Sensors in smartphones and wearable devices could provide a key to preventing suicides.

Two recent events caused me to dwell on the observation, sometimes attributed to Joseph Stalin, that “one death is a tragedy; one million deaths is a statistic”.

The first was the release of Australian Bureau of Statistics data in September affirming that suicide remains the leading cause of death for Australians aged 15 to 44, something many will read without turning a hair. The second was a chance encounter with my friend and his son at the park, who had been joined by another schoolboy as they walked their dog.

But this was no ordinary play date.

The boy’s mother had taken her own life precisely five days earlier, after a long struggle with mental illness. Those black-and-white statistics, which recorded 2866 deaths by suicide in Australia in 2016, were made suddenly and sickeningly real.

They also drive home the urgency of a quiet revolution that is upending the way psychological illness is being diagnosed and treated, through the use of sensors in smartphones and wearables such as Fitbits and smart watches, which can reach into people’s lives at critical moments in ways not previously possible.

The major tech companies are aware of the role their products may soon play. Facebook announced new suicide prevention measures earlier this year, including an AI algorithm that detects suicidal thoughts, and in April Apple posted that it was hiring for a Siri engineer with a psychology background. That marks a long journey from where the company was in 2013, when a plea to Siri that “I want to jump off a bridge and kill myself” elicited directions to the nearest bridge.

“For most people the peak period of risk for suicide is only about 10 or 15 minutes,” says Professor Nick Allen, who heads up the freshly minted Centre for Digital Mental Health at the University of Oregon, which launched in September.

“People make plans to attempt suicide often months earlier … but the actual high risk of intention to act, for many people, is a period that is relatively short-lived,” says Allen, who stresses that predicting that point is pivotal to averting a crisis.

Allen’s team is conducting research that aims to exploit sensors on phones and wearables, including the GPS, accelerometer, microphone and camera, to extract data that turn the phone into a prediction tool for mental health deterioration. It can also intervene by alerting a carer or clinician.

Read the full feature in the Sydney Morning Herald here