Poster No:
1998
Submission Type:
Abstract Submission
Authors:
romain valabregue1, Katja Heuer2, david meunier3, Ines Khemir4, Olivier Coulon5, Francois Rousseau6, Guillaume Auzias7, Roberto Toro8, eric bardinet1
Institutions:
1Centre de NeuroImagerie de Recherche–CENIR, Institut du Cerveau - Paris Brain Institute - ICM, Inser, paris, 2Institut Pasteur, Paris, 3Aix-Marseille Université, Institut de Neurosciences de la Timone, UMR 7289, marseille, marseille, 4Data Analysis Core facility, Institut du Cerveau - Paris Brain Institute - ICM, Inserm U 1127, CNRS, paris, 5Aix-Marseille Université, Institut de Neurosciences de la Timone, UMR 7289, Marseille, marseille, France, 6IMT Atlantique, LaTIM INSERM U1101, brest, 7Institut de Neurosciences de la Timone, UMR 7289, Aix Marseille Université, CNRS, Marseille, France, 8Institut Pasteur, Paris, France
First Author:
romain valabregue
Centre de NeuroImagerie de Recherche–CENIR, Institut du Cerveau - Paris Brain Institute - ICM, Inser
paris
Co-Author(s):
david meunier
Aix-Marseille Université, Institut de Neurosciences de la Timone, UMR 7289
marseille, marseille
Ines Khemir
Data Analysis Core facility, Institut du Cerveau - Paris Brain Institute - ICM, Inserm U 1127, CNRS
paris
Olivier Coulon
Aix-Marseille Université, Institut de Neurosciences de la Timone, UMR 7289, Marseille
marseille, France
Guillaume Auzias
Institut de Neurosciences de la Timone, UMR 7289, Aix Marseille Université, CNRS
Marseille, France
eric bardinet
Centre de NeuroImagerie de Recherche–CENIR, Institut du Cerveau - Paris Brain Institute - ICM, Inser
paris
Introduction:
Atlas-based methods for brain segmentation rely on specific templates in order to define spatial priors. Automatic segmentation is thus available only for the few species mostly used in neuroimaging experiments, and adaptation to unseen species requires building of specific atlases, which use might in addition introduce biases in subsequent inter-species comparison studies. Having an efficient single model to segment the brain of many species would be of great interest for evolutionary biology [Milham 2022]. An example of today's limitation is the Brain Catalog database (https://braincatalogue.org) [Heuer 2019] where manual segmentation is still necessary in most cases.
Deep learning methods offer promising alternatives for brain tissue segmentation, but current models are trained for a specific species only. In this work, we evaluate the potential of training on synthetic data generated from a few species, and study whether the model can generalize to unseen species.
Methods:
We used the synthetic framework [Billot 2023] because 1) it alleviates the need for real data for training and 2) the predictions are robust to variations in MRI contrast. We used the same generative process and training model (3D Unet) as described previously [Valabregue 2023]. In this study, we considered 5 different species for training: 1 adult human, 3 newborn humans from the dHCP, 3 fetal rhesus macaque, 1 adult rhesus macaque and 1 mouse lemur [Data]. The objective was to segment 13 structures in all these species (Head / CSF / Ventricles / Gray Matter / White Matter/ Cerebellum Gray Matter / 7 deep nuclei (Caudate Putamen, Pallidum, Thalamus, amygdala, accumbens, Substantia nigra ), all structures being present n the different species. The only exception was for fetal data where the deep nuclei are merged in a single structure. When such examples were seen during training, we merged the models predictions and computed the loss regarding the deep nuclei as a whole.
We validated our results on 5 different test sets : rhesus macaque (5 brains, private dataset), adult HCP (40 brains from the HCP project), dHCP Young (20 youngest subjects from the dHCP), dHCP Old (20 oldest subject from the dHCP) and new species (dog, horse, and 6 different primate species) [ref Data]. For the human test sets, we chose the Freesurfer segmentation as pseudo ground truth, for the macaque we used gray matter extracted with https://github.com/Macatools/macapype and for the new species we used a manual segmentation of the whole brain (without CSF). In this case, we added the predicted label to construct a brain mask for the evaluation.
Results:
Figure 1a) shows the Dice score for the Gray Matter (GM) for the different test sets. Although the performance of the model trained on multiple species was a bit lower compared to a model trained on a specific species, its performance was constant among all species in contrast to the species-specific models. Figure 1b) shows the Dice score for the brain mask, including the same test sets as above with the addition of the new species test set. We obtained similar results as for the GM but now the model trained on multiple species was the best-performing on new unseen species.
A full validation on each tissue for new species is a difficult task because of the lack of ground truth labels, we therefore show some examples of segmentations for a visual quality check (Cf Figure 2).
Conclusions:
Having a unique model to segment the brain in any species is an ambitious objective but with a potential high impact for comparative anatomy and evolutionary neurobiology. We demonstrate its feasibility with synthetic training: the model shows good generalization to unseen species. Our preliminary results are promising and we hope they will motivate additional work towards universal brain segmentation methods.
Lifespan Development:
Normal Brain Development: Fetus to Adolescence 2
Modeling and Analysis Methods:
Segmentation and Parcellation 1
Keywords:
Atlasing
Machine Learning
Morphometrics
Segmentation
Other - Synthetic data; synthetic training
1|2Indicates the priority used for review
Provide references using author date format
[Milham 2022]. Toward next-generation primate neuroscience: A collaboration-based strategic plan for integrative neuroimaging. In Neuron (Vol. 110, Issue 1, pp. 16–20).
[Heuer 2019]. Evolution of neocortical folding: A phylogenetic comparative analysis of MRI from 34 primate species. In Cortex (Vol. 118, pp. 275–291)
[Billot (2023)] “SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining. Medical Image Analysis, 86:102789.
[Valabregue (2023)] “Comprehensive analysis of synthetic learning applied to neonatal brain MRI segmentation”. arXiv preprint arXiv:2309.05306.
[Data] for training : Mida template : https://itis.swiss/virtual-population/regional-human-models/mida-model/mida-v1-0/
dhcp : http://www.developingconnectome.org/
fetal rhesus macaque ONPRC.18: https://www.nitrc.org/projects/onprc18_atlas
adult rhesus macaque postmortem: https://www.civm.duhs.duke.edu/rhesusatlas
Mouse Lemur Primate : https://www.nitrc.org/projects/mouselemuratlas
for testing : HCP release_v3 https://www.humanconnectome.org/
Equine: https://www.johnsonlabcornell.com/equine-brain
Canine: https://www.johnsonlabcornell.com/copy-of-corpus-callosum-malformations .
small monkey [Heuer 2019]