cellfinder takes a stitched, but otherwise raw whole-brain dataset with at least two channels:
- Background channel (i.e. autofluorescence)
- Signal channel, the one with the cells to be detected:
Raw coronal serial two-photon mouse brain image showing labelled cells
Cell candidate detection
Classical image analysis (e.g. filters, thresholding) is used to find cell-like objects (with false positives):
Candidate cells (including many artefacts)
Cell candidate classification
A deep-learning network (ResNet) is used to classify cell candidates as true cells or artefacts:
Cassified cell candidates. Yellow - cells, Blue - artefacts
Registration and segmentation (amap)
Using amap, cellfinder aligns a template brain and atlas annotations (e.g. the Allen Reference Atlas, ARA) to the sample allowing detected cells to be assigned a brain region.
This transformation can be inverted, allowing detected cells to be transformed to a standard anatomical space.
ARA overlaid on sample image
Analysis of cell positions in a common anatomical space
Registration to a template allows for powerful group-level analysis of cellular disributions. (Example to come)