Machine learning to identify your ‘anal fingerprint’ and artificial vision for feces: the "revolution" the smart toilet is as comical as it is necessary

When Lewis and Heaton presented the Bristol Stool Scale in 1997 at the Scandinavian Journal of GastroenterologyThey could sense that their idea was going to become one of the most widely used clinical and experimental scales in the world; but by no means could they suspect that they were falling short. Seven types of stool (ranging from “sheep droppings” to “totally liquid” through all kinds of “sausages”) seemed enough.

But it’s never enough. We are, if you allow me the expression, insatiable beings. So it is not surprising that in recent years research groups and companies around the world have started a race to find the definitive standard for analyzing (in real time) things like quantity, density, volume, clarity, the consistency or color of human feces. In real time and with deep learning.