perform cellpose segmentation

Perform the Giotto Wrapper of cellpose segmentation. This is for a model inference to generate segmentation mask file from input image. main parameters needed

doCellposeSegmentation(
  input,
  mask_output,
  python_env = "giotto_segmentation",
  channel_1 = 0,
  channel_2 = 0,
  model_name = "cyto3",
  batch_size = 8,
  resample = TRUE,
  channel_axis = NULL,
  z_axis = NULL,
  normalize = TRUE,
  invert = FALSE,
  rescale = NULL,
  diameter = NULL,
  flow_threshold = 0.4,
  cellprob_threshold = 0,
  do_3D = FALSE,
  anisotropy = NULL,
  stitch_threshold = 0,
  min_size = 15,
  max_size_fraction = 0.4,
  niter = NULL,
  augment = FALSE,
  tile_overlap = 0.1,
  bsize = 224,
  dP_smooth = 0L,
  interp = TRUE,
  compute_masks = TRUE,
  progress = NULL,
  verbose = NULL,
  ...
)

Arguments

input

character, required. Provide a path to a gray scale or a three channel image.

mask_output

required. Provide a path to the output mask file.

channel_1

channel number for cytoplasm, default to 0(gray scale)

channel_2

channel number for Nuclei, default to 0(gray scale)

model_name

Name of the model to run inference. Default to 'cyto3', if you want to run cutomized trained model, place your model file in ~/.cellpose/models and specify your model name.

batch_size

Cellpose Parameter, Number of 224x224 patches to run simultaneously on the GPU. Can make smaller or bigger depending on GPU memory usage. Defaults to 8.

resample

(logical, optional) – run dynamics at original image size (will be slower but create more accurate boundaries). Defaults to True.

channel_axis

(int, optional) – channel axis in element of list x, or of np.ndarray x. if NULL, channels dimension is attempted to be automatically determined. Defaults to NULL.

z_axis

(int, optional) – z axis in element of list x, or of np.ndarray x. if NULL, z dimension is attempted to be automatically determined. Defaults to NULL.

normalize

Logical or list. Controls image normalization: If TRUE, normalizes to 1st-99th percentile Can be a list with parameters: * lowhigh: Numeric vector c(low, high) for manual normalization values * sharpen: Numeric, image sharpening factor (1/4-1/8 of cell diameter in pixels) * normalize: Logical, whether to run normalization * percentile: Numeric vector c(low_perc, high_perc) * tile_norm: Integer, window size in pixels for tile normalization * norm3D: Logical, normalize across full z-stack Default: TRUE

invert

Logical. Inverts pixel intensity before processing. Default: FALSE

rescale

Numeric. Resize factor for images. Default: NULL (sets to 1.0)

diameter

Numeric. Cell diameter for each image. Default: NULL (uses diam_mean or diam_train if available)

flow_threshold

Numeric. Flow error threshold for cell retention (2D only). Default: 0.4

cellprob_threshold

Numeric. Threshold for mask pixel retention. Default: 0.0

do_3D

Logical. Enable 3D segmentation for 3D/4D inputs. Default: FALSE

anisotropy

Numeric. Z-axis rescaling factor for 3D segmentation. Default: NULL

stitch_threshold

Numeric. Threshold for 3D mask stitching (2D mode only). Default: 0.0

min_size

Integer. Minimum ROI size in pixels. Default: 15

max_size_fraction

Numeric. Maximum mask size as fraction of image. Default: 0.4

niter

Integer. Number of dynamics computation iterations. Default: NULL (set based on diameter)

augment

Logical. Enable tile augmentation with overlapping. Default: FALSE

tile_overlap

Numeric. Tile overlap fraction for flow computation. Default: 0.1

bsize

Integer. Block size for tiles (recommended: 224). Default: 224

dP_smooth

Integer. Gaussian smoothing standard deviation for 3D flows.

interp

Logical. Enable interpolation for 2D dynamics. Default: TRUE

compute_masks

Logical. Compute dynamics and return masks. Default: TRUE

progress

progress bar. Defaults to NULL

python_path

python environment with cellpose installed. default = "giotto_segmentation".

Value

No return variable, as this will write directly to output path provided.

Examples

doCellposeSegmentation(
    input = input_image,
    mask_output = output, channel_1 = 2,
    channel_2 = 1, model_name = "cyto3", batch_size = 4
)
#> python already initialized in this session
#>  active environment : '/usr/bin/python3'
#>  python version : 3.12
#> specified py env from `envname` = 'giotto_segmentation' not found
#> Error: package 'cellpose' is not yet installed
#> 
#>  To install:
#> ## active python env: '/usr/bin/python3' 
#> ## python version: 3.12
#> ## restart session then use GiottoClass::set_giotto_python_path() if this is incorrect
#> reticulate::conda_install(envname = '/usr/bin/python3', packages = c('cellpose'), pip = TRUE)