Physip anticipates a marathon runner’s downtime by analyzing his brain activity (Meegaperf algo)

In partnership with Billat Training and the Integrative Biology of Adaptations to Exercise unit of Evry University, Physip has succeeded in establishing the link between brain activity and the risk of a high-level athlete stopping during a marathon, using the technology developed during the MEEGAPERF project, supported by the Defense Innovation Agency (AID/DGA) as part of the RAPID program.
The Meegaperf algorithm (AI rule-based, explainable) analyzes the brain’s electrical activity (electroencephalogram, EEG) recorded via a single EEG lead (2 sensors), to produce every 2 seconds an assessment of the risk of degraded performance in 4 levels, from no risk to high risk of stopping exercise due to exhaustion.

On the occasion of the Fête de la Science 2023, a marathon runner was equipped with sensors to measure physiological and mechanical variables (e.g. blood flow, oxygen consumption, temperature, heart rate, speed) as well as … EEG sensors.

Even if the subject has been equipped with 16 EEG sensors for the purposes of the research laboratory, Physip’s algorithm produces its fully automatic interpretation solely from the 2 sensors located in Cz and Pz, no other information being used.

This conclusive experiment is fully in line with Physip’s mission to take the EEG out of the lab and into “real life”. This result is in line with the central governor theory, in which the brain plays the role of effort regulator, as opposed to performance regulation by fatigue, attributed until recently to reaching cardiac limits.

Measurements

The graph below shows the sequence of measurements taken during the run, as well as the results obtained in terms of the risk of stopping exercise.

Figure – From top to bottom: 1) evolution of the EEG marker, produced by the Meegaperf algorithm from the analysis of a single EEG channel. 2) Runner’s cadence. 3) Heart rate. 4) Runner speed. 5) Volume of O2 (blue) and CO2 (red). 6) Meegaperf algorithm output, Risk of Impaired Performance (RoIP) in 4 risk levels, from no risk (dark green) to high risk of exercise cessation due to exhaustion (red).

It shows the red alert of the EEG marker as early as the 40-60th minute (1), with a consequent reduction in pace (2) and running speed (3), enabling the marathon runner to finish his race without breaking his metabolic steady state (the last 45 minutes of the race are missing due to a technical problem during the runner’s last refuelling ).

This analysis by Meegaperf was carried out within the more general framework of the MISS experiment, where other scientific and technical questions were raised

> Watch the video of the first marathon CerVO2max

> Read an article on Meegaperf (in French) Un système pour détecter l’épuisement physique et mental du soldat