Download the internship proposal

Stage type

Master 2 Biomed Engineering or Neuroscience or Cognitive Science. The candidate should be enrolled in French Master 2 or last year of an engineering school in biomedical engineering or signal processing or electronics.

Date and duration of the internship

6 months, first semester 2026, approximately from february to july (plus-minus one month)

Context

Motor adaptation is a crucial process in human life, enabling us to adjust our movements based on sensory feedback. For instance, imagine being asked to play your favorite game of petanque with a set of balls that are heavier than usual. You would need several trials to adapt your movements and achieve the same performance. Understanding motor adaptation has broad applications, from rehabilitation in patients after brain or body injuries to optimizing motor skills in sports, work, and robotics.

Despite its ubiquity, motor adaptation is still not fully understood at the neural level. Specifically the neural representations of the dependence of motor adaptation on the context are poorly understood. Namely, how does one transfer learned behaviours between distinct tasks and movements? E.g. how exactly does it work that one knows how to throw an unfamiliar object to an unfamiliar target, generalizing previous experience with other similar objects and targets?

On the methodical level the difficulty is related to the complexity of the data required: neural and behavioural data has to be collected and analyzed simultaneously, and in addition neither of them is stationary. This requires cleverly designed machine learning approaches going beyond the classical ones.

Objective

The objective of this internship is to prepare a machine learning pipeline to analyse simultaneously collected behavioral and electroencephalography (EEG) and previously collected magnetoencephalography (MEG) data from healthy participants performing a motor adaptation task, as well as using this pipeline to test neuroscience hypotheses about this data. Note that the intern will not collect the data (another intern in the group will), this internship is specifically about programming the data processing pipeline.

Missions

the intern will
● Read the literature on motor adaptation and its neural signatures
● Read and understand SPoC algorithm (source power comodulation)
● Extend and adapt the python code for the existing MEG data processing pipeline
● Perform statistical analysis of the data in Python
● Interact with the PI and a PhD student working on the related project
● Interact with another intern collecting the EEG data
● (Optional, time permitting) Perform advanced machine learning analysis of the data in Python, maximizing cross subject transfer
● (Optional, time permitting) Perform mathematical modeling of the acquired data
● Gain experience in cognitive and motor neuroscience research, python programming, brain computer interfaces, enhancing their skills and CV for future academic research or ML/AI industry opportunities.

Skills required

solid programming skills, Python, machine learning, cognitive science, basic statistics, good level of English

Desired

MNE python, experience with EEG data processing and/or collection, BCIs, signal processing, motor control understanding, basic level of French (only for the better integration within the wider team), git, open science

Monetary compensation

Gratifications de stage (selon la réglementation en vigueur) + remboursement partiel du titre de transport mensuel.

Contact

Dmitrii Todorov, PhD; Chaire Professeur Junior
dmitrii.todorov[at]inserm.fr

Host laboratory

Laboratoire d’Imagerie Biomédicale, LIB, Sorbonne Université, Campus des Cordeliers, 15 rue de l’École de Médecine, 75006 Paris.