“Electronic nose” accurately sniffs out difficult-to-detect cancers

A new study by researchers at the University of Pennsylvania and the University of Pennsylvania Perelman School of Medicine found that odor-based tests that sniff out steam released from blood samples distinguish between benign and pancreatic and ovarian cancer cells with up to 95% accuracy. I was able to …

The findings show that a tool developed by Penn, which uses artificial intelligence and machine learning to decode mixtures of volatile organic compounds (VOCs) released from cells in plasma samples, allows for more difficult screening. It suggests that it may serve as a non-invasive approach. -Detects cancers such as the pancreas and ovaries.

The results will be presented by Dr. Rebecca W. Bushnell, Professor of Physics, University of Pennsylvania, Faculty of Arts and Sciences, at the American Society of Clinical Oncology’s Annual Meeting on June 4.Abstract # 5544).

“This is an early study, but the results are very promising,” Johnson said. “The data show that these tumors can be identified both in the advanced and early stages. This is exciting. Clinical background, This could be a test performed with a standard blood draw that may be part of your annual body. “

Co-authors include Dr. Erica L. Carpenter, director of the Institute for Circulating Tumor Substances and assistant professor of research at Perelman School of Medicine. Cynthia Otto, DVM, Ph.D. , Director of Working Dog Center, Professor of Pennsylvania Veterinary School (Pennsylvania). The late Dr. George Preti of the Monell Chemical Senses Center is also a co-author.

Penn’s research team is currently working with VOC Health to commercialize devices along with other devices for research and clinical applications.

The electronic sense of smell (“e-nose”) system is equipped with nanosensors that are tuned to detect the composition of VOCs emitted by all cells. Previous studies by researchers have shown that VOCs are released from tissues and plasma. Ovarian cancer patients Different from those released from the patient’s sample Benign tumor..

Of 93 patients, including 20 patients Ovarian cancer, 20 benign ovarian tumors, 20 cancer-free age-matched controls, 13 pancreatic cancers, 10 benign pancreatic diseases, 10 controls, steam sensor VOCs from ovarian cancer 95 Identified with% accuracy, Pancreatic cancer With 90% accuracy. The tool also correctly identified all patients with early-stage cancer (8 in total).

The pattern recognition approach of this technology is similar to how people’s own sense of smell works, telling what odor a clear mixture of compounds conveys to the brain. This tool was trained and tested to identify VOC patterns associated with cancer cells and VOC patterns associated with healthy cells. Blood sample Within 20 minutes.

A team collaboration with VOC Health CEO and Chief Innovation Officer Richard Postrel has increased detection speed by a factor of 20.

“By working with researchers from the Faculty of Physics and Astronomy, Perelman School of Medicine, and the University of Penvenia, we were able to complete and integrate our own innovations, facilitating the commercialization process,” Postrel said. I am. “Initial prototypes of commercial devices capable of detecting cancer from liquids and vapors will soon be available and will be provided to these pen researchers to facilitate their research.”

Johnson, along with collaborator Benjamin Avella, MD, grants MPhil, a professor of emergency medicine, a two-year $ 2 million grant from the National Institute of Public Health, National Advanced Translational Center, in a related effort with VOC Health. I was awarded money. Science for the development of handheld devices that can detect the characteristic “smell” of COVID-19 people based on COVID-19 cancer-Detection techniques applied in this study.