M2
Characterization of the role of cardiac interoception in neurofeedback efficacy
Context: In military environments, where high-stress situations are common, maintaining or enhancing cognitive performance is critical for ensuring both the safety and success of combat operations. Similar to physical training, emerging techniques like neurofeedback offer the potential to optimize cognitive abilities in soldiers. Neurofeedback relies on a brain-computer interface that allows individuals to consciously modulate specific brain activities, training their brains to function more efficiently1. By promoting brain plasticity, neurofeedback training can lead to lasting functional changes in the brain, ultimately enhancing cognitive performance1. However, despite its promise, neurofeedback has shown variable efficacy: studies report that between 30% and 50% of participants do not experience cognitive improvement after neurofeedback training2. This variability among individuals poses a significant challenge to developing effective cognitive optimization strategies through neurofeedback. This internship project will test a novel hypothesis: that the quality of interoception—the processes by which the nervous system processes bodily signals3, may be a key factor in determining the effectiveness of neurofeedback. Specifically, individuals with better interoceptive abilities may achieve greater control over brain activity during training, leading to improved learning outcomes. In contrast, those with lower interoceptive quality may struggle to benefit from neurofeedback, limiting its efficacy for them.
Objective: The primary objective of the internship project is to determine whether the quality of interoception can serve as a predictor of neurofeedback efficacy.
Material: The analysis will focus on an existing dataset of 106 participants, 75 of whom underwent neurofeedback training, while 31 participated in sham training as the control group. The data were collected between 2023 and 2024 as part of a research project conducted by the French Armed Forces Biomedical Research Institute (IRBA; ID-RCB: 2022-A02306-37). This dataset includes multilevel measures of cardiac interoception: (1) neurophysiological data, collected using concurrent electroencephalography (EEG) and electrocardiogram (ECG), enabling the analysis of the heartbeat-evoked potential (HEP)4; (2) behavioral data, particularly performance in the heartbeat counting task5; and (3) psychometric data, including self-reported interoception as measured by the Multidimensional Assessment of Interoceptive Awareness questionnaire (MAIA, Version 2)6,7. Among the participants who underwent neurofeedback training, responders and non-responders were identified based on their learning curves, specifically their ability to voluntarily modulate brain activity during the training sessions.
Method: The comparison between neurofeedback responders and non-responders in terms of heartbeat-evoked potential (HEP) features—both in the temporal domain (amplitude and latency) and the frequency domain (spectral power)—will be assessed using a cluster-based permutation approach8. For the behavioral and psychometric measures of cardiac interoception, we will apply classical statistical models as well as more advanced techniques, including machine learning (e.g., classification analysis using artificial neural networks).
Supervision: The internship will be jointly supervised by Michael Quiquempoix, PhD, an expert in neurofeedback and the study's principal investigator, and Charles Verdonk, MD, PhD, a specialist in interoceptive neuroscience.
Aims of the internship
1. Processing of pre-collected data, including EEG, ECG, behavioral, and psychometric data.
2. Implementation of statistical models (such as cluster-based permutation, frequentist, and Bayesian statistics) and machine learning techniques, while conducting analyses in close collaboration with the PhD student involved in data collection and the identification of neurofeedback responders and non-responders.
3. Participation in scientific dissemination activities, with active contributions to the writing of scientific articles.
Profile
We are seeking a talented, creative, and highly motivated student in the fields of neuroscience or cognitive science. The ideal candidate should possess programming skills in MATLAB, R, and/or Python for data processing and analysis, along with a solid understanding of standard statistical methods. Basic knowledge of neurofeedback and interoceptive neuroscience will be considered an asset during the recruitment process. Additionally, the candidate should be able to work effectively in a team and collaborate with researchers from various disciplines.
Recruitment criteria
• Must be enrolled in a Master's program (Master 2 level)
• Selection will be based on the evaluation of the application and interviews
• Candidates must successfully pass the primary security clearance required by the French Ministry of Army' primary security clearance
Practical aspects
• Duration: 6 months
• Starting date: according to the curriculum of the Master
• Gratification: according to legal standards
• Internship place: French Armed Forces Biomedical Research Institute, Brétigny-sur-Orge (91), accessible by public transport (RER C to Brétigny station, followed shuttle service)
Application process
Please send your CV (up to 2 pages) and Master 1 transcripts (if available), and a motivation letter via email to the two supervisors (see contact details below).
The application deadline is October 1, 2026.
Supervisors’ contact information
Michael QUIQUEMPOIX (michael.quiquempoix@gmail.com)
French Armed Forces Biomedical Research Institute (IRBA) - Website of IRBA (in French)
Department REF-AERO | Unit Fatigue and Vigilance
Charles VERDONK (verdonk.charles@gmail.com) - Personal Website
French Armed Forces Biomedical Research Institute (IRBA) - Website of IRBA (in French)
Department Neuroscience and cognitive science | Unity Neurophysiology of stress
References
1 Sitaram, R. et al. Closed-loop brain training: the science of neurofeedback. Nat Rev Neurosci 18, 86 (2017).
2 Alkoby, O., Abu-Rmileh, A., Shriki, O. & Todder, D. Can we predict who will respond to neurofeedback? A review of the inefficacy problem and existing predictors for successful EEG neurofeedback learning. Neuroscience 378, 155-164 (2018).
3 Khalsa, S. S. et al. Interoception and mental health: a roadmap. Biol Psychiatry Cogn Neurosci Neuroimaging 3, 501-513 (2018).
4 Park, H. D. & Blanke, O. Heartbeat-evoked cortical responses: Underlying mechanisms, functional roles, and methodological considerations. Neuroimage 197, 502-511 (2019).
5 Larsson, D. E. O., Esposito, G., Critchley, H. D., Dienes, Z. & Garfinkel, S. N. Sensitivity to changes in rate of heartbeats as a measure of interoceptive ability. Journal of neurophysiology 126, 1799-1813 (2021).
6 Mehling, W. E., Acree, M., Stewart, A., Silas, J. & Jones, A. The Multidimensional Assessment of Interoceptive Awareness, Version 2 (MAIA-2). PLoS One 13, e0208034 (2018).
7 Da Costa Silva, L. et al. Self-reported body awareness: validation of the Postural Awareness Scale and the Multidimensional Assessment of Interoceptive Awareness (version 2) in a non-clinical adult French-speaking sample. Frontiers in Psychology 13, 946271 (2022).
8 Maris, E. Statistical testing in electrophysiological studies. Psychophysiology 49, 549-565 (2012).