Functional Brain Mapping with Deep Learning
The functional organization of human visual cortex is tightly coupled with underlying anatomy. While previous approaches could predict gross retinotopic organization using atlas-based templates, they were unable to capture the detailed idiosyncrasies seen in individual brains. To address this, I developed a geometric deep learning model — DeepRetinotopy — capable of exploiting the actual structure of the cortex to predict the retinotopic organization of visual cortex from anatomy alone, capturing nuanced individual variations. I have since expanded this work into a toolbox for retinotopic mapping (see News).
Publications
- Ribeiro, Benson & Puckett (2025). Human retinotopic mapping: From empirical to computational models of retinotopy. Journal of Vision.
- Ribeiro, Bollmann, Cunnington & Puckett (2022). An explainability framework for cortical surface-based deep learning. arXiv preprint.
- Ribeiro, Bollmann & Puckett (2021). Predicting the retinotopic organization of human visual cortex from anatomy using geometric deep learning. NeuroImage.
Interindividual Variability in Brain Organization
Are all human brains identical in terms of cortical organization? Not quite. While the general layout of the visual cortex is consistent across individuals, the precise organization of visual field maps can vary significantly. My research has uncovered a previously unknown degree of variability in the organization of V2 and V3 among individuals. Contrary to the traditional view of stereotypical arrangements, only one-third of the studied individuals exhibited the expected pattern; the remaining individuals showcased more complex geometric mappings of the retina onto the cortex. Beyond variability in topographic organization, I'm interested in understanding more broadly what makes individual brains so unique.
Publications
- Ribeiro, York, Zavitz, Bollmann, Rosa & Puckett (2023). Variability of visual field maps in human early extrastriate cortex challenges the canonical model of organization of V2 and V3. eLife.
- Ribeiro, dos Santos, Sato, Pinaya & Biazoli (2021). Inferring the heritability of large-scale functional networks with a multivariate ACE modeling approach. Network Neuroscience.
- Rodrigues, Ribeiro, Sato, Mesquita & Biazoli (2019). Identifying individuals using fNIRS-based cortical connectomes. Biomedical Optics Express.
Tools & Data for Medical Imaging & Open Science
Beyond fundamental neuroscience, I contribute to building and sharing tools and datasets for medical image segmentation and reproducible neuroimaging. This includes VesselBoost, a toolbox for small blood vessel segmentation; the first annotated, openly available MRI dataset of tongue musculature; and contributions to Neurodesk, a community-oriented platform for accessible and reproducible neuroimaging analysis.
Publications
- Ribeiro et al. (2025). An annotated multi-site and multi-contrast MRI dataset for the study of the human tongue musculature. Scientific Data.
- Shaw, Ribeiro et al. (2025). Segmentation of the human tongue musculature using MRI. Computers in Biology and Medicine.
- Dao, Masson-Trottier, Ribeiro et al. (2025). Democratizing open neuroimaging: Neurodesk's approach to open data accessibility and utilization. Aperture Neuro.
- Xu*, Ribeiro* et al. (2024). VesselBoost: A Python toolbox for small blood vessel segmentation in human magnetic resonance angiography data. Aperture Neuro.
- Renton, Dao, ..., Ribeiro et al. (2024). Neurodesk: An accessible, flexible and portable data analysis environment for reproducible neuroimaging. Nature Methods.
Other Interests
Beyond the research areas above, I am broadly interested in geometric deep learning, fairness in AI, and applications of artificial intelligence for neuroimaging. For a complete list of publications, please check out my CV or my Google Scholar.