Property presets

Presets are named bundles of property functions that cover the most common phenotyping needs. Pass a preset name to compute_properties and PhenoMe will compute the corresponding features for every image in the dataset.

Start with a preset and add custom functions only when you need something the preset does not cover.

Choosing a preset

Preset	Requires	Best for
"none"	—	Custom-only pipelines; pair with additional_property_functions.
"intensity"	Image only	No segmentation masks; global illumination/signal QC.
"shape"	Mask only	Shape analysis without intensity information.
"basic"	Image + mask	General-purpose starting point; covers shape and signal in one pass.
"standard"	Image + mask	More complete description; adds axis lengths and radial intensity rings.
"complete"	Image + mask	Maximum feature set; adds texture, QC metrics, and extended shape.

The “Requires” column indicates what the preset needs to run. Presets that require a mask return NaN for mask-dependent features if no mask is found for an image.

Properties per preset

"none"

No properties are computed. Use this when you want to supply all functions yourself via additional_property_functions.

---

"intensity"

Image-level statistics over all pixels (no mask required).

Column	Description
intensity_mean	Mean pixel intensity across the whole image.
intensity_std	Standard deviation of pixel intensity.
intensity_min	Minimum pixel intensity.
intensity_max	Maximum pixel intensity.

---

"shape"

Shape descriptors of the segmented object (mask required).

Column	Description
area	Foreground pixel count.
major_axis_length	Length of the major axis of the fitted ellipse.
minor_axis_length	Length of the minor axis of the fitted ellipse.
perimeter	Boundary length of the object.
eccentricity	Elongation of the fitted ellipse (0 = circle, 1 = line).
solidity	Ratio of area to convex hull area (1 = fully convex).

---

"basic"

A lightweight but well-rounded set of intensity and shape features.

Image-level (image required):

Column	Description
intensity_mean	Mean pixel intensity across the whole image.
intensity_std	Standard deviation of pixel intensity.

Shape (mask required):

Column	Description
area	Foreground pixel count.
perimeter	Boundary length of the object.
eccentricity	Elongation of the fitted ellipse (0 = circle, 1 = line).
solidity	Ratio of area to convex hull area (1 = fully convex).

Masked intensity (image + mask required):

Column	Description
intensity_mean_masked	Mean intensity inside the segmented object.
intensity_std_masked	Intensity standard deviation inside the object.

---

"standard"

Extends "basic" with axis lengths, full intensity range, and a 3-ring radial intensity profile. Recommended default for most analyses.

Image-level (image required):

Column	Description
intensity_mean	Mean pixel intensity across the whole image.
intensity_std	Standard deviation of pixel intensity.
intensity_min	Minimum pixel intensity.
intensity_max	Maximum pixel intensity.

Shape (mask required):

Column	Description
area	Foreground pixel count.
major_axis_length	Length of the major axis of the fitted ellipse.
minor_axis_length	Length of the minor axis of the fitted ellipse.
perimeter	Boundary length of the object.
eccentricity	Elongation of the fitted ellipse (0 = circle, 1 = line).
solidity	Ratio of area to convex hull area (1 = fully convex).

Masked intensity + radial profile (image + mask required):

Column	Description
intensity_mean_masked	Mean intensity inside the segmented object.
intensity_std_masked	Intensity standard deviation inside the object.
ring_1_mean	Mean intensity in the outermost ring (membrane region).
ring_1_std	Intensity std in the outermost ring.
ring_2_mean	Mean intensity in the middle ring (cytoplasm region).
ring_2_std	Intensity std in the middle ring.
ring_3_mean	Mean intensity in the innermost ring (core/nucleus region).
ring_3_std	Intensity std in the innermost ring.

---

"complete"

The most comprehensive preset. Adds texture (GLCM), QC metrics, extended shape descriptors, and masked min/max on top of "standard".

Image-level (image required):

Column	Description
intensity_mean	Mean pixel intensity across the whole image.
intensity_std	Standard deviation of pixel intensity.
intensity_min	Minimum pixel intensity.
intensity_max	Maximum pixel intensity.
sharpness_metric	Focus quality estimate (0 = sharp, 1 = blurry).
intensity_entropy	Shannon entropy of the intensity distribution.

Shape (mask required):

Column	Description
area	Foreground pixel count.
major_axis_length	Length of the major axis of the fitted ellipse.
minor_axis_length	Length of the minor axis of the fitted ellipse.
perimeter	Boundary length of the object.
eccentricity	Elongation of the fitted ellipse (0 = circle, 1 = line).
solidity	Ratio of area to convex hull area (1 = fully convex).
orientation	Angle of the major axis in radians.
extent	Ratio of object area to bounding-box area.
equivalent_diameter	Diameter of a circle with the same area as the object.
euler_number	Number of objects minus number of holes (topology).
circularity	4π · area / perimeter² (1 = perfect circle).

Masked intensity + radial profile (image + mask required):

Column	Description
intensity_mean_masked	Mean intensity inside the segmented object.
intensity_std_masked	Intensity standard deviation inside the object.
intensity_min_masked	Minimum intensity inside the object.
intensity_max_masked	Maximum intensity inside the object.
ring_1_mean	Mean intensity in the outermost ring (membrane region).
ring_1_std	Intensity std in the outermost ring.
ring_2_mean	Mean intensity in the middle ring (cytoplasm region).
ring_2_std	Intensity std in the middle ring.
ring_3_mean	Mean intensity in the innermost ring (core/nucleus region).
ring_3_std	Intensity std in the innermost ring.
texture_contrast	GLCM contrast (local intensity variation).
texture_dissimilarity	GLCM dissimilarity (weighted distance between intensity pairs).
texture_homogeneity	GLCM homogeneity (closeness of distribution to diagonal).
texture_energy	GLCM energy (uniformity of texture).
texture_correlation	GLCM correlation (linear dependency of grey levels).

Note

For multi-channel images, all per-channel properties are suffixed with _ch0, _ch1, etc. Pass channel_names=["DAPI", "GFP"] to find_files (or set_file_df) to use those labels instead (e.g. intensity_mean_DAPI). Metadata and mask-derived shape columns are not duplicated per channel.

Usage

file_df = pheno.find_files("path/to/images")
pheno.process_images(wrapper)

# Use a preset
df = pheno.compute_properties(property_preset="basic")

# Custom-only (no preset)
df = pheno.compute_properties(
    property_preset="none",
    additional_property_functions={"image": [my_intensity_fn]},
)

# Extend a preset with extra functions
df = pheno.compute_properties(
    property_preset="standard",
    additional_property_functions={
        "both": [create_concentric_ring_function(5, ["mean"])]
    },
)

Extending a preset

Inspect the functions a preset would produce, then add to them:

from phenome import (
    get_preset_property_functions,
    create_regionprops_function,
    create_concentric_ring_function,
)

# Inspect what "basic" includes
print(get_preset_property_functions("basic"))
# {'image': [...], 'mask': [...], 'both': [...]}

# Add a custom shape property on top of "basic"
extra_shape = create_regionprops_function(["aspect_ratio"])

df = pheno.compute_properties(
    property_preset="basic",
    additional_property_functions={"mask": [extra_shape]},
)

Built-in helper factories

These factories create property functions that can be passed directly to additional_property_functions. All presets are built from them.

Factory	Output columns
create_regionprops_function(names)	One column per name in names. Available: area, perimeter, major_axis_length, minor_axis_length, equivalent_diameter, eccentricity, solidity, orientation, euler_number, extent, circularity, roundness, aspect_ratio.
create_intensity_function(stat, fn)	intensity_{stat} — whole-image statistic.
create_masked_intensity_function(stat, fn)	intensity_{stat}_masked — statistic inside the object.
create_texture_function()	texture_contrast, texture_dissimilarity, texture_homogeneity, texture_energy, texture_correlation.
create_blur_effect_function()	sharpness_metric — focus quality (0 = sharp, 1 = blurry).
create_entropy_function()	intensity_entropy — Shannon entropy of the intensity distribution.
create_concentric_ring_function(N, stats)	ring_1_{stat} … ring_N_{stat} — radial intensity profile from boundary to core.
phenome.plugins.get_blob_properties	Blob/cell-level properties from a labelled mask.

Pass the returned callable via additional_property_functions, for example:

{"mask": [create_regionprops_function(["area", "circularity"])]}

See also

• Writing property functions — define your own.
• Advanced recipes — external libraries, GLCM, multi-channel.
• Property interpretation — what each built-in property measures biologically.