niftynet.evaluation.pairwise_measures module

class PairwiseMeasures(seg_img, ref_img, measures=None, num_neighbors=8, pixdim=(1, 1, 1), empty=False, list_labels=None)

Bases: object

check_binary()

Checks whether self.seg and self.ref are binary. This is to enable measurements such as ‘false positives’, which only have meaning in the binary case (what is positive/negative for multiple class?)

n_pos_ref

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

n_neg_ref

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

n_pos_seg

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

n_neg_seg

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

fp

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

fn

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

tp

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

tn

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

n_intersection

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

n_union

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

sensitivity()

specificity()

accuracy()

false_positive_rate()

positive_predictive_values()

negative_predictive_values()

This function calculates the negative predictive value ratio between the number of true negatives and the total number of negative elements

Returns:

dice_score()

This function returns the dice score coefficient between a reference and segmentation images

Returns:dice score

intersection_over_union()

This function the intersection over union ratio - Definition of jaccard coefficient

Returns:

jaccard()

This function returns the jaccard coefficient (defined as intersection over union)

Returns:jaccard coefficient

informedness()

This function calculates the informedness between the segmentation and the reference

Returns:informedness

markedness()

This functions calculates the markedness :return:

com_dist()

This function calculates the euclidean distance between the centres of mass of the reference and segmentation.

Returns:

com_ref()

This function calculates the centre of mass of the reference segmentation

Returns:

com_seg()

This functions provides the centre of mass of the segmented element

Returns:

list_labels()

vol_diff()

This function calculates the ratio of difference in volume between the reference and segmentation images.

Returns:vol_diff

border_distance

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

measured_distance()

This functions calculates the average symmetric distance and the hausdorff distance between a segmentation and a reference image

Returns:hausdorff distance and average symmetric distance

measured_average_distance()

This function returns only the average distance when calculating the distances between segmentation and reference

Returns:

measured_hausdorff_distance()

This function returns only the hausdorff distance when calculated the distances between segmentation and reference

Returns:

connected_elements()

This function returns the number of FP FN and TP in terms of connected components.

Returns:Number of true positive connected components, Number of false positives connected components, Number of false negatives connected components

connected_errormaps

this provides a decorator to cache function outputs to avoid repeating some heavy function computations

outline_error()

This function calculates the outline error as defined in Wack et al.

Returns:OER: Outline error ratio, OEFP: number of false positive outlier error voxels, OEFN: number of false negative outline error elements

detection_error()

This function calculates the volume of detection error as defined in Wack et al.

Returns:DE: Total volume of detection error, DEFP: Detection error false positives, DEFN: Detection error false negatives

header_str()

to_string(fmt='{:.4f}')

class PairwiseMeasuresRegression(reg_img, ref_img, measures=None)

Bases: object

mse()

rmse()

mae()

r2()

header_str()

to_string(fmt='{:.4f}')