We are happy to host students from EPFL and other Universities for projects and internships.
If you are interested in working on a project or doing an internship in our laboratory, please apply as follows: Please send an e-mail to the responsible of the project with Caroline Magnin ([email protected]) and Lisa Fleury ([email protected]) in cc.
Please provide the following information in the e-mail:
- Which project type are you interested in (e.g. bachelor project)?
- In which University program are you enrolled and what year are you in (e.g. 3rd year bachelor student in SV)?
- How long does this type of project usually last and how many ECTS credits do you receive for it (e.g. 3 weeks full-time, 6 ECTS credits)?
- What are the deadlines and presentation modes for the project (e.g. presentation during project and written report afterwards)?
- In what period would you like to perform the project in our laboratory (e.g. 1st of May until 19th of May 2022, full-time)?
- What are your main interests (e.g. healthy aging)?
Please attach your CV and your motivation letter.
After reviewing your application, we will contact you via e-mail regarding further details.
Students from external universities will further need
- Two letters of recommendation
- Funding information
Be aware that, due to extensive administrative paperwork, external only projects lasting at least 6 months will be considered for external students. Further be aware that you might need to meet further administrative requirements (e.g. visum, insurance). You will find information to processes and detailed requirements at: http://sae.epfl.ch/exchange-incoming. Please note that our lab is located in the cantons of Geneva and Valais, NOT Vaud (where the main campus is located). In that sense, the visa/residence permits will have to be dealt with in Geneva or Valais respectively.
IMPORTANT: Please note that our laboratory is located at the Clinique Romande de Réadaptation (CRR/SUVA) in Sion as well as at Campus Biotech in Geneva, Switzerland. All of the projects/internships will therefore be taking place in Sion or in Geneva respectively.
Title: Dynamic coactivation patterns of motor networks in stroke patients during hand-gripper tasks
Availability: From Spring/Summer 2026
Keywords: Stroke, task-based fMRI, hand-gripper task, Motor Cortex, Coactivation Patterns (CAPs), Dynamic Functional Connectivity, Temporal Dynamics, Network Reorganization, Recovery
Location: CRR Suva, Sion
Profile: master semester project
Contact: Giorgia Baron ([email protected]) & Lisa Fleury ([email protected])
Duration: semester
Objective: Stroke disrupts the spatiotemporal organization of motor networks, but understanding how motor activity unfolds over time during movement remains a challenge. Task-based fMRI provides a direct window into motor network dynamics during action, revealing transient bursts of activity and coactivation across brain regions.
This project will focus on time-resolved analysis of motor network dynamics during hand-gripper tasks in stroke patients, using coactivation pattern (CAP) analysis. CAPs identify recurring patterns of coactivation across the brain during task execution, allowing characterization of how motor networks are engaged, reorganized, or impaired after stroke.
Specific goals:
1. Extract task-based CAPs: Identify transient coactivation patterns in the motor cortex and connected networks during hand-gripper execution in stroke patients.
2. Characterize network dynamics: Quantify temporal properties of CAPs such as duration, frequency, and transitions to assess how stroke alters the organization of motor networks.
3. Relate CAP dynamics to recovery outcomes: Correlate CAP measures with behavioral assessments of motor performance (e.g., grip strength, motor scores) and rehabilitation progress.
Prerequisite: Strong computational skills, interest in neuroimaging and proficiency in Python or MATLAB. Familiarity with fMRI preprocessing, time-resolved functional connectivity is a plus.
How to apply: Please send your CV and motivation letter to Giorgia Baron & Lisa Fleury (see Contact)
Title: From slow waves to perturbational complexity: nonequilibrium brain dynamics after stroke
Availability: From January 2026
Keywords: Stroke, TMS-EEG, fMRI, Slow Waves, Effective Connectivity, Recovery, Perturbational Complexity Index (PCI)
Location: CRR Suva, Sion
Profile: master thesis
Contact: Giorgia Baron ([email protected]) & Lisa Fleury ([email protected])
Duration: 6 up to 12 months
Objective: After stroke, slow waves can appear during wakefulness, reflecting sleep-like brain activity that disrupts communication across neural networks and impacts behavior. These slow waves represent nonequilibrium brain dynamics and quantifying their temporal asymmetry can provide insight into network dysfunction and recovery potential. The current project aims to bridge slow-wave dynamics with perturbational complexity, investigating how temporal irreversibility and effective connectivity relate to network-level reorganization after stroke.
Specific goals
1. Characterize slow-wave propagation using TMS-EEG and resting-state fMRI, examining how these patterns relate to observed network dysfunction such as ipsilesional hypersynchrony and contralesional decoupling.
2. Quantify temporal irreversibility of slow-wave dynamics and compare it with perturbational complexity index (PCI) to assess the nonequilibrium nature of post-stroke networks.
3. Link slow-wave topology and timing to behavior, investigating how these dynamics relate to motor, cognitive, or sensorimotor performance.
4. Investigate functional roles of slow waves, determining whether they are solely disruptive or potentially restorative, through frequency-specific analyses of PCI and directed functional connections.
Through this multimodal approach, the student will contribute to understanding how nonequilibrium slow-wave dynamics shape perturbational complexity and network recovery after stroke. If the results are relevant for publication, the student will also be involved in preparing manuscripts and disseminating the findings.
Prerequisite: Strong computational skills, interest in neuroimaging and TMS/EEG/fMRI anaylsis. Proficiency in Python or MATLAB is required.
How to apply: Please send your CV and motivation letter to Giorgia Baron & Lisa Fleury (see Contact)
Title: Linking structural covariance and EEG dynamics after stroke
Availability: From January 2026
Keywords: Stroke, Structural–Functional Connectivity, Structural Covariance Networks, EEG, Recovery
Location: CRR Suva, Sion
Profile: master thesis
Contact: Giorgia Baron ([email protected])
Duration: 6 up to 12 months
Objective: Stroke is a leading cause of long-term disability, often resulting in widespread alterations to both structural and functional brain organization. While focal lesions cause immediate damage, stroke triggers large-scale network reorganization that affects areas beyond the lesion site. The current project aims to investigate how cortical structural organization constrains large-scale functional brain dynamics after stroke, by combining high-density EEG and structural MRI. Specifically, the project will integrate subject-specific structural covariance networks (SCN), derived from cortical thickness similarity, with functional networks obtained from EEG using neuronal avalanche propagation (ATM).
Specific objectives
1. Characterize how SCN–ATM coupling is disrupted in perilesional versus contralesional regions immediately after stroke.
2. Track longitudinal changes in SCN–ATM coupling from acute to chronic stages post-stroke.
3. Assess whether baseline or early changes in SCN–ATM coupling predict recovery trajectories in motor or cognitive domains.
4. Examine the influence of lesion characteristics (location, volume, type) on the disruption and recovery of SCN–ATM coupling.
Through this multimodal approach, the student will contribute to understanding how structural connectivity shapes functional dynamics post-stroke, with potential implications for individualized prognosis and personalized rehabilitation strategies. If the results are relevant for publication, the student will also be actively involved in preparing manuscripts and contributing to the dissemination of the findings.
Prerequisite: Strong computational skills, interest in neuroimaging and proficiency in Python or MATLAB.
How to apply: Please send your CV and motivation letter to Giorgia Baron (see Contact)
Title: Hybrid EEG–fMRI mapping after stroke
Availability: From January 2026
Keywords: Stroke, EEG–fMRI, Functional Connectivity, connICA, Resting-State Networks, Recovery
Location: CRR Suva, Sion
Profile: master thesis
Contact: Giorgia Baron ([email protected])
Duration: 6 up to 12 months
Objective: Stroke often leads to widespread disruptions of functional brain networks, with effects extending beyond the primary lesion. Understanding how these disruptions manifest across different timescales is crucial for predicting recovery and designing individualized rehabilitation strategies. The current project aims to investigate joint EEG–fMRI functional patterns after stroke, coupling resting-state recordings to capture both fast electrophysiological and slower hemodynamic dynamics. Specifically, the project will apply hybrid connectivity independent component analysis (connICA) to resting-state EEG and fMRI data. This approach allows the identification of spatially independent functional networks that are linked across modalities, moving beyond simple correlations between EEG and fMRI.
Specific objectives
1. Identify joint EEG–fMRI connectivity components that differentiate stroke-affected regions from healthy regions.
2. Track longitudinal changes in these joint connectivity networks during recovery after stroke.
3. Determine whether specific EEG frequency fingerprints are linked to disrupted fMRI intrinsic connectivity networks in stroke patients.
Through this multimodal analysis, the student will help uncover how functional connectivity across timescales is altered by stroke and how these changes evolve during recovery. If the results are relevant for publication, the student will also be actively involved in preparing manuscripts and contributing to the dissemination of the findings.
Prerequisite: Strong computational skills, interest in neuroimaging and proficiency in Python or MATLAB.
How to apply: Please send your CV and motivation letter to Giorgia Baron (see Contact)
Title: Longitudinal analysis of post-stroke brain dynamics using DySCo
Availability: From January 2026
Keywords: Stroke, fMRI, TMS-EEG, Dynamic Functional Connectivity, DySCo, Temporal Covariance, Network Reorganization, Recovery
Location: CRR Suva, Sion
Profile: master thesis
Contact: Giorgia Baron ([email protected])
Duration: semester
Objective: After a stroke, the brain undergoes complex reorganization, with network interactions evolving over time. Capturing these dynamics is crucial for understanding recovery mechanisms and predicting functional outcomes. Dynamic Functional Connectivity (dFC) offers a time-resolved view of brain networks, but traditional methods are computationally intensive and difficult to interpret for high-dimensional or longitudinal datasets. The Dynamic Symmetric Connectivity Matrix (DySCo) framework provides a unified, efficient, and theoretically grounded approach to study dFC. This project aims to apply DySCo longitudinally to stroke patient datasets to characterize how brain network dynamics evolve during recovery.
Specific goals:
- Compute longitudinal dFC: Use DySCo to derive time-varying connectivity matrices from fMRI across multiple post-stroke time points.
- Quantify network properties: Calculate DySCo measures such as connectivity strength, matrix similarity, and informational complexity to track changes in network organization over time.
- Relate network dynamics to recovery: Correlate DySCo measures with behavioral, motor, and cognitive assessments to identify patterns predictive of functional recovery.
- Evaluate computational efficiency: Demonstrate DySCo’s ability to handle high-dimensional, voxel-level datasets efficiently, enabling analyses not feasible with conventional dFC methods.
Through this project, the student will gain hands-on experience in longitudinal neuroimaging analysis, advanced computational methods, and brain network modeling. Contributions may lead to publications and presentations.
Prerequisite: Strong computational skills, interest in neuroimaging and proficiency in Python or MATLAB. Familiarity with high-dimensional data analysis is highly recommended.
How to apply: Please send your CV and motivation letter to Giorgia Baron (see Contact)
Title: Dissecting stimulation-induced mechanisms of post-stroke brain reorganization through BOLD co-fluctuation dynamics
Availability: From January 2026
Keywords: Stroke, Resting-state fMRI, BOLD Dynamics, Neuronal Cascades, Co-Activation Events, Functional Connectivity States, Temporal Interference Stimulation, Striatum
Location: CRR Suva, Sion
Profile: master thesis
Contact: Giorgia Baron ([email protected]) and Camille Proulx ([email protected])
Duration: 6 up to 12 months
Objective: At rest, the brain exhibits rich spatiotemporal dynamics characterized by recurrent functional connectivity (FC) states evolving on slow timescales. Recent theoretical and experimental work suggests that these dynamics emerge from fast neuronal cascades that propagate through the structural connectome BOLD co-fluctuations that shape resting-state networks (RSNs). After stroke, these multiscale mechanisms are disrupted, and neuromodulation techniques such as temporal interference (TI) stimulation may alter the pathological propagation of neuronal cascades and the resulting BOLD dynamics. However, it remains unclear which aspects of post-stroke network reorganization are driven by spontaneous recovery versus direct stimulation-induced mechanisms. This project aims to apply a cascade-based framework of BOLD dynamics to resting-state fMRI data from stroke patients acquired with and without temporal interference stimulation of the striatum, in order to identify stimulation-specific alterations in brain dynamics and functional connectivity.
Specific goals
1. Characterize BOLD co-fluctuation events
Identify and quantify intermittent epochs of high-amplitude BOLD co-activation (CA) bursts in resting-state fMRI data from stroke patients, comparing stimulated and non-stimulated conditions.
2. Analyze FC state dynamics
Examine how CA bursts shape time-resolved FC states and resting-state networks, assessing their stability, recurrence, and spatial organization across conditions.
3. Link stimulation to neuronal cascade mechanisms
Test whether temporal interference stimulation alters the frequency, spatial extent, or propagation patterns of BOLD co-fluctuations, consistent with changes in underlying neuronal cascades.
4. Disentangle recovery vs stimulation effects
Compare longitudinal or cross-condition data to distinguish spontaneous post-stroke reorganization from stimulation-driven modifications of network dynamics.
Prerequisite: Strong computational skills and interest in neuroimaging. Proficiency in Python or MATLAB is required. Familiarity with resting-state fMRI analysis, time-resolved functional connectivity, or signal processing is a plus.
How to apply: Please send your CV and motivation letter to Giorgia Baron or Camille Proulx (see Contact).
Title: Behavioral dynamics of hand motor performance in stroke patients: longitudinal and virtual lesion analyses
Availability: From Spring 2026
Keywords: Stroke, task-based fMRI, hand-gripper task, Motor Cortex, Coactivation Patterns (CAPs), Dynamic Functional Connectivity, Temporal Dynamics, Network Reorganization, Recovery
Location: CRR Suva, Sion
Profile: bachelor or master
Contact: Giorgia Baron ([email protected]) & Lisa Fleury ([email protected])
Duration: Semester
Objective: Motor recovery after stroke can be tracked not only through imaging but also through behavioral performance during controlled tasks. The hand-gripper task provides precise measurements of motor output, including force control, timing, and variability. In combination with TMS-induced virtual lesions, these data allow for causal investigation of motor network function and compensation after stroke.This project will focus on analyzing behavioral data from stroke patients performing an isometric hand-gripper task, comparing performance across time points (longitudinal) and/or conditions (baseline vs. TMS virtual lesion).
Specific goals:
1. Characterize motor performance over time: Quantify grip force accuracy, variability, and timing across multiple post-stroke sessions to identify trajectories of recovery within patients.
2. Compare conditions within subjects: Assess how TMS-induced virtual lesions affect task performance, testing hypotheses about motor network compensation and residual function.
3. Identify behavioral markers of recovery: Explore whether specific metrics (e.g., force consistency, reaction time) predict rehabilitation outcomes or correlate with lesion location or severity.
Prerequisite: Computational skills and proficiency in Python, R, or MATLAB. Experience with behavioral data analysis, statistics, or time-series analysis is highly recommended. Familiarity with motor neuroscience or stroke rehabilitation is a plus.
How to apply: Please send your CV and motivation letter to Giorgia Baron & Lisa Fleury (see Contact).