Electronic nose for lung cancer detection

Abstract

The breath of patients with lung cancer contain particles that could be used to detect the disease. In this project, Valerio Bruschi and I, applied machine learning techniques to breath samples to train a model able to discriminate healthy and ill patience. We used two real datasets, obtained as measures in a medical environment from two electronic noses named Cyranose and ROTV. We successfully accomplished the following goals: 1) Instrument calibration using a set of key compounds. 2) High predictive quality on the collected data from the medical setting.