Zebrafish Brain Image Registration

This is a pipeline for zebra fish brain image registration.

Description

This easy-to-use pipeline which can process fish one by one, around 30 minutes for each fish.

Preparation

Input

1. CSV files of ROI coordinates.
2. Atlas downloaded from mapzebrain.
3. 9 Planes tiffile for each fish (or any number of planes you want ).
4. One zstack tiffile for each fish.

Tools

1. Jupyter notebook.
2. Fuji (imageJ).

Prerequisite

1. Motion correction.
2. Confirm the direction of atlas and your zstack being consistent.
3. Copy and save each zstack as nrrd format.

Packages

import matplotlib.pyplot as plt
import os
import numpy as np
import cv2
import tifffile
import glob
import bg_space as bg
import pandas as pd
import time

Path to change

2D registration

In this folder, there are all your inputs of one fish, like zstack and planes.

main_dir = "/media/semmelhacklab/David_Behavior_Experiment/testfish/2023-07-06_F1_lowintensity_test"

Change fish number here.

img2 = tifffile.imread(main_dir+'/F1_zscan.tif')

3D registration

Change your atlas path here.

reference_path = "/media/semmelhacklab/David_Behavior_Experiment/testfish/HSA.nrrd"

Also, the fish name.

with open(sh_reg3d_path,'w') as f:
    s = 'reference='+reference_path+'\n'+\
        'directory_3d='+directory_3d+'\n'+\
        'moving_file=${directory_3d}/F1_zscan.nrrd\n'+\

Check registration

2D: Peaks in similarity plots. Open warped tiffiles and plane tiffiles in ImageJ and compare.

3D: Open warped zstack and atlas in ImageJ, merge them with different colors and check if they match well.

Apply transformation

Check your ROI csv paths, and check the row names before applying transformation.

-You can use this code for changing titles.