Our Technology

With our design we are capable of converting the airwaves generated from your breathing into sound waves, thereby allowing us to analyze your breath with our acoustic analytics and artificial intelligence, providing an innovative and indispensable tool to those in the fields of medical diagnostics, wellness, and fitness. Our solution can be either standalone or integrated into other respiratory devices and equipment, allowing considerable versatility with regard to how healthcare professionals and providers leverage the data, providing new data insights to detect and manage respiratory health compared to other existing devices.

The latest evolution of our spirometer also simultaneously measures lung flow volumes in conjunction with sounds generated by the lungs

To date, we have authored a total seven (7) patents.


The Analyzer for Pathology consists of  a number of modes that analyze different kinds of breathing protocols including tidal breathing (at rest), slow vital capacity and forced vital capacity.

Our machine learning algorithms report and track statistical information that combines data about respiratory flow and harmonic frequencies to understand the behavior of healthy breathing and the signs of pathology in different stages of respiratory disease. We have developed over 35 descriptors for wheeze, crackles, coughs and breath pattern behavior that can be applied to respiratory disease.


We have developed an artificial neural network that ingests lung flow volume and lung sound descriptors simultaneously to determine pathology severity such as: asthma, COPD, pulmonary fibrosis, lung cancer, respiratory infections. The neural network interprets data and subsequently recognizes patterns in order to help cluster and classify samples subsequently grouping them into the pathologies, while simultaneously learning and evolving to more accurately recognize pathology and disease severity. To date, we have successfully trained our artificial neural network to evaluate asthma and COPD severity.


Breath Research has been developing a cardiopulmonary exercise test (CPET) based on pure audio input only. Our goal is the accurate detection of ventilatory thresholds (VE/VO2 and VE/VCO2 inflection points) which are the basis for exercise intensity or heart rate training zones and can be used for early detection and tracking of cardiopulmonary diseases.

Our custom framework successfully detects ventilatory thresholds for a broad range of adult populations during exercise tests of varying length with a 93% accuracy rate.


We have benefited from cultivating recording and analyzing respiratory sounds from a number of different sources, including the onboard microphone in mobile phones. As a result, we have developed a breathing app over the years that integrates breath measurements and coaching. Our current app is available on the iTunes store.