Artificially Intelligent Nanoarray Detects Various Cancers by Liquid Biopsy of Volatile Markers

Reef Einoch Amor, Assaf Zinger, Yoav Y. Broza, Avi Schroeder, Hossam Haick

Research output: Contribution to journalArticlepeer-review

Abstract

Cancer is usually not symptomatic in its early stages. However, early detection can vastly improve prognosis. Liquid biopsy holds great promise for early detection, although it still suffers from many disadvantages, mainly searching for specific cancer biomarkers. Here, a new approach for liquid biopsies is proposed, based on volatile organic compound (VOC) patterns in the blood headspace. An artificial intelligence nanoarray based on a varied set of chemi-sensitive nano-based structured films is developed and used to detect and stage cancer. As a proof-of-concept, three cancer models are tested showing high incidence and mortality rates in the population: breast cancer, ovarian cancer, and pancreatic cancer. The nanoarray has >84% accuracy, >81% sensitivity, and >80% specificity for early detection and >97% accuracy, 100% sensitivity, and >88% specificity for metastasis detection. Complementary mass spectrometry analysis validates these results. The ability to analyze such a complex biological fluid as blood, while considering data of many VOCs at a time using the artificially intelligent nanoarray, increases the sensitivity of predictive models and leads to a potential efficient early diagnosis and disease-monitoring tool for cancer.

Original languageEnglish (US)
Article number2200356
JournalAdvanced Healthcare Materials
Volume11
Issue number17
DOIs
StatePublished - Sep 2022

Keywords

  • breast cancer
  • liquid biopsies
  • machine- learning
  • nanotechnology
  • ovarian cancer
  • pancreatic cancer
  • sensors

ASJC Scopus subject areas

  • Biomaterials
  • Biomedical Engineering
  • Pharmaceutical Science

Fingerprint

Dive into the research topics of 'Artificially Intelligent Nanoarray Detects Various Cancers by Liquid Biopsy of Volatile Markers'. Together they form a unique fingerprint.

Cite this