Can AI Be Trained To Spot Depression In People’s Voices?

MIT scientists have trained an AI model to spot depression by analyzing the way people speak.

(Getty Images)
Getty Images/Westend61

Diagnosing depression is difficult, since there is no blood test, scan or biopsy that can provide hard evidence of the condition, which can be exhibited in multiple ways. The full weight of a diagnosis lies on the skill of a trained clinician to make an evaluation based mostly on a person’s responses to a series of standard questions.

But researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) think that artificial intelligence could help predict if a person is suffering from depression. Tuka Alhanai, a researcher at CSAIL, says that this is possible if a machine can learn a model that identifies speech and language patterns associated with depression. The AI developed by Alhanai and her team was able to recognize depression with a relatively high degree of accuracy by analyzing how people speak, rather than what they say to a clinician, according to Smithsonian Magazine.

“Instead of telling the model to focus on answers to particular questions, it’s programmed to figure out on its own what it wants to focus on,” she said to Smithsonian. 

The InsideHook Newsletter.

News, advice and insights for the most interesting person in the room.