o
i- @ sr d dl mZ d dlmZmZ d dlZd dlZd dlm Z d&d'd
dZ
d(d)ddZ d*dd Z d+d,d$d%Z
dS )- ) annotations)AnycastN)NdarrayOrTensorprobsr coordsevaluation_masklabels_to_exclude list | Nonereturn,tuple[NdarrayOrTensor, NdarrayOrTensor, int]c
C s` t | t |kstd| j d|j dt |jdkr&|jd t |jks3td|j d|j dt| tjrA| } t|tjrO| }t|tjr]| }|du rcg }t
|}t j |ft j d}|t
|j }| t |d k }td|d D ]}||vr||v r| t ||k
||d < q|t | } ||tt| fS )
ar
This function is modified from the official evaluation code of
`CAMELYON 16 Challenge `_, and used to distinguish
true positive and false positive predictions. A true positive prediction is defined when
the detection point is within the annotated ground truth region.
Args:
probs: an array with shape (n,) that represents the probabilities of the detections.
Where, n is the number of predicted detections.
coords: an array with shape (n, n_dim) that represents the coordinates of the detections.
The dimensions must be in the same order as in `evaluation_mask`.
evaluation_mask: the ground truth mask for evaluation.
labels_to_exclude: labels in this list will not be counted for metric calculation.
Returns:
fp_probs: an array that contains the probabilities of the false positive detections.
tp_probs: an array that contains the probabilities of the True positive detections.
num_targets: the total number of targets (excluding `labels_to_exclude`) for all images under evaluation.
zthe length of probs z", should be the same as of coords . zcoords z9 need to represent the same number of dimensions as mask N)dtyper )len
ValueErrorshape
isinstancetorchTensordetachcpunumpynpmaxzerosfloat32tupleTwhereranger int)
r r r r max_labeltp_probsZ hittedlabelfp_probsi num_targets r' T/home/dell461/cl/sdc2/last_ska_mid/HISourceFinder-master-l/src/monai/metrics/froc.pycompute_fp_tp_probs_nd s0 "
r) y_coordx_coordresolution_levelr! c C s t |tjr| }t |tjr | }|td| t}|td| t}t j
||gdd}t | |||dS )a
This function is modified from the official evaluation code of
`CAMELYON 16 Challenge `_, and used to distinguish
true positive and false positive predictions. A true positive prediction is defined when
the detection point is within the annotated ground truth region.
Args:
probs: an array with shape (n,) that represents the probabilities of the detections.
Where, n is the number of predicted detections.
y_coord: an array with shape (n,) that represents the Y-coordinates of the detections.
x_coord: an array with shape (n,) that represents the X-coordinates of the detections.
evaluation_mask: the ground truth mask for evaluation.
labels_to_exclude: labels in this list will not be counted for metric calculation.
resolution_level: the level at which the evaluation mask is made.
Returns:
fp_probs: an array that contains the probabilities of the false positive detections.
tp_probs: an array that contains the probabilities of the True positive detections.
num_targets: the total number of targets (excluding `labels_to_exclude`) for all images under evaluation.
r )axis)r r r r ) r r r r r r powastyper! r stackr) )r r* r+ r r r, stackedr' r' r( compute_fp_tp_probsN s r3 r$ np.ndarray | torch.Tensorr# r&
num_images tuple[np.ndarray, np.ndarray]c
C s t | t|s tdt | tjr| } t |tjr'| }g g }}tt t
| t
| }|dd D ]}| | |k | ||k q>| d | d t
|t| }t
|t| } || fS )a
This function is modified from the official evaluation code of
`CAMELYON 16 Challenge `_, and used to compute
the required data for plotting the Free Response Operating Characteristic (FROC) curve.
Args:
fp_probs: an array that contains the probabilities of the false positive detections for all
images under evaluation.
tp_probs: an array that contains the probabilities of the True positive detections for all
images under evaluation.
num_targets: the total number of targets (excluding `labels_to_exclude`) for all images under evaluation.
num_images: the number of images under evaluation.
z&fp and tp probs should have same type.r Nr )r typeAssertionErrorr r r r r sortedsetlistappendsumr asarrayfloat)
r$ r# r& r5 Z total_fpsZ total_tpsZ all_probsthresh
fps_per_imagetotal_sensitivityr' r' r( compute_froc_curve_dataz s
rC g ?g ?r r- rA
np.ndarrayrB eval_thresholdsr r c C s, t || ddd |ddd }t |S )a
This function is modified from the official evaluation code of
`CAMELYON 16 Challenge `_, and used to compute
the challenge's second evaluation metric, which is defined as the average sensitivity at
the predefined false positive rates per whole slide image.
Args:
fps_per_image: the average number of false positives per image for different thresholds.
total_sensitivity: sensitivities (true positive rates) for different thresholds.
eval_thresholds: the false positive rates for calculating the average sensitivity. Defaults
to (0.25, 0.5, 1, 2, 4, 8) which is the same as the CAMELYON 16 Challenge.
N)r interpmean)rA rB rH Z interp_sensr' r' r( compute_froc_score s "
rL )N)
r r r r r r r r
r r )Nr )r r r* r r+ r r r r r
r, r! r r )
r$ r4 r# r4 r& r! r5 r! r r6 )rD )rA rG rB rG rH r r r )
__future__r typingr r r r r monai.configr r) r3 rC rL r' r' r' r( s =
,%