This repo implements a cell segmentation, tracking, and analysis pipeline using common Python libaries.
python3 segment_and_track.py --config-file=<path to an analysis YAML configuration file>
Analysis runs can be configured using the YAML syntax. For example:
##################################################################################################
# This file defines the settings for the demo cell segmentation and tracking analysis pipeline.
##################################################################################################
---
# STEP 1: Cell segmentation related settings.
segmentation_settings:
# Set to True if we want to run segmentation. If False, segmentation will be skipped.
run_cellpose: True
# The path of the directory that holds the input raw image files to be segmented.
raw_data_dir: "./demo_data"
# The path of the directory that masks will be written to.
output_data_dir: "./demo_data/masks"
# The file pattern to match against (e.g. all files in raw_data_dir matching "*.tif" will be included).
file_pattern: ".tif"
# The cellpose model to use.
cellpose_model: "cyto3"
# An estimated average cell diameter, measured in pixels.
cell_diameter_est: 140
# Other settings.
flow_threshold: 2.19
min_size_mask: 40
cell_prob_threshold: -1
cell_body_channel: 0
cell_nucleus_channel: 2
# STEP 2: Cell tracking related settings.
tracking_settings:
# Set to True if we want to run tracking. If False, tracking will be skipped.
run_trackastra: True
# The path of the directory that holds the input raw image files to be tracked.
raw_data_dir: "./demo_data/masks"
# The path of the directory that tracks will be written to.
output_data_dir: "./demo_data/tracks"
# Whether to use a GFP filter.
use_gfp_filter: True
# Which model to use.
trackastra_model: "greedy_nodiv"
# The maximum distance in pixels the cells are expected to move. Keep low if mostly stationary.
max_track_distance: 50
...