When you’re sick, you act differently. You likely move around less, don’t talk on the phone as regularly, perhaps get more or less sleep than usual. And when you’re really sick, those changes can be even more pronounced.

That’s the insight behind Ginger.io, a mobile health analytics startup working out of the Cambridge Innovation Center, and spun out of the MIT Media Lab.

“The basic idea is there’s sort of an interesting time where everybody has one of these mobile phones,” said Anmol Madan, co-founder and CEO of Ginger.io. “And these things are incredibly powerful diaries of a person’s life.”

The modern smartphone is equipped with numerous sensors. It gathers data on how you communicate, where you travel, and more. Today’s devices have altimeters to measure whether you change elevation (like climbing stairs) and can integrate with a pedometer to track how many steps you take.

But Ginger.io isn’t focused on helping fitness nuts count their steps. As I wrote this morning, the bulk of healthcare costs are driven by patients with chronic diseases, and this is where Ginger.io hopes it can help.

One of the main conditions the company is targeting is depression. If you’re prone to severe depression and suddenly stop texting your friends and don’t move all day, that’s a sign that you may be in trouble. Ginger.io hopes to eventually be able to ping the professional managing your health and let them know.

And that’s just the beginning. Because mental health issues are tied to a variety of other chronic diseases like diabetes – the company won the Data Design Diabetes Challenge last year – this kind of monitoring has value there too. More broadly, people act differently when they’re sick; if Ginger.io can combine data with predictive models to tip off healthcare professionals when patients are symptomatic, that’s valuable.

The company raised $1.7 million in seed funding in October, and in addition to the DDDC, also won the SXSW Accelerator and was a finalist in the 2012 Alzheimer’s Challenge.

Ginger.io’s value proposition fits nicely with the trend in healthcare toward managing a population’s health outcomes overall, which means caring what they do when they’re at home rather than just treating them when they show up to the hospital.

As the company collects more data, it can be more powerful still. And while parts of the medical community may be hesitant about machine learning, co-founder Karan Singh told me he believes that is changing.

I asked Madan about the Boston ecosystem and its impact on his startup, and he told me that in few other places could he hire multiple PhD’s with expertise at the intersection of medicine and computer science. And he sees Boston as having tremendous potential in big data, in one area in particular.

“The opportunity that is wide open right now that Boston could be really good at is making sense of the data,” he told me. “The analysis.” He cited not just Harvard and MIT but also Northeastern as having prowess in that area.

Madan also told me that a handful of other CIC occupants are helping to test Ginger.io to see if the platform has any value for the average healthy user. But that’s not his focus. “We’re more of a chronic patient company,” he said.

Which is good. Because that’s where the real challenge lies.