negative thoughts prompt a whole host of physiological changes, some of which can be detected fairly easily:

skin conductance -- e.g. http://onlinelibrary.wiley.com/doi/10.1111/j.1469-8986.1991.tb03392.x/abstract

heart rate chance -- e.g. http://www.sciencedirect.com/science/article/pii/S0167876006002194

(not so easy to detect) eeg -- e.g. http://www.sciencedirect.com/science/article/pii/S0191886905000413

On their own, this data is not so useful.I suspect though, that if they were used together, they would a lot better at detecting short term and long term patterns leading to a 'downward swing'.

Mashing in data other sources could improve this too:

voice pattern analysis. -- e.g. http://knoceq.blogspot.co.uk/2013/12/detection-of-disease-through-sound.html

fine movement changes. -- e.g. http://www.ncbi.nlm.nih.gov/pubmed/17515173

How to combine this information? One option is to use artificial neural networks like used by Smartphones for understanding speech http://www.wired.com/2014/06/siri_ai/ .

THE CHALLENGES

--enthusiasm for this project! Join us and spread the word! Citizen Science!

--are there other bio-signals that can be used? E.g. could we develop an unobtrusive headset to be worn to detect errant brain patterns via EEG? e.g. http://www.newscientist.com/article/mg21829146.000-etattoo-monitors-brainwaves-and-baby-bump.html#.VQfp_I6sWw4.

--we will need thousands of people to take part in studies. I expect an app on your phone would monitor your vitals and periodically ask you about your mood. e.g. http://www.ted.com/talks/matt_killingsworth_want_to_be_happier_stay_in_the_moment?language=en

WHO CAN HELP?

--everyone, by taking part in our studies.

--software engineers, hackers, statisticians, psychologists, citizen scientists.

CAN I HELP NOW?

Yes! Start chatting about this. Sign up to the project.