| Title: | Breast cancer detection by analyzing the volatile organic compound (VOC) signature in human urine |
| Author(s): | Giro Benet J; Seo M; Khine M; Guma Padro J; Pardo Martnez A; Kurdahi F; |
| Address: | "Center for Embedded Cyber-Physical Systems (CEPS), University of California Irvine (UCI), Irvine, 92697, USA. jgirbene@uci.edu. Center for Embedded Cyber-Physical Systems (CEPS), University of California Irvine (UCI), Irvine, 92697, USA. Department of Biomedical Engineering, University of California Irvine (UCI), Irvine, 92697, USA. South Catalonia Oncology Institute (IOCS), Sant Joan de Reus University Hospital, IISPV, Rovira i Virgili University, 43204, Reus, Spain. Department of Electronic and Biomedical Engineering, Universitat de Barcelona (UB), 08028, Barcelona, Spain" |
| DOI: | 10.1038/s41598-022-17795-8 |
| ISSN/ISBN: | 2045-2322 (Electronic) 2045-2322 (Linking) |
| Abstract: | "A rising number of authors are drawing evidence on the diagnostic capacity of specific volatile organic compounds (VOCs) resulting from some body fluids. While cancer incidence in society is on the rise, it becomes clear that the analysis of these VOCs can yield new strategies to mitigate advanced cancer incidence rates. This paper presents the methodology implemented to test whether a device consisting of an electronic nose inspired by a dog's olfactory system and olfactory neurons is significantly informative to detect breast cancer (BC). To test this device, 90 human urine samples were collected from control subjects and BC patients at a hospital. To test this system, an artificial intelligence-based classification algorithm was developed. The algorithm was firstly trained and tested with data resulting from gas chromatography-mass spectrometry (GC-MS) urine readings, leading to a classification rate of 92.31%, sensitivity of 100.00%, and specificity of 85.71% (N = 90). Secondly, the same algorithm was trained and tested with data obtained with our eNose prototype hardware, and class prediction was achieved with a classification rate of 75%, sensitivity of 100%, and specificity of 50%" |
| Keywords: | Animals Artificial Intelligence *Breast Neoplasms/diagnosis Dogs Electronic Nose Female Gas Chromatography-Mass Spectrometry/methods Humans *Volatile Organic Compounds/analysis; |
| Notes: | "MedlineGiro Benet, Judit Seo, Minjun Khine, Michelle Guma Padro, Josep Pardo Martnez, Antonio Kurdahi, Fadi eng England 2022/09/02 Sci Rep. 2022 Sep 1; 12(1):14873. doi: 10.1038/s41598-022-17795-8" |
