This large-scale study (n=2092) used Fourier Transform Infrared (FTIR) spectroscopy and machine learning to develop a blood test (Dxcovr® Cancer Liquid Biopsy) for detecting eight cancer types. The test achieved high area under the receiver operating characteristic curve (ROC) values for individual cancers (0.76-0.91) and demonstrated the ability to detect 64% of Stage I cancers with 99% specificity when classifying 'any cancer' against non-cancer controls. The test can be adjusted for higher sensitivity or specificity depending on clinical needs, offering a potential low-cost strategy for earlier cancer diagnosis.
Publisher
British Journal of Cancer
Published On
Nov 15, 2023
Authors
James M. Cameron, Alexandra Sala, Georgios Antoniou, Paul M. Brennan, Holly J. Butler, Justin J. A. Conn, Siobhan Connell, Tom Curran, Mark G. Hegarty, Rose G. McHardy, Daniel Orringer, David S. Palmer, Benjamin R. Smith, Matthew J. Baker
