Measuring the level and type of volatile organic compounds (VOC’S) emitted from the lungs with each breath is a technique which is being tested in many disease states. This technology, commonly known as eNose, produces a “breath print” which is then analyzed. The system functions much like our own olfactory system. Artificial intelligence(AI) views this “breath print” and compares it against previous data from patients with a particular disease. The AI learns the pattern of VOC’s from other patients and compares it to the one it is testing, and if the matching has a high correlation, a degree of confidence is generated.

Recently, this type of analysis was employed in separating the diagnosis of sarcoidosis (a disease of inflammation we have discussed before) to nterstitial lung disease (a scarring disease of the lung, ILD). Depending on the presentation, these two diseases can appear very similar on CT scanning of the chest. There is no simple non-invasive diagnostic test to separate these two disease processes. The important issue being that they would have different treatment options.

Researchers collect a lot of data from the VOC’S in exhaled breath, utilizing multiple sensors that detect different elements. Breath profiles are then produced and these profiles are compared with groups of patients with one or more of the diseases in question. In the case of sarcoid and ILD, recent studies have shown eNose technology to have excellent sensitivity, specificity in distinguishing sarcoid from ILD