Devices, systems, and methods for improving image consistency
View Patent ↗Certain exemplary embodiments can comprise a method, which can comprise automatically rendering an improved image of a target object. The improved image obtained based upon a principal mode of the target object. The principal mode of the target object can be provided to an algorithm that is adapted to derive the improved image of the target object.
1. A method comprising:
obtaining data from a computed tomography device;
receiving a user selection of a set of voxels of a target object comprised by said initial image;
deriving an initial segmentation of said target object, said initial segmentation based on a prominent intensity of said target object;
rendering an initial image based upon said initial segmentation;
determining a principal mode from a prior segmentation via a mean shift algorithm that comprises an evaluation of a determined voxel intensity change within said initial segmentation, wherein said mean shift algorithm comprises an evaluation of an equation:
(
x
i
+
1
,
y
i
+
1
,
z
i
+
1
)
=
∑
(
u
,
v
,
w
)
∈
Si
k
(
(
u
-
x
i
)
,
(
v
-
y
i
)
,
(
w
-
z
i
)
2
)
ω
(
u
,
v
,
w
)
(
u
,
v
,
w
)
∑
(
u
,
v
,
w
)
∈
Si
k
(
(
u
-
x
i
)
,
(
v
-
y
i
)
,
(
w
-
z
i
)
2
)
ω
(
u
,
v
,
w
)
where
(x i ,y i ,z i ), i=0, 1, 2, . . . , is a sequence of positions starting from a given position (x 0 ,y 0 ,z 0 )
Si is said initial segmentation of said object;
k is a kernel function; and
ω is a weight function;
providing said principal mode to an algorithm that is adapted to derive an improved image of said target object; and
rendering said improved image of said target object.
2. The method of claim 1 , wherein the step of deriving an improved image of a target object further comprises:
automatically determining an improved image of a target object, said improved image determined based upon a principal mode of said target object, said principal mode of said target object provided to an algorithm that is adapted to derive said improved image of said target object, said principal mode determined from data obtained from a computed tomography device, said principal mode determined via an application of a mean shift algorithm that comprises an evaluation of a determined voxel intensity change within an initial segmentation of said target object.
3. The method of claim 1 , further comprising:
obtaining said data from said computed tomography device.
4. The method of claim 1 , further comprising:
receiving a user selection of a set of voxels from said target object.
5. The method of claim 1 , further comprising:
receiving a user identification of said target object.
6. The method of claim 1 , further comprising:
determining said initial segmentation of said target object.
7. The method of claim 1 , further comprising:
determining said initial segmentation of said object based upon a user identification of said data from said computed tomography device.
8. The method of claim 1 , further comprising:
flipping a segmentation of one or more background structures to obtain said initial segmentation of said target object.
9. The method of claim 1 , further comprising:
deriving said initial segmentation of said object from an intensity distribution from a neighborhood around a predetermined prominent feature of said data obtained from said computed tomography device.
10. The method of claim 1 , further comprising:
deriving said initial segmentation via a region growing algorithm.
11. The method of claim 1 , further comprising:
deriving said initial segmentation via a watershed algorithm.
12. The method of claim 1 , further comprising:
deriving said initial segmentation via a level set algorithm.
13. The method of claim 1 , further comprising:
deriving said initial segmentation via a graph cut algorithm.
14. The method of claim 1 , wherein said weight function is a Gaussian weight function.
15. The method of claim 1 , wherein said weight function is a triangle weight function.
16. The method of claim 1 , wherein said weight function is a cosine weight function.
17. The method of claim 1 , wherein said weight function is a step weight function.
18. A computer-readable medium embodying computer executable instructions for activities comprising:
automatically determining an improved image of a target object, said improved image determined based upon a principal mode of said target object, said principal mode of said target object provided to an algorithm that is adapted to derive said improved image of said target object, said principal mode determined from data obtained from a computed tomography device, said principal mode determined via an application of a mean shift algorithm that comprises an evaluation of a determined voxel intensity change within an initial segmentation of said target object, wherein said mean shift algorithm comprises an evaluation of an equation:
(
x
i
+
1
,
y
i
+
1
,
z
i
+
1
)
=
∑
(
u
,
v
,
w
)
∈
Si
k
(
(
u
-
x
i
)
,
(
v
-
y
i
)
,
(
w
-
z
i
)
2
)
ω
(
u
,
v
,
w
)
(
u
,
v
,
w
)
∑
(
u
,
v
,
w
)
∈
Si
k
(
(
u
-
x
i
)
,
(
v
-
y
i
)
,
(
w
-
z
i
)
2
)
ω
(
u
,
v
,
w
)
where
(x i ,y i ,z i ), i=0, 1, 2, . . . , is a sequence of positions starting from a given position (x 0 ,y 0 ,z 0 )
Si is said initial segmentation of said object;
k is a kernel function; and
ω is a weight function.
19. A system comprising:
a processing means for determining an improved image of a target object, said improved image determined based upon a principal mode of said target object, said principal mode of said target object provided to an algorithm that is adapted to derive said improved image of said target object, said principal mode determined from data obtained from a computed tomography device, said principal mode determined via an application of a mean shift algorithm that comprises an evaluation of a determined voxel intensity change within an initial segmentation of said target object, wherein said mean shift algorithm comprises an evaluation of an equation:
(
x
i
+
1
,
y
i
+
1
,
z
i
+
1
)
=
∑
(
u
,
v
,
w
)
∈
Si
k
(
(
u
-
x
i
)
,
(
v
-
y
i
)
,
(
w
-
z
i
)
2
)
ω
(
u
,
v
,
w
)
(
u
,
v
,
w
)
∑
(
u
,
v
,
w
)
∈
Si
k
(
(
u
-
x
i
)
,
(
v
-
y
i
)
,
(
w
-
z
i
)
2
)
ω
(
u
,
v
,
w
)
where
(x i ,y i ,z i ), i=0, 1, 2, . . . , is a sequence of positions starting from a given position (x 0 ,y 0 ,z 0 )
Si is said initial segmentation of said object;
k is a kernel function; and
ω is a weight function; and
a user interface adapted to render said improved image.