BreathObserver – Non-invasive point-of-care respiratory gas analysis for lung cancer detection

 


 

Overview

Project description:

Geratherm Respiratory

Lung and bronchus cancers are among the most common cancer types and are responsible for 1/5 of the tumor-associated diseases in Germany. The established methods for initial diagnosis of lung cancer include Computer Tomopgraphy (CT) and bronchoscopy, which are stressful for patients, time-consuming and expensive. Moreover, these procedures are invasive and limited due to the extensive resources required.
The objective of BreathObserver is the research on a novel modular and multimodal medical diagnostic device, which directly analyses the human exhalation (breath). We aim to detect indications of lung cancer early as well as distinguish between different severities, with the detection and analysis of volatile organic compounds (VOCs). The completely new, cost-effective, non-invasive, mobile and modular exhalation analysis system, which is to be explored, is supposed to be used for screening and the deployment of a rapid initial diagnosis in hospitals and medical practices.

The consortium of the project consists of four partners with complimentary expertise:

  • Geratherm Respiratory GmbH - Industrial partner, project coordinator: software, overall integration of the software and hardware components to a functional demonstrator
  • UST Umweltsensortechnik GmbH - Industrial partner: modular, multimodal gas-sensor technology with exchangeable disposable sensor
  • Technische Universität Ilmenau - University: methods for sensor calibration and data analysis for the detection of biomarkers in the exhalation gas
  • University hospital Jena - University / clinical partner: clinical requirements analysis, biomarker detection and reference studies

To achieve the aims of this project, a novel overall system will be investigated, which will be based on a specific metal-oxide (MOX) semiconductor gas-sensor technology. This technology will be integrated in a modular and mobile measurement system. With the help of multimodal sensors and the combination of feature engineering data analysis as well as machine learning algorithms we aim to achieve a detection of pathology-specific biomarkers in the exhalation gas with high sensitivity and specificity (≥ 85 %).
This novel exhalation analysis is a significant contribution to the early point-of-care diagnostic of lung diseases, especially lung cancer, and their therapeutic follow-ups. Because of the early and simplified diagnostic opportunities, the system supports the chances of treatment and also reduces cost and resources for the health care system.