Which sensor to use to identify if a person stopped breathing?

Hi guys wonder if anyone can help me. I am working on a project where i am improving health care of elderly people. One of the tasks that i am doing is to monitor the user's breathing especially if they are asleep.

I have done some research but nothing good came up. Which sensor to use to identify if the user stopped breathing?

Thank you

Look at the polysomnography technology...

Breathing is a movement of the chest that produces an air flow through the nose or mouth...

You can use a resistive stretch belt around the chest to get the movement or a fast thermistor that detect changes in the temperature of exhaled or inhaled gaz (but you need one at the nose and one at the mouth wich is not very comfortable)

http://www.sleepreviewmag.com/2013/05/worth-the-effort

What about those little LED heartrate sensors, they are measuring the different color of blood which should be proportional to the oxygen in blood?

Ah! The old Heisenberg quandry raises its head again!

** IF ** your situation is such that the subjects would be willing to subject themselves to the nuisance of the sort of facemask that sleep apnea patients have to endure, it becomes so much easier! Sundry sensors can be introduced into either the ingress or egress channels. A simple "flow meter" on the ingress would be my first thought.

By the way...
If that subject line was deliberate: Congratulations! You DID get my attention. But maybe too many possible respondents just passed over the post, not realizing that it might be serious? And it it wasn't... shame on you for sucking me in!! (^_^)

To my fellow forum writers: How good and restrained you have been! (Note that I have saved my "witty" remarks for a footnote to a serious reply. One of the reasons I enjoy the Arduino forum is the high standards here. I often have to give up on forums when dragged into checking too many replies which turn out to be nothing more than "Ha, ha!", etc.)

I would give the idea of sensitive microphones, listening to breathing noise. they listen while a patient is awake, and take record, through coding should a change arise it will in turn trip alarms. this would be good so long as the unit understands a delay between breath when awake or sleeping. this makes the system easily repaired and easy to replicate because the component is simply a microphone. it is possible to conjoin an O2 meter along with the mic. or to make their data coincide to determine sleep, awake, emergency levels.