You are here : Home > Elastic Light Scattering for low cost and non-destructive clinical pathogens identification

Europe | Health & life sciences


LUMINT technology is an innovative diagnostic tool that implements imagingtechnology and artificial intelligence. The project aim is to identify automaticallybacterial pathogens and help clinicians in their therapeutic decisions. The LUMINT@CLINICS project aims to transfer this technology from the laboratory tothe hospital.

Published on 9 April 2021

LUMINT@CLINICS :  Elastic Light Scattering for low cost and non-destructive clinical pathogens identification

LUMINT technology is an innovative diagnostic tool that implements imaging technology and artificial intelligence. The project aim is to identify automatically bacterial pathogens and help clinicians in their therapeutic decisions. The LUMINT@CLINICS project aims to transfer this technology from the laboratory to the hospital.


Starting date : Jan 2019 > Jun 2020  

Lifetime:18 months

Program in support : 


Status project : complete

CEA-Leti's contact 

Sandra Barbier  

> Pierre Marcoux                               



Project Coordinator: CEA-Leti 


  • Laboratorium voor Medische Microbiologie, Medisch diensthoofd, Universieit Gent (BE)

  • Assistance Publique Hôpitaux de Paris 

  • Hôpital Européen Georges Pompidou,Service de microbiologie


  • « Elastic Light Scattering for clinical pathogens
    identification: application to early screening of
    Staphylococcus aureus on specific medium », E. Schultz, V. Genuer, P. Marcoux, O. Gal, C. Belafdil, D. Decq, M. Maurin, S. Morales, SPIE-10479, BIOS, San Francisco, 2019.

  • « Optical forward-scattering for identification of bacteria within microcolonies », P.-R. Marcoux, M. Dupoy, A. Cuer, J.-L. Kodja, A. Lefebvre, F. Licari, R. Louvet, A. Narassiguin and F. Mallard, Applied Microbiology and Biotechnology, 2014, 98, 2243-2254. doi: 10.1007/s00253-013-5495-4.

Investment:  € 0.7 m.

EC Contribution€ 0.7 m.



  • LUMINT technology is an innovative diagnostic tool thatimplements imaging technology and artificial intelligence. The project aim is to identify automatically bacterial pathogens and help clinicians in their therapeutic decisions. The LUMINT@ CLINICS project aims to transfer this technology from the laboratory to the hospital.

  • The LUMINT@CLINICS project is a unique opportunity for transferring LUMINT technology to the application field by:

  1. Building 2 fully integrated prototypes, designed to be user-friendly, based on a laboratory system previously developed by CEA-Leti and successfully assessed under laboratory conditions for 10 standard bacterial species
  2. Installing prototypes in two microbiological laboratories at Georges Pompidou European Hospital (HEGP) and Ghent University Hospital (UZGENT) and training clinicians to operate the instruments themselves
  3. Assessing these prototypes in real conditions, i.e. during routine clinical practice, to build up a clinical sample database for the 40 most common bacterial species representing over 90% of pathogens in hospitals. Large-scale data collection is the first step towards truly demonstrating feasibility and, although it represents a key achievement, it has not yet been reported in the literature
  4. Ensuring availability of 2 harmonized data sets acquired in 2 different laboratories, which will help in drawing up a coherent LUMINT development plan
  5. Assessing identification performance with respect to today’s standard mass spectrometry analysis.

  • CEA-Leti’s outcomes include a general upgrade of the LUMINT chain from instrument to AI algorithms and availability of a clinical database double-validated using a gold standard test. A 1 point increase in TRL index (from 4 to 5) is expected during the project. In addition to technological achievements, CEA-Leti is benefitting from fruitful collaboration with clinician partners, who can share their knowledge of real practice constraints as well as present and future needs.


  • LUMINT is an innovative pathogen identification tool based on a light scattering technique and artificial intelligence. Elastic light scattering technology relies on the fact that a cell illuminated by a laser beam scatters light in a specific pattern. Using the recorded patterns, AI makes it possible to rapidly identify bacteria. A first proof of concept has already been successfully achieved based on acquisition and laboratory analysis of 10 frequent bacteria species.

  • The LUMINT@CLINICS project aims to translate the tool to a clinical level. It is mandatory to assess its performance on clinical isolates and compare them to the gold standard technique (mass-spectrometry or genetic methods). The study involves two hospitals - Assistance Publique-Hopitaux De Paris (APHP) and Ghent University (UZGENT) - where image database can be acquired on the 40 most common pathogens in clinical practice, representing 90% of the samples collected in hospitals. Automatic identification using optimized algorithms will then be performed. 

  • An organizational impact study as well as a detailed business plan setup are also included in the project.

  • The outcome of Lumint@Clinics is decisive to raise the necessary funding for the next development steps: large dataacquisition campaigns and in vitro diagnostic regulatory trials.


  • Bacteria and viruses resistant to antimicrobials are a growing problem in today’s world. Time is of the essence in identifying rapidly multiplying bacteria. Bacteria identification is crucial to optimal therapy and speedy, reliable diagnostics are at the heart of the fight against resistant superbugs.

  • The LUMINT solution can potentially have a huge impact on microbiology laboratory organization. This project may revolutionize processes at microbiological laboratories because it will perform faster, non-destructive, fully automated analysis, potentially facilitating chain identification and antibiotic susceptibility testing. The development of this new pathogen identification system is therefore at the heart of the fight against resistant superbugs.