How to convert between CCF and IBL coordinates

This guide shows you how to convert points between the Allen CCF and the IBL coordinates relative to bregma, for example to turn the output of a tracing tool into xyz_picks.json.

Prerequisites

  • iblatlas, which is installed with the GUI.

Steps

  1. Create the atlas:

    import numpy as np
from iblatlas.atlas import AllenAtlas

atlas = AllenAtlas(25)

  2. Check the units and axis order of your points. ccf2xyz takes CCF points in µm and returns IBL points in m; xyz2ccf takes IBL points in m and returns CCF points in µm. Both take a ccf_order: 'mlapdv' or 'apdvml', the order of the axes of the CCF points.

  3. Convert:

    CCF to IBL

    ccf_um = np.array([[3000, 4000, 3000], [6000, 6000, 500]], dtype=float)  # ML, AP, DV
xyz_m = atlas.ccf2xyz(ccf_um, ccf_order='mlapdv')

ccf_um = np.array([[3000, 4000, 3000], [6000, 6000, 500]], dtype=float)  # AP, DV, ML
xyz_m = atlas.ccf2xyz(ccf_um, ccf_order='apdvml')

    IBL to CCF

    xyz_m = np.array([[2000, 4000, 0], [4000, -1000, -4000]]) / 1e6   # ML, AP, DV in m

ccf_um = atlas.xyz2ccf(xyz_m, ccf_order='mlapdv')   # returned as ML, AP, DV
ccf_um = atlas.xyz2ccf(xyz_m, ccf_order='apdvml')   # returned as AP, DV, ML

    The IBL points are always ordered ML, AP, DV.

  4. For xyz_picks.json, multiply the IBL points by 1e6 to give µm, and save them under the key xyz_picks (see How to export a track from brainreg-segment for an example).