Projects

    Registration based on learned geometric primitives.

    Feature-based image registration approaches have historically fallen out of favour due to the difficulty of reliably extracting anatomical correspondences. Such challenges are now largely overcome thanks to recent advances in deep learning, with segmentation models capable of delivering reliable, fine-grained anatomical delineations in seconds. We leverage these advances to revive feature-based registration, creating alignment tools that are robust and anatomically grounded.

    NimbleReg: deep-learning multi-region surface-based diffeomorphic registration.

    NimbleReg: deep-learning multi-region surface-based diffeomorphic registration.

    NimbleReg is a non-linear registration framework that operates on region boundary surfaces extracted from segmentations. Using a PointNet backbone, it is lightweight. Embedded in the stationary velocity field framework, it enables seamless multi-region blending to produce an overall diffeomorphic transformation.

    NimbleReg: A light-weight deep-learning framework for diffeomorphic image registration.
    A. Legouhy, R. Callaghan, N. Mazet, V. Julienne, H. Azadbakht, H. Zhang.
    preprint.
    NimbleReg: A lightweight deep-learning framework for multi-region surface-based diffeomorphic registration.
    A. Legouhy, R. Toro, H. Zhang.
    IABM, 2026.

    Polaffini: robust and versatile framework for anatomically grounded polyaffine registration.

    Polaffini uses the centroids of segmented regions to efficiently estimate local affine matchings that are combined into an overall polyaffine transformation. Applicable both as standalone registration and as pre-alignment for non-linear methods, it is fast and robust, making it well-suited for integration into neuroimaging pipelines.

    Polaffini: Efficient feature-based polyaffine initialization for improved non-linear image registration
    A. Legouhy, R. Callaghan, H. Azadbakht, H. Zhang.
    IPMI, 2023.
    Polaffini: A feature-based approach for robust affine and polyaffine image registration.
    A. Legouhy, C. Campo, R. Callaghan, H. Azadbakht, H. Zhang.
    Imaging Neuroscience, under review.

    End-to-end diffusion MRI processing with deep learning.

    Diffusion MRI processing, from distortion correction to microstructure parameter estimation, is traditionally slow, taking hours per subject with iterative optimisation. Deep learning offers orders-of-magnitude speedups. We developed fast deep-learning methods addressing each stage of this pipeline, and explored protocol-agnostic approaches formicrostructure estimation.

    Set deep-learning for protocol-agnostic quantitative parameter estimation.

    Set deep-learning for protocol-agnostic quantitative parameter estimation.

    This work introduces a protocol-agnostic method based on a PointNet architecture which is lightweight and possesses nice symmetries. Each voxel is represented as an unordered set of measurement tokens combining the signal with its acquisition parameters.

    Set deep learning for protocol generalisation in machine-learning based brain microstructure estimation.
    L. Kerkelä, A. Legouhy, N. Kraguljac, H. Zhang.
    ISMRM, 2026.

    Eddeep: fast eddy-current distortion correction with deep learning.

    Eddeep is the first tool tackling eddy-current distortion correction with deep learning. It is composed of 2 models in sequence: 1) an image translator to restore correspondences between DW and b=0 volumes of very different contrast and, 2) a physics-based registration model to estimate the distortion and apply correction.

    Eddeep: Fast eddy-current distortion correction for diffusion MRI with deep learning.
    A. Legouhy, R. Callaghan, W. Stee, P. Peigneux, H. Azadbakht, H. Zhang.
    MICCAI, 2024.
    Eddeep: Fast eddy-current distortion correction for diffusion MRI with deep learning.
    A. Legouhy, R. Callaghan, W. Stee, P. Peigneux, H. Azadbakht, H. Zhang.
    ISMRM, 2025.
    Fast susceptibility distortion correction with deep learning.

    Fast susceptibility distortion correction with deep learning.

    This work proposes a semi-supervised, physics-based deep-learning approach for fast susceptibility distortion correction of blip-up/blip-down acquisitions.

    Correction of susceptibility distortion in EPI: a semi-supervised approach with deep learning
    A. Legouhy, M. Graham, M. Guerreri, W. Stee, T. Villemonteix, P. Peigneux, H. Zhang.
    CDMRI @ MICCAI, 2022.

    Atlasing

    Population atlases capture the typical anatomy of a cohort, enabling spatial normalisation and population-level statistical analyses. Spatio-temporal atlases further model anatomical evolution over time. Their construction relies on non-linear registration and on the manipulation of geometric transformations under topology-preserving constraints.

    anima.readthedocs.io/atlasing
    An iterative centroid approach for online atlasing.

    An iterative centroid approach for online atlasing.

    Classical atlas construction must be restarted from scratch when new data arrive. We introduce an iterative-centroid formulation that provides an online update rule to incorporate new subjects incrementally, based on principled manipulation of diffeomorphic transformations in the log-Euclidean framework.

    An iterative centroid approach for diffeomorphic online atlasing.
    A. Legouhy, F. Rousseau, C. Barillot, O. Commowick.
    Transactions on Medical Imaging, 2022.
    Online atlasing using an iterative centroid.
    A. Legouhy, O. Commowick, F. Rousseau, C. Barillot.
    MICCAI, 2019.
    Unbiased diffeomorphic spatio-temporal atlasing.

    Unbiased diffeomorphic spatio-temporal atlasing.

    Studying brain development requires atlases that simultaneously capture the population mean anatomy and its continuous evolution over time. This work proposes a diffeomorphic spatio-temporal atlasing method that jointly enforces geometric unbiasedness across individuals and temporal fidelity across developmental stages.

    Unbiased longitudinal brain atlas creation using robust linear registration and log-Euclidean framework for diffeomorphisms
    A. Legouhy, O. Commowick, F. Rousseau, C. Barillot.
    ISBI, 2019.