Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jun 23, 2026

New Framework for Understanding Cross-Brain Coherence in Functional Near-Infrared Spectroscopy (fNIRS) Hyperscanning Studies
05:59

New Framework for Understanding Cross-Brain Coherence in Functional Near-Infrared Spectroscopy (fNIRS) Hyperscanning Studies

Published on: October 6, 2023

BiXformer: A Bidirectional Cross Attention Transformer for Disentangling Inter-Regional Neural Dynamics.

Omar El Sayed, Yujin Han, Tudor Dragoi

    Biorxiv : the Preprint Server for Biology
    |June 22, 2026
    PubMed
    Summary

    We developed BiXformer, a novel AI tool, to untangle complex brain communication signals. This method accurately deciphers directed neural interactions and timing delays in large-scale neural recordings.

    Related Concept Videos

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    Diurnal rhythms of choice: a novel state-dependent drift diffusion model uncovers time-dependent changes in rat decision making.

    bioRxiv : the preprint server for biology·2026
    Same author

    Distinct cell-type contributions and network topography of theta-nested gamma oscillations in the medial entorhinal cortex.

    bioRxiv : the preprint server for biology·2026
    Same author

    Low rank adaptation of chemical foundation models generates effective odorant representations.

    bioRxiv : the preprint server for biology·2026
    Same author

    Dynamic engagement of the motor cortex in controlling movement.

    bioRxiv : the preprint server for biology·2026
    Same author

    Mapping neural subspace dynamics onto the structure of the mouse descending motor system.

    bioRxiv : the preprint server for biology·2025
    Same author

    Recurrent connectivity supports carbon dioxide sensitivity in <i>Aedes aegypti</i> mosquitoes.

    bioRxiv : the preprint server for biology·2025

    Area of Science:

    • Computational Neuroscience
    • Neurotechnology
    • Machine Learning for Neuroscience

    Background:

    • High-throughput neural recording technologies generate vast datasets from behaving animals.
    • Interpreting complex, bidirectional, and temporally offset inter-regional neural communication is challenging.
    • Existing methods struggle with superimposed feedforward and feedback signals in neural populations.

    Purpose of the Study:

    • To introduce BiXformer, a bidirectional cross-attention transformer, for disentangling inter-regional neural communication.
    • To decompose neural communication into causal and acausal streams using directionally masked attention.
    • To recover directed latent dynamics and estimate communication delays without linearity or stationarity assumptions.

    Main Methods:

    More Related Videos

    Inter-Brain Synchrony in Open-Ended Collaborative Learning: An fNIRS-Hyperscanning Study
    04:44

    Inter-Brain Synchrony in Open-Ended Collaborative Learning: An fNIRS-Hyperscanning Study

    Published on: July 21, 2021

    Related Experiment Videos

    Last Updated: Jun 23, 2026

    New Framework for Understanding Cross-Brain Coherence in Functional Near-Infrared Spectroscopy (fNIRS) Hyperscanning Studies
    05:59

    New Framework for Understanding Cross-Brain Coherence in Functional Near-Infrared Spectroscopy (fNIRS) Hyperscanning Studies

    Published on: October 6, 2023

    Inter-Brain Synchrony in Open-Ended Collaborative Learning: An fNIRS-Hyperscanning Study
    04:44

    Inter-Brain Synchrony in Open-Ended Collaborative Learning: An fNIRS-Hyperscanning Study

    Published on: July 21, 2021

  • Developed BiXformer, a bidirectional cross-attention transformer model.
  • Utilized directionally masked attention to enforce temporal constraints within attention heads.
  • Validated on synthetic datasets with known ground-truth delays and applied to real neural recordings.

    • Main Results:

    • BiXformer accurately recovered low-dimensional, directed latent dynamics and communication delays from synthetic data.
    • The model demonstrated accurate recovery of latent structure and inter-regional timing.
    • Applied to neural-behavioral data, BiXformer revealed interpretable components reflecting sensory feedback and motor signals.

    Conclusions:

    • BiXformer effectively disentangles bidirectional and temporally offset neural communication.
    • The model provides a flexible framework for estimating communication delays and directed dynamics in neural circuits.
    • BiXformer advances the analysis of complex neural recordings, offering insights into brain circuit function.