Multi-resolution classification of exhaled aerosol images to detect obstructive lung diseases in small airways

Symptomless development of the lung cancer makes it highly elusive to be detected at its early stages. This work presents a new diagnostic approach that uses exhaled aerosols to detect obstructive respiratory diseases. The hypothesis was that each lung structure had a signature distribution of exhaled aerosols (termed as aerosol fingerprint, or AFP). The proposed method was demonstrated to be able to detect small structural changes (<1 mm) in small airways (lung bifurcations G7-9) with a sufficiently high accuracy (92.5%). Compared to standard techniques such as CT, biopsy, and bronchoscope, this new approach has the advantages of being non-invasive, easy to use, and low cost. The proposed aerosol-based breath test shows promise towards the development of an alternative or supplemental tool to the low-dose CT scanning for lung cancer screening.266