Research

Full bibliography on Google Scholar and in my CV.

A vascular code for speed

PhD, 2021–2025. PSL University, at Physics for Medicine Paris, supervised by Sophie Pezet and Mickaël Tanter. CIFRE with Iconeus.

Lagged statistics from the GLM analysis computed from the animal speed contrast.

Place cells, grid cells, head direction cells and speed cells taught us how single neurons encode space. The question of my PhD was whether the same information shows up at the mesoscopic scale, in the blood flow that feeds those neurons, and so becomes visible to whole-brain imaging.

To find out, I imaged freely moving rats with functional ultrasound imaging (fUSI) while they explored an arena, and tracked their behavior with pose estimation. Blood volume in hippocampal and parahippocampal regions follows the animal’s running speed closely enough that speed can be decoded across animals from regional fUSI signals alone. Angular head speed leaves a weaker but visible trace in thalamic nuclei.

Getting there meant solving some practical problems first: reproducible imaging planes across sessions, probe placement, behavioral videos synchronized with the scanner, and the analysis pipeline itself.

Research collaborations

Nature cover featuring the naked mole-rat queen odor study

A queen odour mediates reproductive suppression in a eusocial mammal
Khallaf, M.A., …, Cybis Pereira, F., …, et al. (2026). Nature, 656. doi:10.1038/s41586-026-10772-5

A single queen breeds in a naked mole-rat colony, and a chemical in her scent helps keep it that way.
I analyzed the fUSI data: I registered naked mole-rat brains to the mouse atlas, mapped the brain’s response to the queen odor and ran the statistics.

Cool media: ESPCI news · Nature short

fUSI images before and after motion artifact removal

Robust functional ultrasound imaging in the awake and behaving brain: a systematic framework for motion artifact removal
Le Meur-Diebolt, S., Cybis Pereira, F., et al. (2026). Imaging Neuroscience, 4. doi:10.1162/IMAG.a.1191

This paper benchmarks 792 fUSI denoising strategies across four datasets and recommends two pipelines.
I co-developed the core fUSI analysis software used to build the benchmark framework.

Heart and breathing rate extracted from fUSI data

Physio-fUS: a tissue-motion based method for heart and breathing rate assessment in neurofunctional ultrasound imaging
Zucker, N., Le Meur-Diebolt, S., Cybis Pereira, F., et al. (2025). eBioMedicine, 112. doi:10.1016/j.ebiom.2025.105581

The tissue motion of Doppler recordings is enough to recover heart and breathing rate.
I ran the fUSI recordings on freely moving rats and helped with the related subfigures.

Back to top