pipette
ESEspañol

Whole-bulb LSFEM enables automated reconstruction and quantitative spatial mapping of thousands of olfactory glomeruli

J. E. Rodriguez-Gatica, J. I. Jayakumar, S. I. Gnanasekar, A. Mohammadi, T. Ackels, M. Rothermel, D. Roos, M. K. Schwarz

Preprint

In the authors' words

The glomerular layer of the olfactory bulb (OB) transforms chemical information into the first spatially organized neural representation of odor identity. Yet its three-dimensional architecture remains difficult to quantify at whole-bulb scale. Here, we establish an end-to-end framework integrating light-sheet fluorescence expansion microscopy (LSFEM) with automated reconstruction of thousands of individual glomeruli across intact mouse OBs. LSFEM preserves whole-bulb architecture and supports genetically encoded fluorescence and immunolabelling, providing continuous multichannel input for automated analysis. SegFormer-based segmentation of 2D optical sections was followed by morphology-guided separation and three-dimensional object reconstruction. Across 12 intact bulbs, the workflow yielded a median of 2,677 glomeruli per bulb, approaching the 2,851 reported by complete three-dimensional reconstruction. Object-level maps further revealed that glomeruli spanning a broad range of volumes were intermingled throughout the bulb without evident size-defined anatomical domains. Because each reconstructed glomerulus retains a three-dimensional centroid coordinate, the resulting object map can additionally be used as a specimen-specific spatial reference framework, allowing distances and distributions to be expressed relative to defined glomerular or anatomical positions within the intact OB. Together, whole-bulb LSFEM and automated reconstruction provide a scalable framework for quantitative three-dimensional mapping of glomerular number, morphology, and spatial organization, and the relative positioning of additional anatomical signals within the glomerular architecture.

Main resultThe abstract does not state a limitation.

Appeared: Wednesday, November 18. bioRxiv. Preprint, not yet peer-reviewed.

DOI: 10.64898/2026.11.16.766961