A woman has created an artificial intelligence baby translator app in hopes to help diagnose autism early.
Ariana Anderson is a mother of three. She realized by her third child that she was beginning to distinguish her baby's cries - the difference between hungry, tired, in pain, etc. This notion made Anderson wonder - could she make an algorithm that could listen to a baby's cries and categorize them?
Anderson created an app called "Chatterbaby". It collects data from user's babies and uses the "changes in frequency and patterns in the sound to silence ratio to tell parents why their kiddo is crying". More than 1700 babies have been added over the last five years, creating a data base of wails for the algorithm to pick through.
The app has not reached the capacity to decipher all cries yet, but Anderson is hopeful; the more baby cries that are entered in, the better her algorithm will get. So far it can tell cries from pain, hunger, and fussiness.
The long term hopes of this app and algorithm are high - if enough data is collected, Anderson plans to use it to diagnose autism in very young children.
Currently autism is hard to detect in early life as it is a vast spectrum with many facets. It is even harder to get a diagnosis if you live in parts of the world that suffer from extreme poverty, where there are no doctors to educate and inform parents. But the app has the potential to change all of that.
“We’re trying to bring the lab to the participants where they are,” Anderson explains.
The earlier that parents and educators know that a child has autism, the quicker they can shift day-to-day routines to better accommodate a child.
Chatterbaby is working with UCLA to collect audio data files from parents who use the app, once users sign consent forms that allow them to do so. Users are also asked information about their babies by survey, to help provide more details for the study.
Other researchers have proven that cries hold evidence with them; neurological clues that can signal an array of conditions.
Stephen Sheinkopf, who is a psychologist at the Brown University Center for the Study of Children at Risk, does warn however that there will be false positives. Though he is hopeful that AI could help diagnose, he is hesitant about the effects that eventual misdiagnoses could have on a family.
“We’re talking about changing a way a parent thinks about their child and interacts with them," says Sheinkopf. "Which in and of itself can influence their development”.
What do you think of this app? Let us know in the comments!