• Research,
  • Future Health,

MILCOM Project: developing machine learning methods for the acquisition, annotation and analysis of medical images

Published on June 30, 2021 Updated on July 1, 2021
With funding from the Connect Talent call for projects in the Pays de la Loire Region, Nantes Métropole and the ERDF, the MILCOM project combines data sciences and health, and focuses on the application of machine learning to analyse multimodal medical images for the validation and identification of biomarkers in oncology.

The research work is being carried out by a multidisciplinary team from the LS2N, in collaboration with the Nuclear Medicine Department at Nantes University Hospital and the CRCINA team at Inserm, and is supervised by Diana Mateus, professor at Centrale Nantes and member of the Signal, Image and Sound (SIMS) team at LS2N.

The project aims to develop machine learning methods for the acquisition, annotation and analysis of medical images to assist doctors in their decision-making.
 

To carry out this research work, the team benefits from the research facilities at Centrale Nantes and relies, for example, on the school's LIGER supercomputer for machine learning. The team works in collaboration with several startups, notably Hera Mi, which develops and markets advanced medical imaging software solutions for breast cancer screening, and Keosys, which specialises in medical imaging.

Ultimately, the MILCOM project should help oncologists to diagnose and provide personalised treatment for patients suffering from diseases such as multiple myeloma.


Other research prospects

The MILCOM project team has started working with the MIPS laboratory (Movement, Interactions, Performance) and the STAPS Department at the University of Nantes on the ultrasound imaging of muscles.

MILCOM's work aims to partially automate the analysis of these images in order to measure quantities such as muscle volume, which are indicative of development in the performance of high-level athletes or the state of progress of degenerative diseases such as Duchenne muscular dystrophy. With the purchase of an ultrafast research ultrasound machine, the MILCOM team is also interested in methods of reconstructing these ultrasound images in order to improve the quality and accuracy of 2D and 3D images.


 


Published on June 30, 2021 Updated on July 1, 2021