a serious case of the runs, or somewhere in between. The doctors’ insights into your poop will help train artificial intelligence models to understand the same things the doctors see in the image. Similar training systems are used to teach self-driving cars how to identify a tree or a cat in the road, according to David Hachuel, a co-founder of the startup Auggi, which is building the platform. 

Because there’s no existing database of real-life poop, Hachuel’s team began developing the app using Play-Doh they sculpted to mimic poop around the Bristol scale. “We spent countless hours just making different Play-Doh models,” Hachuel tells The Verge. “We actually 3D-printed a toilet just to emulate how that would happen in real life.” That was fun, he says, but time-consuming. To get the real thing, they first turned to Reddit and other sites to crowdsource photos of feces that people uploaded to the discussion board. Both Hachuel and Katz tell The Verge that you’d be surprised to see how many people are already online sharing and discussing their poop.

you’d be surprised to see how many people are already sharing pictures of their poop

Fun and forums aside, Hachuel says that the goal of the campaign is to eventually make the database an open-source tool for academic researchers. In the meantime, the hope is to empower patients to take back control of their own gut health. He says they’re building technology that helps people with chronic gut disorders better understand the relationship between their lifestyle and their symptoms. “They struggle every day making decisions on what to eat, how much exercise to do to keep their symptoms at bay,” he says. “And so it’s really critical to build this database and to develop these simple monitoring tools to allow those patients to essentially do that at home.”