OptimizeParams: Optimize Detection Params

In ocbe-uio/cellmigRation: Track Cells, Analyze Cell Trajectories and Compute Migration Statistics

Description

Optimize Detection Parameters for running a cell tracking job

Usage

OptimizeParams(
  tc_obj,
  lnoise_range = NULL,
  min.px.diam = 5,
  diameter_range = NULL,
  threshold_range = NULL,
  target_cell_num = NULL,
  threads = 1,
  quantile.val = NULL,
  px.margin = NULL,
  plot = FALSE,
  verbose = FALSE,
  dryrun = FALSE
)

Arguments

tc_obj	a trackedCells object
lnoise_range	numeric vector of lnoise values to be used in the optimization step. Can be NULL
min.px.diam	integer, minimum diameter of a particle (cell). Particles with a diameter smaller than min.px.diam are discarded
diameter_range	numeric vector of diameter values to be used in the optimization step. Can be NULL
threshold_range	numeric vector of threshold values to be used in the optimization step. Can be NULL
target_cell_num	integer, the expected (optimal) number of cells to be detected in each frame
threads	integer, number of cores to use for parallelization
quantile.val	numeric, argument passed to EstimateDiameterRange(). If NULL, it is defaulted to 0.99
px.margin	numeric, argument passed to EstimateDiameterRange(). If NULL, it ia defaulted to 2
plot	if 'TRUE', plots results in the end
verbose	shall information about the progress of the operation be printed to screen/console
dryrun	shall a dryrun be performed

Details

The lnoise param is used to guide a lowpass blurring operation, while the lobject param is used to guide a highpass background subtraction. The threshold param is used for a background correction following the initial image convolution

• lnoise: Characteristic lengthscale of noise in pixels. Additive noise averaged over this length should vanish. May assume any positive floating value. May be also set to 0, in which case only the highpass "background subtraction" operation is performed.

• lobject Integer length in pixels somewhat larger than a typical object. Can also be set to 0, in which case only the lowpass "blurring" operation defined by lnoise is done without the background subtraction defined by lobject

• threshold Numeric. By default, after the convolution, any negative pixels are reset to 0. Threshold changes the threshhold for setting pixels to 0. Positive values may be useful for removing stray noise or small particles.

Value

a trackedCells object

Author(s)

Damiano Fantini, [e-mail]

References

https://www.data-pulse.com/dev_site/cellmigration/ https://www.mathworks.com/matlabcentral/fileexchange/60349-fasttracks

Examples

x <- get(data(TrackCellsDataset))
x <- OptimizeParams(tc_obj = x, lnoise_range = c(5,7,10),
                    diameter_range = c(12,14,18),
                    threshold_range = c(4,7), dryrun = TRUE)
getOptimizedParameters(x)

ocbe-uio/cellmigRation documentation built on Dec. 16, 2021, 10:59 p.m.