Nanoarray to sniff out diseases from your breath?

| December 29, 2016 | 3 Replies

Do you remember the NOKIA Concept, Nokia Morph? As well as being able to change into different devices, it had an array of nano sensors, part of which could detect chemicals in the environment. Yesterday, SlashGear posted about a new medical ‘breathalyzser’ that could detect diseases. Usually, any time similar ‘ground breaking’ headlines appeared on Reddit, the first comment would usually dispel such claims.

But this isn’t that far fetched. You may be familiar with alcohol breathalysers? That’s already become a consumer item, even with smartphone accessory options.

In the medical field, breath samples are used already to help guide a diagnosis. e.g. Hydrogen breath test that might suggest intestinal bacterial overgrowth, lactose intolerance; urea breath test that might indicate presence of H. pylori, a bacteria found in the stomach that often can cause ulcers. It’s a nice non-invasive test (though can be time consuming to collect).

‘Disease Breathalyser’

According to a piece published on American Chemical Society they’ve created a piece of equipment that can essentially sniff and distinguish multiple diseases.

The claim is based on different diseases producing a different chemical profile through the breath which can then be identified and distinguished. Artificial intelligence techniques was said to have been used to classify and diagnose the conditions. The breath components were identified by using mass spectrometry. If you haven’t heard of that before, that’s basically a ‘super nose’ as such. It sort of breaks down/ionises a chemical and sorts them out, producing a signature of each component involved.

As you can imagine, there’s quite a bit of research already  into the metabolic analysis of breath. It’s called ‘breathomics’. There’s also ‘salivaomics’ for, as you guessed it, saliva. The aim, again, is to provide quick, non-invasive diagnostic tools.

I think the last thing we want is to arm Dr Google so that the public become more paranoid when they step foot onto hospitals or family doctor’s reception. But I think it’s good to at least be more aware that these kind of things could become available. They might help actual medical practice (after they’ve been rigorously tested) as an ‘adjunct’, merely assisting and not being the ‘oh it’s this’ final diagnosis machine as it would be useful to have more, non-invasive, quick tests that back up initial clinical suspicions from symptoms/history and examination alone.

Nokia and Health

Other than Nokia’s Morph concept and their Withings health brand, Nokia has of course pioneered Nokia sensing x challenge. Remember another personalised diagnostic tool (also nano based).

Meet the first Nokia Sensing X Challenge Winners! Nanobiosym

This one from 2014:

$525,000 Nokia Sensing XChallenge Winners announced – rHealth – portable blood lab!

Further applications in Medicine

There are great applications to decentralise medicine, especially in rural areas where it might be difficult to get to a doctor. I think it’s great that somehow they’re making it cheaper too. The use of artificial intelligence to decide on the significance of the set of data is also quite intriguing. A lot of medicine is definitely about pattern recognition. But it’s not just that. It’s often said that Medicine is as much an art as it is a science. It’s not always the case of adding up a few things to come to a conclusion, and then there’s the all important human/patient component. You may have watched some episodes of House to know where that leads.

I think it is useful in guiding people to be more aware of their health. Campaigns on recognising symptoms of cancers have been fantastic because most have better outcomes when detected earlier. Same goes for other diseases really. If you’re aware sooner, you can do something about it before it really takes a hold. The original article mentions diabetes being detected for patients having ‘sweet breath’ . There are already quite a few diagnostic techniques used to point towards possible diabetes. Btw, diabetes was also known for having ‘sweet smelling urine’, hence not only can blood be quickly tested, but presence of sugar can also be quickly detected in urine. I’m surprised every day seeing how bad the complications of diabetes can get (e.g. blindness, amputation, kidney disease, sepsis, cardiovascular disease, death) before a patient might seek help, but early detection could lead to more prompt management, better counselling, more frequent routine follow ups etc all of which have been shown to help lower those complications.

I think what scares me as a doctor is not so much the huge volumes of people coming in (you’ll always have a mega busy work load) because they think they have something that they don’t. It’s when they might have something but are swayed against seeking medical help and are missed. But eventually, if they get smart enough, it may also mean than the constantly shrinking available resource could be used a bit better, i.e. unwell people needing help are treated according to that level of need. The UK/NHS is sort of aiming for a ‘diagnosis’ by smartphone. We already have a semi-automated ‘triaging’ system called ‘111’ where people phone in or use the 111 website, plug in/talk about their symptoms and based on (I’m guessing tick boxes or a flow diagram) they’re directed about whether they need to seek a certain level of medical help.

It’s exciting to hear about this sort of progress that could likely be in our phones some day. Beyond simply just having better processors, cameras, screens and whatnot, it’s a great step to focus on having things that could really contribute to a healthier you. We have a lot of those sort of things already today. Imagine what the future holds.





Category: Nokia

About the Author ()

Hey, thanks for reading my post. My name is Jay and I'm a medical student at the University of Manchester. When I can, I blog here at and tweet now and again @jaymontano. We also have a twitter and facebook accounts @mynokiablog and Check out the tips, guides and rules for commenting >>click<< Contact us at tips(@) or email me directly on jay[at]