Sniffing Out Disease With “Scentprints”

Long before symptoms of a disease emerge, distinct metabolic odor signatures can be produced in the breath. Detection of these signatures can offer a novel approach to confirm the presence of a disease early, ensuring timely treatment and, in some cases, avoiding invasive biopsies.

Technology Networks spoke to Karim Aly, CEO of Noze to find out more about replicating human odor perception and the benefits of breath-based disease detection.

Kate Robinson (KR): What is a digital nose and how is it relevant to diagnostics and screening?

Karim Aly (KA): A digital nose is a technological system that essentially mimics the sense of smell. It detects and analyzes chemical compounds in the air, converts these into digital signals and identifies the underlying odors. When we apply this to identify specific compounds released by the human body, the potential for using this technology in disease detection becomes quite significant.

Long before symptoms of disease and illness are evident, they alter the body’s metabolism to produce a distinct metabolic odor signature – a biomarker –present in our breath. So, every time we exhale, we're essentially releasing a unique snapshot of our health right into the air. The process of capturing breath samples, deciphering these "snapshots" and reliably identifying any present biomarkers is the essence of breath-based disease detection. 

The ability to decode human breath holds the potential to revolutionize the diagnostic landscape, opening doors to early discovery and treatment for a wide range of health issues, from cancers and metabolic disorders to infectious diseases. By leveraging the latest advancements in digital odor perception alongside the power of AI, we are on the cusp of a transformative era in diagnostics with breath-based disease detection. This is not just progress; it's a revolution in healthcare, promising better care and superior outcomes for patients.

KR: How are the key functions of human odor perception mimicked by the Noze chip?

KA: At Noze, we realized early on that digital olfaction needs to be viewed as a system, akin to the olfactory mechanism found in mammals. Odor perception in mammals, including humans, involves four primary functions, and Noze's technology accurately replicates each of these. 

Function 1: Capture

In mammals, the front end of the olfactory system is a collection of olfactory receptors. In order to emulate this, we built a sensor chip with a diverse array of chemical receptors. Just as an odor triggers a distinctive neural code when it interacts with mammalian olfactory receptors, a unique "digital scentprint" is generated when an odor moves across our chemical receptor array.

Function 2: Identify

The olfactory bulb and the brain process the incoming neural code and interpret whether it’s the scent of coffee or, say, fresh bread with coffee in the background. Noze’s proprietary multi-layered perceptive AI engine emulates this capability by applying a series of algorithms to accurately interpret “scentprints,” no matter what else might be in the background.

Function 3: Learn

Humans have a remarkable ability to learn new smells – consider how effortlessly a child learns the scent of coffee or burning wood after just one encounter. Similarly, Noze algorithms have been designed to learn from a very limited set of digital "scentprints," compensating for the absence of extensive, well-curated odor datasets.

Function 4: Evolve

Humans possess the power to expand our olfactory repertoires by sharing knowledge and constantly incorporating new aromas into our personal smell “libraries”. In a similar vein, each Noze chip is linked to our centralized, cloud-based platform, enabling  the chip to recognize new odors as more “scentprints” are incorporated into the odor library.

KR: What inspired the development of digital olfaction?

KA: The journey of Noze began with an ambition to digitize the sense of smell by leveraging recent technological advances in material science and artificial intelligence. Dr. Ashok Masilamani, the founder and chief technology officer of Noze, became increasingly intrigued as to why we have managed to digitize sight through cameras, touch through haptics and sound through microphones, yet the sense of smell had been left behind.

Inspired by NASA's work utilizing an electronic nose on the International Space Station to monitor astronaut health, Noze reached out to NASA and would eventually become the first non-US company to be granted an exclusive license to its technology. This milestone provided Dr. Masilamani with the momentum to start building towards his own vision for a digital odor perception platform. Over the following eight years, Noze dedicated itself to advancing this technology through miniaturization, the addition of proprietary layers of aroma data engineering and building sophisticated AI algorithms. The result is a highly advanced digital odor perception platform.  

KR: Can you explain how screening is performed using the chip?

KA: Most significant illnesses produce a metabolic odor signature, which is emitted through our breath. Once an AI model for a specific illness or condition has been developed and activated, the process is actually very simple. The patient exhales into the breathalyzer device to introduce a breath sample. Inside the breathalyzer, the sensor chip processes the sample, transforming it into a digital signal, or "scentprint". This “scentprint” is then processed by our AI algorithms, which analyze it and instantly return a result.

KR: Does this method of diagnostic have any benefits over traditional methods?

KA: Within the field of disease detection, there have been remarkable strides, yet challenges still remain. Misdiagnosis, particularly through false positives, is a common issue in screening tests. Additionally, detecting serious diseases often involves costly and invasive procedures, putting them out of reach for many patients. Furthermore, there's an absence of standardized screening methods for certain conditions, which carry significant health risks.

Breath-based disease detection is a game-changer, providing real-time results that can play a crucial role in early detection, ensuring timely treatment and enhancing patient health outcomes. These screenings are easy, affordable and non-invasive, making them a preferred choice. What's more, being administered through a compact, handheld device means they can be used at any point-of-care, from super hospitals to remote clinics. This can also significantly enhance global healthcare equity, enabling access to early disease screenings in remote and developing regions previously underserved by such medical options.

KR: How easily could technology like this be implemented into clinical practice?

KA: Our technology can seamlessly integrate into current care pathways, acting as a precursor test before more invasive confirmatory tests, or, in certain situations, it may even completely eliminate the need for them. Noze has successfully addressed the technological and clinical hurdles faced by previous endeavors in breath-based disease detection, creating a diagnostic breathalyzer that's simple to operate, cost-effective and adaptable to any care setting, irrespective of geographical location or economic status. Noze’s breathalyzer has already been deployed in several clinical studies, with the results proving exceptionally favorable.

Karim Aly was speaking to Kate Robinson, Assistant Editor for Technology Networks.

About the interviewee:

Karim Aly is CEO of Noze. He is focused on transforming healthcare by empowering machines with the ability to smell. Prior to Noze, Karim established one of the first startup studios in Canada in affiliation with one of the country’s largest universities.