I’ve written a few times about the increasingly predictive power of our mobile phones.  A recent study even claimed that your phone could tell when you’re bored by monitoring your usage.

A recent study suggests that similar methods can also detect the onset of bipolar disorder.

Bipolar disorder is characterized by severe mood swings that range from euphoria at one end, to depression at the other.  With the mood swings often very unpredictable, it’s often a significant challenge to detect them fast enough to begin treatment.

Spotting mood swings

For instance, there aren’t any reliable biomarkers for bipolar, so current testing usually revolves around psychological testing to determine ones state of mind.  The self reporting nature of this process often means diagnosis lags somewhat behind the actual mood change.

Deploying mobile phones to assist in this endeavor was the brainchild of Venet Osmani at the Center for Research and Telecommunication Experimentation for Networked Communities (CREATE-NET) in Trento, Italy.

He believes that the sensors built into the average smartphone are ideal for picking up on the mood changes users undergo, as they occur.

For instance, the manic stage of the condition is typified by hyperactivity, which can manifest itself in rapid speech, high levels of movement and excessive phone usage.

People in depressive states, by contrast, tend to show similar behaviors, albeit at the other end of the spectrum.

Proving the pudding

The process was tested out on 12 people diagnosed with bipolar disorder.  Their activity was tracked over a 12 week period, during which the patient was asked to visit a clinic where their mental state was tested using traditional methods.

Interestingly, the activity and location data given off by our phone usage was a good indicator of our present mood, and was capable of detecting a change in mood 94% of the time.

When this was then combined with phone call data, the predictive success rose to 97%.

“Almost all changes were detected with almost no false alarms,” the researchers say.

If that kind of track record can be sustained in the long-term, it could have huge implications for people with the condition.

“One of the important aspects of this work is the possibility of the early detection of changes in a patient’s state with high precision and recall, facilitating timely intervention and thus leading to better treatment outcomes,” the team say.

More needs to be done before such confidence can be instilled in the findings however.  The study had a very small sample over a relatively short timeframe, so a more thorough analysis will be required before concrete recommendations can emerge regarding the approach.

Other studies have suggested there is validity in using phone data to understand and predict our moods however, so it will be interesting to follow the project as they test the hypothesis more rigorously.