A preliminary study on the possibility to diagnose urinary tract cancers by an electronic nose

In this paper a study about the possibility of an early and non-invasive diagnosis of urinary tract cancers with an electronic nose is presented. Measurements of urine headspace were performed by means of an electronic nose based on eight quartz crystal microbalance (QCM) gas sensors coated with different metalloporphyrins. A total of 113 patients affected by various urological pathologies were involved in the experiment, 19 of them were measured twice, before and after the surgical treatment of the cancer. 18 healthy volunteers were used as a reference population. The e-nose data were processed by both principal component analysis (PCA) and discriminant analysis solved by partial least square (PLS-DA). The PLS-DA model related to the data of the healthy and the ills individuals, except the post-surgery samples, shows a complete discrimination between these two groups and a gradual differentiation between prostate and bladder cancer. The application of PCA to the whole data set, shows a migration of the post-surgery class towards the healthy group, suggesting that the headspace of urine samples of patients with the urinary tract cancer is different from that of healthy individuals. These preliminary results confirm a strict correlation between urine headspace and urological pathologies, enhancing the possibility to perform early diagnosis, considered of upmost importance along the medical diagnostic path. (c) 2007 Elsevier B.V. All rights reserved