Metric measures (`medpy.metric`)#

This package provides a number of metric measures that e.g. can be used for testing and/or evaluation purposes on two binary masks (i.e. measuring their similarity) or distance between histograms.

Binary metrics (`medpy.metric.binary`)#

Metrics to compare binary objects and classification results.

Compare two binary objects#

`dc`(result, reference)	Dice coefficient
`jc`(result, reference)	Jaccard coefficient
`hd`(result, reference[, voxelspacing, ...])	Hausdorff Distance.
`asd`(result, reference[, voxelspacing, ...])	Average surface distance metric.
`assd`(result, reference[, voxelspacing, ...])	Average symmetric surface distance.
`precision`(result, reference)	Precison.
`recall`(result, reference)	Recall.
`sensitivity`(result, reference)	Sensitivity.
`specificity`(result, reference)	Specificity.
`true_positive_rate`(result, reference)	True positive rate.
`true_negative_rate`(result, reference)	True negative rate.
`positive_predictive_value`(result, reference)	Positive predictive value.
`ravd`(result, reference)	Relative absolute volume difference.

Compare two sets of binary objects#

`obj_tpr`(result, reference[, connectivity])	The true positive rate of distinct binary object detection.
`obj_fpr`(result, reference[, connectivity])	The false positive rate of distinct binary object detection.
`obj_asd`(result, reference[, voxelspacing, ...])	Average surface distance between objects.
`obj_assd`(result, reference[, voxelspacing, ...])	Average symmetric surface distance.

Compare to sequences of binary objects#

`volume_correlation`(results, references)	Volume correlation.
`volume_change_correlation`(results, references)	Volume change correlation.

Image metrics (`medpy.metric.image`)#

Some more image metrics (e.g. sls and ssd) can be found in medpy.filter.image.

mutual_information(i1, i2[, bins])

Computes the mutual information (MI) (a measure of entropy) between two images.

Histogram metrics (`medpy.metric.histogram`)#

`chebyshev`(h1, h2)	Chebyshev distance.
`chebyshev_neg`(h1, h2)	Chebyshev negative distance.
`chi_square`(h1, h2)	Chi-square distance.
`correlate`(h1, h2)	Correlation between two histograms.
`correlate_1`(h1, h2)	Correlation distance.
`cosine`(h1, h2)	Cosine simmilarity.
`cosine_1`(h1, h2)	Cosine simmilarity.
`cosine_2`(h1, h2)	Cosine simmilarity.
`cosine_alt`(h1, h2)	Alternative implementation of the `cosine` distance measure.
`euclidean`(h1, h2)	Equal to Minowski distance with \(p=2\).
`fidelity_based`(h1, h2)	Fidelity based distance.
`histogram_intersection`(h1, h2)	Calculate the common part of two histograms.
`histogram_intersection_1`(h1, h2)	Turns the histogram intersection similarity into a distance measure for normalized, positive histograms.
`jensen_shannon`(h1, h2)	Jensen-Shannon divergence.
`kullback_leibler`(h1, h2)	Kullback-Leibler divergence.
`manhattan`(h1, h2)	Equal to Minowski distance with \(p=1\).
`minowski`(h1, h2[, p])	Minowski distance.
`noelle_1`(h1, h2)	Extension of `fidelity_based` proposed by [R0007bcd883ab-1].
`noelle_2`(h1, h2)	Extension of `fidelity_based` proposed by [R256fa0dd8776-1].
`noelle_3`(h1, h2)	Extension of `fidelity_based` proposed by [Rd5b1e2c5cbe2-1].
`noelle_4`(h1, h2)	Extension of `fidelity_based` proposed by [Rcc5be5568670-1].
`noelle_5`(h1, h2)	Extension of `fidelity_based` proposed by [R3c080ba261ce-1].
`quadratic_forms`(h1, h2)	Quadrativ forms metric.
`relative_bin_deviation`(h1, h2)	Calculate the bin-wise deviation between two histograms.
`relative_deviation`(h1, h2)	Calculate the deviation between two histograms.