Svm features¶

Prerequisites: Valid cell geometry and inverse-phase signal; these quantities are used by the curation classifier.

A failed cell calculation omits that family row and records the failure. Missing values in a merged table are not automatically zero. Numerical NaN can also arise within a returned measurement. The object retention option explicitly sets invalid mesh-dependent aggregates to NaN.

contour¶

Exported column: <channel>_contour

Meaning: Stored cell contour coordinates.

Calculation: Store the cell contour coordinates directly in the table. Explanation: This is mainly kept for tracking, visualization, and linking feature rows back to the original segmented cell.

Units: pixels. Value shape: N × 2 geometry array in pixel coordinates.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

cell.contour

Source assignment, line 394; bactoscoop.features.Features.svm().

cell_length¶

Exported column: <channel>_cell_length

Meaning: Cell length along the mesh midline.

Calculation: Sum the step lengths between successive mesh midpoints along the cell axis. Explanation: This is the measured pole-to-pole length of the cell, following the fitted midline rather than a simple straight line.

Units: µm. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_length(step_length)

Calculation helpers: bactoscoop.utilities.get_length().

Source assignment, line 397; bactoscoop.features.Features.svm().

cell_width¶

Exported column: <channel>_cell_width

Meaning: Cell body width.

Calculation: Compute the width of every mesh segment as the distance between (x1, y1) and (x2, y2), sort those widths, and take the mean of the widest one-third. Explanation: This gives a robust estimate of the cell body width while reducing the influence of the narrow poles.

Units: µm. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_avg_width(cell.x1, cell.y1, cell.x2, cell.y2, px)

Calculation helpers: bactoscoop.utilities.get_avg_width().

Source assignment, line 398; bactoscoop.features.Features.svm().

cell_area¶

Exported column: <channel>_cell_area

Meaning: Cell projected area.

Calculation: Compute the area of the polygon defined by the cell contour. Explanation: This is the 2D projected size of the segmented cell in the image.

Units: µm². Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_area(cell.contour, px)

Calculation helpers: bactoscoop.utilities.get_area().

Source assignment, line 401; bactoscoop.features.Features.svm().

cell_volume¶

Exported column: <channel>_cell_volume

Meaning: Cell estimated volume.

Calculation: Treat each mesh segment as a cylinder with radius width/2, integrate pi*(width/2)^2 along the cell length using the trapezoidal rule, and sum the segments. Explanation: This estimates the 3D volume of the bacterium from its 2D mesh.

Units: µm³. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_volume(cell.x1, cell.y1, cell.x2, cell.y2, step_length, px)

Calculation helpers: bactoscoop.utilities.get_volume().

Source assignment, line 402; bactoscoop.features.Features.svm().

cell_surface_area¶

Exported column: <channel>_cell_surface_area

Meaning: Cell estimated surface area.

Calculation: For each mesh segment compute 2*pi*(width/2)*step_length and sum the segment values. Explanation: This estimates the outer surface area of the cell envelope along the reconstructed cell shape.

Units: µm². Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_surface_area(width_not_ordered, step_length)

Calculation helpers: bactoscoop.utilities.get_surface_area().

Source assignment, line 405; bactoscoop.features.Features.svm().

max_curvature¶

Exported column: <channel>_max_curvature

Meaning: Maximum cell contour curvature.

Calculation: Compute curvature as |x’y’’ - y’x’’| / (x’^2 + y’^2)^(3/2) along the relevant contour or midline, then report the requested summary statistic. Explanation: Curvature describes how strongly the line bends; higher values indicate sharper turns or tighter bends.

Units: µm⁻¹. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_curvature_characteristics(cell.contour, px)

Calculation helpers: bactoscoop.utilities.get_curvature_characteristics().

Source assignment, line 408; bactoscoop.features.Features.svm().

min_curvature¶

Exported column: <channel>_min_curvature

Meaning: Minimum cell contour curvature.

Units: µm⁻¹. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_curvature_characteristics(cell.contour, px)

Calculation helpers: bactoscoop.utilities.get_curvature_characteristics().

Source assignment, line 408; bactoscoop.features.Features.svm().

mean_curvature¶

Exported column: <channel>_mean_curvature

Meaning: Mean cell contour curvature.

Units: µm⁻¹. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_curvature_characteristics(cell.contour, px)

Calculation helpers: bactoscoop.utilities.get_curvature_characteristics().

Source assignment, line 408; bactoscoop.features.Features.svm().

cell_perimeter¶

Exported column: <channel>_cell_perimeter

Meaning: Cell contour perimeter.

Calculation: Sum the distances between consecutive points of the subpixel contour. Explanation: This is the total contour length around the cell boundary in 2D.

Units: µm. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_cell_perimeter_measurements(cell.contour, cell.svm_features['cell_area'], px)

Calculation helpers: bactoscoop.utilities.get_cell_perimeter_measurements().

Source assignment, line 415; bactoscoop.features.Features.svm().

cell_circularity¶

Exported column: <channel>_cell_circularity

Meaning: Cell circularity.

Calculation: (4*pi*cell_area) / (cell_perimeter^2). Explanation: Values closer to 1 indicate a more circle-like outline, while lower values indicate a more elongated or irregular shape.

Units: Dimensionless (GLCM descriptors use quantized intensity bins). Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_cell_perimeter_measurements(cell.contour, cell.svm_features['cell_area'], px)

Calculation helpers: bactoscoop.utilities.get_cell_perimeter_measurements().

Source assignment, line 415; bactoscoop.features.Features.svm().

cell_POA¶

Exported column: <channel>_cell_POA

Meaning: Cell compactness.

Calculation: (cell_perimeter^2) / cell_area. Explanation: This describes how much boundary length is needed to enclose the area; larger values reflect less compact shapes.

Units: Dimensionless. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_cell_perimeter_measurements(cell.contour, cell.svm_features['cell_area'], px)

Calculation helpers: bactoscoop.utilities.get_cell_perimeter_measurements().

Source assignment, line 415; bactoscoop.features.Features.svm().

cell_sphericity¶

Exported column: <channel>_cell_sphericity

Meaning: Cell sphericity descriptor.

Calculation: (1.5*pi*(cell_perimeter/(2*pi))^1.5) / cell_area. Explanation: This is a 2D shape descriptor intended to capture how close the outline is to a compact rounded object. The implementation uses a 2D perimeter/area formula and carries µm^(-1/2); it is not standard dimensionless 3D sphericity.

Units: µm⁻¹ᐟ² under the implemented 2D formula. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_cell_perimeter_measurements(cell.contour, cell.svm_features['cell_area'], px)

Calculation helpers: bactoscoop.utilities.get_cell_perimeter_measurements().

Source assignment, line 415; bactoscoop.features.Features.svm().

phaco_total_intensity¶

Exported column: <channel>_phaco_total_intensity

Meaning: Total phase-contrast intensity inside the cell.

Calculation: Measure the requested statistic from the inverted, background-subtracted phase-contrast signal inside the cell, along the contour, or along expanded/eroded contour versions depending on the feature. Explanation: These features describe how the phase-contrast signal is distributed around and within the segmented cell.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

total_int

Source assignment, line 428; bactoscoop.features.Features.svm().

phaco_max_intensity¶

Exported column: <channel>_phaco_max_intensity

Meaning: Maximum phase-contrast intensity inside the cell.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

max_int

Source assignment, line 429; bactoscoop.features.Features.svm().

phaco_mean_intensity¶

Exported column: <channel>_phaco_mean_intensity

Meaning: Mean phase-contrast intensity inside the cell.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

mean_int

Source assignment, line 430; bactoscoop.features.Features.svm().

phaco_contour_peaks¶

Exported column: <channel>_phaco_contour_peaks

Meaning: Number of peaks in the phase-contrast contour profile.

Units: count. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.find_signal_peaks(phaco_contour_intensity, max_int)

Calculation helpers: bactoscoop.utilities.find_signal_peaks().

Source assignment, line 441; bactoscoop.features.Features.svm().

phaco_max_contour_intensity¶

Exported column: <channel>_phaco_max_contour_intensity

Meaning: Maximum phase-contrast intensity along the contour.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.max(phaco_contour_intensity)

Source assignment, line 444; bactoscoop.features.Features.svm().

phaco_mean_contour_intensity¶

Exported column: <channel>_phaco_mean_contour_intensity

Meaning: Mean phase-contrast intensity along the contour.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.mean(phaco_contour_intensity)

Source assignment, line 447; bactoscoop.features.Features.svm().

phaco_min_contour_intensity¶

Exported column: <channel>_phaco_min_contour_intensity

Meaning: Mean of the 10 lowest phase-contrast contour intensities.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.mean(sorted_phaco_contour_intensity[:10])

Source assignment, line 450; bactoscoop.features.Features.svm().

phaco_max_contour_variability¶

Exported column: <channel>_phaco_max_contour_variability

Meaning: Maximum local variability of phase-contrast contour intensity.

Units: A.U.. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.max(phaco_contour_intensity_variability)

Source assignment, line 453; bactoscoop.features.Features.svm().

phaco_mean_contour_variability¶

Exported column: <channel>_phaco_mean_contour_variability

Meaning: Mean local variability of phase-contrast contour intensity.

Units: A.U.. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.mean(phaco_contour_intensity_variability)

Source assignment, line 456; bactoscoop.features.Features.svm().

midline_kurtosis¶

Exported column: <channel>_midline_kurtosis

Meaning: Kurtosis of the phase-contrast midline profile.

Calculation: Compute the requested distribution statistic from the phase-contrast axial intensity profile along the midline. Explanation: These values describe whether the midline intensity profile is symmetric, heavy-tailed, or dominated by peaks.

Units: Dimensionless (GLCM descriptors use quantized intensity bins). Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_kurtosis(cell.profiling_data['phaco_axial_intensity'])

Calculation helpers: bactoscoop.utilities.get_kurtosis().

Source assignment, line 468; bactoscoop.features.Features.svm().

midline_skewness¶

Exported column: <channel>_midline_skewness

Meaning: Skewness of the phase-contrast midline profile.

Units: Dimensionless (GLCM descriptors use quantized intensity bins). Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

u.get_skew(cell.profiling_data['phaco_axial_intensity'])

Calculation helpers: bactoscoop.utilities.get_skew().

Source assignment, line 471; bactoscoop.features.Features.svm().

phaco_max_expanded_contour_intensity¶

Exported column: <channel>_phaco_max_expanded_contour_intensity

Meaning: Maximum phase-contrast intensity along the expanded contour.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.max(cell.profiling_data['phaco_expanded_contour_intensity'])

Source assignment, line 485; bactoscoop.features.Features.svm().

phaco_mean_expanded_contour_intensity¶

Exported column: <channel>_phaco_mean_expanded_contour_intensity

Meaning: Mean phase-contrast intensity along the expanded contour.

Units: Image intensity units after the family-specific preprocessing. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.mean(cell.profiling_data['phaco_expanded_contour_intensity'])

Source assignment, line 488; bactoscoop.features.Features.svm().

phaco_cell_edge_gradient¶

Exported column: <channel>_phaco_cell_edge_gradient

Meaning: Mean phase-contrast edge gradient.

Units: A.U.. Value shape: Scalar.

Availability: emitted whenever this family calculation succeeds; all_data does not gate this key.

Normalization: No additional export-level normalization; the calculation below determines family-specific preprocessing.

Current assignment expression (a tuple expression can produce several named outputs):

np.average(cell.profiling_data['phaco_eroded_contour_intensity'] - cell.profiling_data['phaco_expanded_contour_intensity'])

Source assignment, line 493; bactoscoop.features.Features.svm().