In-field detection of hands contamination by AI analysis of living bacteria – LAIveBactDetect

The COVID-19 pandemic puts the emphasis on the essential role that hand hygiene plays in the prevention of infections. For instance, in hospitals, nosocomial infections due to hand hygiene issue are considered as the cause of more than 5000 deaths per year in France and Germany. The deployment of a portable device that identifies living bacteria on hand and alerts in case of potential contamination would be an efficient way to decrease this number The LAIveBactDetect project aims at providing such a device. The device operates in four steps. First, a samples containing the pathogens to analyze is collected during the hand-washing process at a hand-sanitising dispenser. Second, an integrated optical microscope takes pictures and movies of that sample. Third, the raw data are processed by sophisticated AI-based algorithms embedded in a processing unit inside the device. Finally, a decision about a potential contamination is made and given to the user. The project will be carried on by a French-German consortium composed of two academic laboratories and two companies. The University of Freiburg will build a reliable and annotated database to train the AI. The University of Strasbourg will develop the AI algorithm. ADDI-DATA FRANCE, will design the embedded AI processing unit in which they will implement the algorithms. Finally, OPHARDT Hygiene, will integrate all parts of the system in a prototype that will be used at the end of the project for in-field testing. Altogether, we master the entire chain of skills required for the development of the embedded AI processing: microbiology, theoretical aspects if AI, electronics, embedded AI, system integration. If successful, LAIveBactDetect is expected to save 500 lives and 2.5 M€ of public health expenditure per year. The return on investment for involved companies is estimated to 5 years.