Machine Learning and Muscle Injury Prediction


Project context:

The current project is part of the recruitments provided for in the FULGUR project, led by Gaël GUILHEM (SEP laboratory, INSEP), supported by the National Research Agency as part of the Priority Very High Performance Sports Research Program. The post-doctoral student is placed under the direct authority of the director of the Inter-university Laboratory of Human Movement Science (LIBM EA 7424). FULGUR brings together 9 research establishments, 3 sports federations and 2 companies.

This project has three main objectives:

♦ Describe the mechanics of sprinting at the center of mass and joint segments, in order to quantify the specific training load for sprinting, at these scales, in real training or even competition conditions (work package 1);

♦ Determine the musculoskeletal profile of each very high-level athlete with a view to proposing "tailormade" training programs aimed at optimizing the effectiveness of propulsion in sprinting (work package 2);

♦ Estimate the level of risk of injury and suggest individualized prevention strategies based on a multifactorial approach including mechanical aspects, the environment (nutrition, sleep) and the behavior of athletes (work package 3).
These objectives will be supported by transversal work packages aimed at improving the analysis of musculoskeletal imaging and sports gestures using ultrasound techniques and machine learning. The work carried out must contribute to produce new knowledge capable to change the practices of sportspeople / coaches through innovative transfer supports made available to high-level sport players. The coordination of the project between the different partners will be ensured by a research engineer (project manager).


Main activities

♦ Participation in the drafting of research protocols and requests for authorization to the ethics and / or Personal Protection committees on the data analysis part;

♦ Support of databases in connection with the 2nd post-doctoral fellow;

♦ Processing and analysis of numerical experimental data collected in WP1 and WP2, numerical and qualitative experimental data collected in WP3, development and validation of predictive algorithms using machine learning techniques.