Breath Taking AI

Eupnoos is an innovative audio-phenotyping company dedicated to capturing and examining breath sound patterns via the mobile phone. By applying proprietary algorithms to this audio data, the platform can identify unique biomarkers and signatures that indicate the presence of respiratory diseases, such as asthma, enabling early diagnosis and personalized treatment plans.
Smartphone enabled asthma diagnosis in minutes, not months
Asthma Misdiagnosis: A Critical Problem
Asthma is a prevalent condition, with misdiagnosis rates reaching up to 1 in 3 people. Early and accurate diagnosis, as endorsed by international guidelines, is crucial as it increases the likelihood of receiving preventative therapy, reducing harm from hospital admissions and potentially preventing asthma-related fatalities. However, diagnostic challenges remain, as tests such as spirometry and fractional exhaled nitric oxide (FeNO) testing often provide suboptimal sensitivity.
There is an increasing demand for universally accessible, affordable diagnostics with exceptional sensitivity and specificity. These diagnostic tools must address health inequities and so need to cater to individuals of all races, genders, ethnicities, and locations, safely, and reliably.
The Eupnoos difference
By blowing into a smartphone (no additional accessories, hardware or mouthpieces needed), Eupnoos's proprietary algorithms identify signatures associated with asthma to rule in or rule out a diagnosis. The result is presented as a simple yes or no answer to the patient while the clinician interface includes a probability score. By ruling in or ruling out the disease, clinicians have the decision support they need instantly to decide if the patient needs confirmatory tests thereby cutting down the time to diagnosis and reducing incidence of under and mis-diagnosis..
Our proprietary algorithms and cloud-based software architecture provide clinical practitioners with essential information for making informed decisions across various healthcare interfaces.
In a pilot study (n=40), our spectral algorithm identified self-reported asthma with 100% sensitivity, 87.5% specificity.
Screening for Chronic Obstructive Pulmonary Disease without spirometry
Misdiagnosis and under-diagnosis of COPD
Some estimates suggest that COPD is under-diagnosed in up to 50% of cases or even more in some regions. Misdiagnosis rates are harder to quantify but remain a concern due to the lack of specificity of spirometry in differentiating COPD from other respiratory conditions, such as asthma.
Factors contributing to misdiagnosis and under-diagnosis using spirometry include improper test administration, incorrect interpretation of results, and patients not undergoing spirometry despite presenting with respiratory symptoms. Additionally, the effort-dependent nature of the test may lead to inaccurate results for some patients, particularly the elderly or those with severe respiratory symptoms.
These diagnostic challenges highlight the need for more accurate and accessible diagnostic tools to improve the detection and management of COPD
Eupnoos as an adjunct to spirometry has the potential to vastly improve specificity
In a pilot study (n=41), Eupnoos demonstrated promising results in detecting self-reported COPD with 75% sensitivity and 98.72% specificity thus addressing a recognised shortcoming with spirometry.