Mu‐ming Poo

BCIJelly provides a unified ecosystem for BCI research, standardizing data, integrating decoders, and enabling efficient deployment on neuromorphic hardware.

This paper develops a connectome-informed neural network algorithm that improves RNN and LLM performance by incorporating natural cortical connectivity motifs.

This study creates a single-cell spatial transcriptomic atlas of the human cortex, revealing gene expression patterns, aging effects, and disease associations across layers and species.

The paper creates a single-cell atlas of amniote brains to study cell type diversification and identifies species-specific gene expression patterns and novel cell subtypes.

The paper constructs a multimodal brain cell atlas of adult macaques across their lifespan, revealing aging-related molecular changes and identifying aging hotspots and regulatory loci.

The paper identifies 31 projectome-defined subtypes of hypothalamic neurons and six modular subnetworks, revealing the organization of axon projections and their functional roles.

This study shows that anodal tDCS during motor task performance enhances learning specifically for that task in mice, with effects dependent on task-related neuronal activity in the primary motor cortex.

The paper investigates feedback projections in primate early visual cortex, revealing retinotopic organization and differences in connection strength between ventral and dorsal stream areas.

The paper identifies syllable-selective neuronal ensembles in marmoset primary auditory cortex that respond to specific vocalizations, showing sensitivity to global sound properties and sequence structure.

The paper shows that marmoset primary auditory cortex neurons have uniform frequency tuning within local regions, indicating a tonotopic organization at both macro and microscopic levels.