Depression is a mental problem that urban people are prone to suffer from, and patients are often unaware that they have related symptoms. A start-up has developed an algorithm that can use the analysis of speech to judge the signs of depression.
The algorithm developed by Kintsugi, a start-up from California, can use voice and artificial intelligence analysis technology to find the symptoms of depression. Co-founder Grace Chang said that using the API makes it easier to identify people who might need treatment and direct help. The principle is that regardless of cultural background or language, depressed people tend to have flatter voices, and they collected tens of thousands of voice samples from depressed patients and established two scales to measure the likelihood of developing depression.
They initially made a voice diary app (same name as the startup itself: Kintsugi), and then collected a large number of voice samples for training models with the consent of users. Since they only care about how they speak, not what they say, there is no need for natural language processing and privacy concerns are avoided. Kintsugi has raised $28 million in funding and has contracts with several large medical companies to develop the technology.
Read more here: