Diagnosing by smell

Smelling the breath and body odor of a sick person has long been considered an important part of the herbalist's examination, regardless of your tradition and approach. Modern scientists have also been aware for a while that "bad breath" is not simply a condition of overgrowth of oral bacteria, but that it can also be linked to a variety of unbalanced internal processes (see this review for a summary). Now, a machine that analyzes volatile compounds in a patient's breath has been shown to detect the presence of lung cancer pretty reliably (3/4 of the time, with a misdiagnosis rate of about 25%). Neat idea -- but it's interesting to note that folks trained in Traditional Chinese Medicine started using the ultra-sensitive noses of cute poodles to do the exact same thing a few years ago.
Though our olfactory apparatus isn't as sharp as a dog's, we may still have some capacity to use information garnered from the breath and body odor of sick folks to understand what's going on. I recall a brief conversation I had at the last International Herb Symposium, while sipping Kava punch at the herbalist's Ball, where a gentleman whose name I can't remember was discussing olfactory diagnosis. An expert in the process had apparently teamed up with a perfumist and created a whole kit of scents, each of which exactly mimicked the "scent" of an individual disease. If anyone knows more about this, please let me know! I'd love to look, or smell, a little deeper...