3D ball skinning using partial differential equations for generation of smooth tubular surfaces
View Patent ↗A method of computing a continuous interpolation of a discrete set of three-dimensional (3D) balls, including generating an initial skin, wherein the initial skin is a surface comprised of splines and wherein the splines touch each ball along a circle that is tangent to the ball, solving a first differential equation to minimize the initial skin's surface area or solving a second differential equation to minimize a squared mean curvature of the initial skin's surface, wherein the result of solving the first or second differential equations is an updated skin; and repeating the steps of solving the first or second differential equations for the updated skin, and then, repeating the steps of solving the first or second differential equations for each subsequently updated skin until a desired skin is realized.
1. A method of computing a continuous interpolation of a discrete set of three-dimensional (3D) balls, comprising:
generating an initial skin using a processor of a computer, wherein the initial skin is generated by using machine code excutable by the processor to form a surface of the initial skin comprised of splines, wherein the splines touch each ball along a circle that is tangent to the ball;
solving a first differential equation to minimize the initial skin's surface area or solving a second differential equation to minimize a squared mean curvature of the initial skin's surface,
wherein the result of solving the first or second differential equations is an updated skin; and
repeating the steps of solving the first or second differential equations for the updated skin, and then, repeating the steps of solving the first or second differential equations for each subsequently updated skin until a desired skin is realized,
where a circle on a ball B i is in a plane with mormal N i =[cos θ i sin φ i , sin φ i sin φ i , cos φ i ] T that passes through the ball center c i and intersects the ball B i , q(u) is a point on the circle and is defined by q(u)={circumflex over (x)} cos u+ŷ sin u, and a point p i (u) on a circle through which spline passes is defined by p i (u)=c i +r i q(u), where r i is a radius of an i th ball, and R i is a rotation matrix specified by N i .
2. The method of claim 1 , wherein the rotation matrix R i is
R
i
=
R
(
θ
i
,
ϕ
i
)
=
[
cos
θ
i
cos
ϕ
i
-
sin
θ
i
cos
θ
i
sin
ϕ
i
sin
θ
i
cos
ϕ
i
cos
θ
i
sin
θ
i
sin
ϕ
i
-
sin
ϕ
i
0
cos
ϕ
i
]
.
3. The method of claim 1 , wherein the initial skin is represented by S(u, v) , where S(u, v) is a collection of continuous segments S i (u, v), where an S i (u, v) is specified by A i (u),B i (u),C i (u),D i (u), where A i (u)=−2p i+1 (u)+2p i (u)+t i N i +t i+1 N i+1 , B i (u)=3p i+1 (u)−3p i (u)−2t i N i −t i+1 N i+1 , C i (u)=t i N i , and D i (u)=p i (u).
4. The method of claim 3 , wherein the first differential equation is represented by
∂
J
a
∂
w
k
=
1
2
∫
∫
∂
E
i
∂
w
k
G
i
+
E
i
∂
G
i
∂
w
k
-
2
F
i
∂
F
i
∂
w
k
[
E
i
G
i
-
(
F
i
)
2
]
1
2
ⅆ
u
ⅆ
v
+
1
2
∫
∫
∂
E
i
-
1
∂
w
k
G
i
-
1
+
E
i
-
1
∂
G
i
-
1
∂
w
k
-
2
F
i
-
1
∂
F
i
-
1
∂
w
k
[
E
i
-
1
G
i
-
1
-
(
F
i
-
1
)
2
]
1
2
ⅆ
u
ⅆ
v
.
5. The method of claim 4 , where J a =∫∫√{square root over (EG−F 2 )}dudv, w k ε[θ i ,φ i ]∀i, E=S u ·S u , F=S u ·S v , and G=S v ·S v .
6. The method of claim 3 , wherein the second differential equation is represented by
∂
J
c
∂
w
k
=
∫
∫
2
H
i
[
∂
ε
i
∂
w
k
G
i
+
e
i
∂
G
i
∂
w
k
-
2
∂
f
i
∂
w
k
F
i
-
2
f
i
∂
F
i
∂
w
k
+
∂
g
i
∂
w
k
E
i
+
g
i
∂
E
i
∂
w
k
2
(
E
i
G
i
-
(
F
i
)
2
)
-
(
ε
i
G
i
-
2
f
i
F
i
+
g
i
E
i
)
·
(
∂
E
i
∂
w
k
G
i
+
E
i
∂
G
i
∂
w
k
-
2
∂
F
i
∂
w
k
F
i
)
2
(
E
i
G
i
-
(
F
i
)
2
)
2
]
ⅆ
u
ⅆ
v
+
∫
∫
2
H
i
-
1
[
∂
ε
i
-
1
∂
w
k
G
i
-
1
+
e
i
-
1
∂
G
i
-
1
∂
w
k
-
2
∂
f
i
-
1
∂
w
k
F
i
-
1
-
2
f
i
-
1
∂
F
i
-
1
∂
w
k
+
∂
g
i
-
1
∂
w
k
E
i
-
1
+
g
i
-
1
∂
E
i
-
1
∂
w
k
2
(
E
i
-
1
G
i
-
1
-
(
F
i
-
1
)
2
)
-
(
e
i
-
1
G
i
-
1
-
2
f
i
-
1
F
i
-
1
+
g
i
-
1
E
i
-
1
)
·
(
∂
E
i
-
1
∂
w
k
G
i
-
1
+
E
i
-
1
∂
G
i
-
1
∂
w
k
-
2
∂
F
i
-
1
∂
w
k
F
i
-
1
)
2
(
E
i
-
1
G
i
-
1
-
(
F
i
-
1
)
2
)
2
]
ⅆ
u
ⅆ
v
.
7. The method of claim 6 , where J c =∫∫H 2 dudv, w k ε[θ i , ]∀i, E=S u ·S u , F=S u ·S v , G=S v ·S v , e=M·S uu , f=m·S uv , and g=M·S vv , where M is the surface normal.
8. The method of claim 3 , further comprising:
combining the first and second differential equations to generate the updated skin, wherein the first and second differential equations are combined by the following equation
∂
J
∂
w
k
=
(
1
-
k
)
∂
J
a
∂
w
k
+
∂
J
c
∂
w
k
,
where
∂
J
a
∂
w
k
is the first differential equation and
∂
J
c
∂
w
k
is the second differential equation.
9. The method of claim 8 , wherein the first differential equation is represented by
∂
J
a
∂
w
k
=
1
2
∫
∫
∂
E
i
∂
w
k
G
i
+
E
i
∂
G
i
∂
w
k
-
2
F
i
∂
F
i
∂
w
k
[
E
i
G
i
-
(
F
i
)
2
]
1
2
ⅆ
u
ⅆ
v
+
1
2
∫
∫
∂
E
i
-
1
∂
w
k
G
i
-
1
+
E
i
-
1
∂
G
i
-
1
∂
w
k
-
2
F
i
-
1
∂
F
i
-
1
∂
w
k
[
E
i
-
1
G
i
-
1
-
(
F
i
-
1
)
2
]
1
2
ⅆ
u
ⅆ
v
.
10. The method of claim 9 , where J a =∫∫√{square root over (EG−F 2 )}dudv, w k ε[θ i ,φ i ,]∀i, E=S u ·S u , F=S u ·S v , and G=S v ·S v .
11. The method of claim 8 , wherein the second differential equation is represented by
∂
J
c
∂
w
k
=
∫
∫
2
H
i
[
∂
ε
i
∂
w
k
G
i
+
e
i
∂
G
i
∂
w
k
-
2
∂
f
i
∂
w
k
F
i
-
2
f
i
∂
F
i
∂
w
k
+
∂
g
i
∂
w
k
E
i
+
g
i
∂
E
i
∂
w
k
2
(
E
i
G
i
-
(
F
i
)
2
)
-
(
e
i
G
i
-
2
f
i
F
i
+
g
i
E
i
)
·
(
∂
E
i
∂
w
k
G
i
+
E
i
∂
G
i
∂
w
k
-
2
∂
F
i
∂
w
k
F
i
)
2
(
E
i
G
i
-
(
F
i
)
2
)
2
]
ⅆ
u
ⅆ
v
+
∫
∫
2
H
i
-
1
[
∂
ε
i
-
1
∂
w
k
G
i
-
1
+
e
i
-
1
∂
G
i
-
1
∂
w
k
-
2
∂
f
i
-
1
∂
w
k
F
i
-
1
-
2
f
i
-
1
∂
F
i
-
1
∂
w
k
+
∂
g
i
-
1
∂
w
k
E
i
-
1
+
g
i
-
1
∂
E
i
-
1
∂
w
k
2
(
E
i
-
1
G
i
-
1
-
(
F
i
-
1
)
2
)
-
(
e
i
-
1
G
i
-
1
-
2
f
i
-
1
F
i
-
1
+
g
i
-
1
E
i
-
1
)
·
(
∂
E
i
-
1
∂
w
k
G
i
-
1
+
E
i
-
1
∂
G
i
-
1
∂
w
k
-
2
∂
F
i
-
1
∂
w
k
F
i
-
1
)
2
(
E
i
-
1
G
i
-
1
-
(
F
i
-
1
)
2
)
2
]
ⅆ
u
ⅆ
v
.
12. The method of claim 11 , where J c =∫∫H 2 dudv, w k ε[θ i ,φ i ,]∀i, E=S u ·S u , F=S u ·S v , G=S v ·S v , e=M ·S uu , f=M·S uv , and g=M·S vv , where M is the surface normal.
13. The method of claim 8 , wherein the first and second differential equations are used in a gradient descent procedure to find the desired skin.
14. The method of claim 13 , wherein the gradient descent procedure manipulates parameters w i =[θ i ,φ i ] T of each ball i, where w i (n+1)=w i (n)−Δt∇J w i (n) , ∀i to find the desired skin.
15. A method of modeling a tubular structure, comprising:
imaging a tubular structure;
placing a plurality of balls in the tubular structure; and
finding a skin that smoothly interpolates the balls,
wherein finding the skin comprises:
generating an initial skin, wherein the initial skin is a surface comprised of splines and wherein the splines touch each ball along a circle that is tangent to the ball;
solving a first differential equation to minimize the initial skin's surface area or solving a second differential equation to minimize a squared mean curvature of the initial skin's surface,
wherein the result of solving the first or second differential equations is an updated skin; and
repeating the steps of solving the first or second differential equations for the updated skin, and then, repeating the steps of solving the first or second differential equations for each subsequently updated skin until the skin that smoothly interpolates the balls is realized,
wherein the skin that smoothly interpolates the balls is a model of the tubular structure,
where a circle on a ball B i is in a plane with normal N i =[cos θ i sin φ i , sin φ i sin φ i , cos φ i ] T that passes through the ball center c i and intersects the ball B i , q(u) is a point on the circle and is defined by q(u)={circumflex over (x)} cos u+ŷ sin u, and a point p i (u) on a circle through which a spline passes is defined by p i (u)=c i +r i R i q(u), where r i is a radius of an i th ball, and R i is a rotation matrix specified by N i .
16. The method of claim 15 , wherein the tubular structure is an anatomical structure.
17. The method of claim 15 , wherein the tubular structure is imaged by a scanner.
18. The method of claim 15 , wherein the balls are ordered.
19. A system of computing a continuous interpolation of a discrete set of three-dimensional (3D) balls, comprising:
a memory device for storing a program; and
a processor in communication with the memory device, the processor operative with the program to perform a method, the method comprising:
generating an initial skin, wherein the initial skin is a surface comprised of splines and wherein the splines touch each ball along a circle that is tangent to the ball;
solving a first differential equation to minimize the initial skin's surface area or solving a second differential equation to minimize a squared mean curvature of the initial skin's surface,
wherein the result of solving the first or second differential equations is an updated skin; and
repeating the steps of solving the first or second differential equations for the updated skin, and then, repeating the steps of solving the first or second differential equations for each subsequently updated skin until a desired skin is realized,
where a circle on a ball B i is in a plane with normal N i =[cos θ i sin φ i , sin φ i sin φ i , cos φ i ] T that passes through the ball center c i and intersects the ball B i , q(u) is a point on the circle and is defined by q(u)={circumflex over (x)} cos u+ŷ sin u, and a point p i (u) on a circle through which spline passes is defined by p i (u)=c i +r i R i q(u), where r i is a radius of an i th ball, and R i is a rotation matrix specified by N i .