Sniffing out cancer with improved 'electronic nose' sensors

Diagnosing cancer today usually involves various imaging techniques, examining tissue samples under a microscope, or testing cells for proteins or genetic material. In search of safer and less invasive ways to tell if someone has cancer, scientists have recently started analyzing breath and defining specific profiles of compounds in breath samples. But translating these exhaled disease fingerprints into a meaningful diagnosis has required a large number of sensors, which makes them impractical for clinical use. Hossam Haick and colleagues sought to address this problem.

The researchers developed a small, breath-diagnostic array based on flexible gold-nanoparticle sensors for use in an "electronic nose." The system -- tested on breath samples from 43 volunteers, 17 of whom had ovarian cancer -- showed an accuracy rate of 82 percent. The researchers say developing this method further would require larger-scale clinical testing. They add that the approach could also apply to diagnostics for other diseases.

The authors acknowledge funding from the Horizon 2020 ICT Program.