run_otsu_discordance {mxnorm}

Calculate Otsu discordance scores using specified threshold for an mx_dataset object.

Description

The Otsu discordance analysis quantifies slide-to-slide agreement by summarizing the distance between slide-level Otsu thresholds and the global Otsu threshold for a given marker in a single metric.

Usage

run_otsu_discordance(
  mx_data,
  table,
  threshold_override = NULL,
  plot_out = FALSE,
  ...
)

Arguments

mx_data

mx_dataset object used to compute Otsu discordance scores

table

dataset in mx_data used to compute metrics. Options include: c("raw","normalized","both"), e.g. a y-axis parameter.

threshold_override

optional user-defined function or alternate thresholding algorithm adaptable from Python skimage module filters (Note: not all algorithms in filters adapted). Options include supplying a function or any of the following: c("isodata", "li", "mean", "otsu", "triangle","yen"). More detail available here: https://scikit-image.org/docs/dev/api/skimage.filters.html. If using a user-defined function, it must include a thold_data parameter.

plot_out

boolean to generate Otsu discordance plots (default=FALSE)

...

optional additional arguments for Otsu discordance functions

Value

mx_dataset object with analysis results of Otsu discordance in otsu_data table

References

Otsu, N. (1979). A threshold selection method from gray-level histograms. IEEE transactions on systems, man, and cybernetics, 9(1), 62-66.

Examples

mx_data = mx_dataset(mxnorm::mx_sample, "slide_id", "image_id",
  c("marker1_vals","marker2_vals","marker3_vals"),
  c("metadata1_vals"))
mx_data = mx_normalize(mx_data, transform="log10",method="None")
## using `threshold_override` here in case users haven't installed `scikit-image`
mx_data = run_otsu_discordance(mx_data, table="normalized",
threshold_override = function(thold_data){quantile(thold_data, 0.5)})