Detection of Lung Cancer using Expired Breath Analysis by Ion Mobility Spectrometry

Edgar Williams

Research output: Contribution to journalArticlepeer-review


Introduction: Volatile organic compounds (VOC's) present in exhaled breath reflect the metabolic activity of the body and change in the presence of both respiratory disease and lung cancer. Human breath analysis is a non-invasive technique for the rapid identification of gas-phase analytes, such as VOC's, and offers a new clinical diagnostic tool. Ion Mobility Spectrometry coupled with a Multi Capillary Column (MCC/IMS) allows exhaled breath to be rapidly analysed and characterised with great precision.

Aim: To derive a database of exhaled breath profiles from participants with and without respiratory disease and to use this database to determine if breath borne VOC's can be used to differentiate disease states.

Methodology: Breath samples from 323 respiratory patients with and without lung cancer were measured using the MCC/IMS and analysed in real time (mean age 68 years, 195 males and 128 females). Demographic data, medical diagnosis and medical history were also collected.

Results: The respiratory data were compared to control data from 182 healthy participants of the general public and NHS staff. All breath profiles collected were analysed using advanced chemometric techniques developed in collaboration with Radboud University, Nijmegen, Netherlands. This analysis determined which exhaled breath VOC's have the potential to differentiate disease.

Conclusion: The study has shown that lung cancer can be identified with 87% specificity and 70% sensitivity from healthy controls, reiterating the potential of breath analysis as a diagnosis tool.
Original languageEnglish
JournalEuropean Respiratory Journal
Issue numbersuppl 59
Publication statusPublished - 2015


  • Biomarkers
  • Breath Test
  • Lung cancer
  • Oncology


Dive into the research topics of 'Detection of Lung Cancer using Expired Breath Analysis by Ion Mobility Spectrometry'. Together they form a unique fingerprint.

Cite this