Tell me how you use your smartphone and I’ll tell you if you’re depressed. It sounds like an impressive title, but it’s the latest invention from Silicon Valley: an app that can diagnose mood disorders from the way we use our mobile phone.

Mindstrong Health, a startup founded in Palo Alto, California, by a team of doctors, including Tom Insel, former director of the National Institute of Mental Health in the United States, is trying to prove that our obsession with technology brings with it something positive: it can help people detect (and partly treat) depression, schizophrenia, bipolar disorder, post-traumatic stress disorder and substance abuse.

All thanks to an app, which would be able to collect data on people’s emotional health by analyzing how they use their smartphones. Once installed, the Mindstrong app monitors everything a person does when they have a phone in their hand: the way they type words on the screen, the frequency and positions of touches, the trajectories of “scrolling”… The data is then encrypted and analyzed using the so-called machine learning. The results are then shared with the patient and the doctor.

This is not the first time that technology has been used to treat mood and behavioral disorders: in the past, antianxiety apps and therapies based on video games have been used; moreover, there are studies that have also taken into consideration the photos published on Instagram as a sign of depression. Mindstrong promises to go one step further, because it analyzes not so much what we do over the phone, but how we do it.

To test the app, at the beginning of the project, about 5 years ago, 150 volunteers were involved in a neurocognitive evaluation to analyze, with established methods, aspects related to episodic memory (the way we remember events) and executive function (mental abilities that include the ability to control impulses, manage time and focus on a task): these are the brain functions that are weakened in people suffering from behavioral disorders. Volunteers then returned home with an app, installed on their smartphone, which began to collect data that it was sent to a remote server.

Analyzing the results, the researchers of the Mindstrong Health team claim to have found many connections that confirmed the results the classic tests administered to volunteers. For example, memory problems, which are common in brain disorders, could be detected by observing the speed with which we type words or analyzing the errors we make (and the frequency with which we eliminate certain characters), but also by measuring the speed with which we scroll down a list of contacts.

In the scientific community, the app generated a little skepticism. But Paul Dagum, founder of Mindstrong Health, says that thousands of people are already using the app.

And that the company, which has now put together 5 years of clinical study data that can confirm its theories, has been working with patients and doctors in clinics since last March. Not only that, but important research centres are also studying its potential.

An ongoing study at the University of Michigan is, for example, examining whether Mindstrong could be useful for people who do not suffer from a mental disorder, but who have a high risk of depression and suicide.

Led by Srijan Sen, a professor of psychiatry and neuroscience, the study monitors the moods of doctors in their first year of work across the country: they are high-risk people who experience situations of intense stress, with frequent sleep deprivation and high rates of depression.

Participants record their moods every day and wear a Fitbit activity detector to record data on sleep, activity and heart rate. And about 1,500 of them also use the Mindstrong app to collect data on how they type and understand how their mental conditions change throughout the year.

It is a study that will provide very important data: Professor Sen assumes that people’s memory patterns and speed of thought change subtly before they realize they are depressed, although it is not easy to understand how long after symptoms occur, or what cognitive patterns can really predict depression.