'Smartwatch' sensors to help diagnose depression

Scientists in the UK and US are conducting clinical trials with depressed patients, using ultra-sensitive wrist and ankle sensors, detecting tiny fluctuations in emotional arousal.

‘Psychiatry is 30 to 40 years behind physical health, especially in terms of diagnosis,’ said Dr Szymon Fedor, who received a Marie Skłodowska-Curie grant to research the potential of depression monitoring at the Massachusetts Institute of Technology in the US.

‘In physical health we are using blood pressure to predict heart attacks, but in mental health we tend to base it on verbal reports from patients.

‘We want to be more objective and accurate with diagnosis and treatment of mental health - as we are with heart attacks or strokes.’

Dr Fedor is running a trial with five depressed patients who are receiving either electroconvulsive therapy or transcranial magnetic stimulation.

Electroconvulsive therapy induces seizures in patients to provide relief from symptoms of depression while transcranial magnetic stimulation uses magnetic fields to stimulate nerve cells in the brain.

Patients in the study receive the treatments each day for two months while agreeing to wear sensors continuously on their wrists.

The devices measure electrodermal activity - electrical characteristics of the skin that reflect levels of emotional arousal.

“

‘The sooner you realise that someone’s behaviour is changing, the sooner someone can make an intervention.’ 

The sensors, which look like smartwatches, also pick up heart rate and skin temperature, as well as recording movement - providing data on sympathetic nervous system activity, exercise and sleep.

At the end of each day, participants are asked to download the information from the sensors into a web application on a tablet or smartphone, which is then forwarded to researchers.

Algorithms

The physiological data is interpreted with the help of specially developed algorithms, calculating wellbeing, treatment efficacy and risk of relapse.

Depression is further evaluated by a clinician, twice weekly, and questions are sent to participants on their smartphones to see whether the physiological measures are corroborated by subjective experience.

‘By looking at different aspects of a person’s life we can get a more accurate picture of their mood,’ Dr Fedor explained.

‘We can begin to build the bridge between the diagnosis of mental health conditions and physical diseases.’

Although individuals in the current study have to download the data at the end of each day, Dr Fedor says live streaming of information will also be possible.

‘Live streaming could be especially helpful when dealing with individuals who are at risk of suicide,’ Dr Fedor said.

‘The sooner you realise that someone’s behaviour is changing, the sooner someone can make an intervention.

‘For example, when the composition of an individual’s anti-depressant drugs is changed, this can initially have adverse effects on behaviour.

‘They may become very agitated or anxious - the data will show early warning signs for this.’

Movement disorder

One condition which has been linked to suicide, and particularly the taking of anti-depressant medication, is the movement disorder akathisia.

The condition is linked to increased levels of the neurotransmitter norepinephrine, leading individuals to feel restless and agitated, sometimes rocking back and forth and stamping their legs.

Studies have shown that individuals on certain anti-depressant medications – known as SSRIs - are 10 times as likely to experience the distressing symptoms of akathisia than others.

‘The sensors will allow researchers to detect when individuals are experiencing akathisia and to act quickly, as this has been linked to suicidal ideation,’ Dr Fedor said.

‘The sensors will be able to tell when individuals are pacing about or sleeping, for example.’

Dr Fedor's project, known as PERSONA, runs until 2017 and follows on from the EU-funded PSYCHE project, which ended in 2013.

This project looked specifically at individuals with bipolar disorder, using physiological data obtained from electronic devices contained within clothing to detect changes in mood.