28 lines (25 with data), 2.3 kB

3

mΏ€a{γ@s*dd
lZ
dd
lZddlmZ
ddd„
Zd
S)ιN)
Ϊ__surface_distancesι
cCsR
td
d„t
|j|
jƒDƒ
ƒ
s8tdt|jƒ
t|
jƒ
fƒ
‚
d||dk<d||dk<d|
|
dk<d|
|
dk<|jtjƒ
}|jj	ƒj
ƒ}|
jj	ƒj
ƒ}
|dkr΄tdd„tt
|jƒ
ƒ
Dƒ
ƒ
}t||
||ƒ}t|
|||ƒ}t
|ƒ
}t
|ƒ
}	tj||k
ƒ
|}
tj||k
ƒ
|	}tj||kƒ
|}tj||kƒ
|	}
|
||
|||
d	}|j|ƒ

|S)
aυ
    This implementation differs from the official surface dice implementation! These two are not comparable!!!!!

    The normalized surface dice is symmetric, so it should not matter whether a or b is the reference image

    This implementation natively supports 2D and 3D images. Whether other dimensions are supported depends on the
    __surface_distances implementation in medpy

    :param a: image 1, must have the same shape as b
    :param b: image 2, must have the same shape as a
    :param threshold: distances below this threshold will be counted as true positives. Threshold is in mm, not voxels!
    (if spacing = (1, 1(, 1)) then one voxel=1mm so the threshold is effectively in voxels)
    must be a tuple of len dimension(a)
    :param spacing: how many mm is one voxel in reality? Can be left at None, we then assume an isotropic spacing of 1mm
    :param connectivity: see scipy.ndimage.generate_binary_structure for more information. I suggest you leave that
    one alone
    :return:
    c
Ssg|]\}
}|
|k‘qS©r)Ϊ.0Ϊ
iΪ
jrrϊ+/home/ws/yanghan/Newcode/metrics/meanNSD.pyϊ
<listcomp>szmeanNSD.<locals>.<listcomp>z:a and b must have the same shape. a.shape= %s, b.shape= %sr
gΰ?rNc
Ssg|]}
d‘qS)
rr)rΪ
_rrrr	$sg:Œ0βŽyE>)ΪallΪzipΪshapeΪAssertionErrorΪstrΪtypeΪtorchΪ
LongTensorΪdataΪcpuΪnumpyΪtupleΪrangeΪlenrΪnpΪsumΪappend)Ϊ
aΪ
bZnsdsΪ	thresholdΪspacingZconnectivityZa_to_bZb_to_aZnumel_aZnumel_bZtp_aZtp_bΪfpΪfnΪdcrrrΪmeanNSDs, 














r#)rNr)rrrZmedpy.metric.binaryrr#rrrrΪ<module>
s