Learning by Research

Thesis Title: Development of electrochemical sensors for bloodstream infection diagnostic

Blood culture is the gold standard for detecting bloodstream infections, which are widespread and potentially fatal. The time between the appearance of symptoms and the administration of appropriate treatment is therefore crucial. To perform a blood culture, cumbersome and expensive equipment is required but only offers a binary answer (infected or sterile) on the condition of the collected blood. The transport time between the collection site and the laboratory is time lost for diagnosis, resulting in a delay in therapeutic choices. This thesis aimed at evaluating the interest of electrochemical monitoring of a blood culture. First, a prototype of blood culture bottle instrumented by an electrochemical sensor array was developed. Many bacterial and fungal species were tested. The potentiometric measurements provided a multiparametric response, ensuring the detection without having to use a centralized and complex automaton.

Moreover, these signals could be exploited by machine learning in order to identify the detected pathogen, without manipulation or chemical reagent. This technology would thus allow to valorize the transport time and to automate the first steps of identification to generate a relevant diagnosis in the shortest time. The last works carried out during this thesis allowed to study the biological and chemical phenomena involved during the detection and identification. These results open the way to numerous applications and research themes, particularly in the field of bacterial biofilms.

Key Words: Electrochemical sensors, biological diagnostic, blood culture, in vitro diagnostic, machine learning, label-free identification

Doctoral School:  ED I-MEP² - Engineering - Materials, Mechanical, Environment, Energetics, Processes, Production
Research laboratory: Laboratoire d'électronique et de technologie de l'information (CEA- Leti)
Thesis supervision: Pascal MAILLEY, Pierre MARCOUX and Maxime GOUGIS

> To find out about all the 2024 thesis prizes