You are on page 1of 6

Equations

• Slice Selection : • Better Approximation : ideal O2D (x, y ) =

∫O

3D

(x, y,z)δ (z − zs )dz

   z − zs  dz O2D (x, y ) = O3D (x, y,z)TopHat  1 ∆z  Ω  2 

• Radon Transform :
~

P (θ,z) =

∫ ∫ O(x, y)δ(x cos(θ ) + y sin(θ ) + z)
dx

• Fourier Transform : g(k ) = • Central Slice Theorem :

∫ G(x)e

−ikx

k z = k x cos(θ ) + ky sin(θ ) z = x cos(θ ) + y sin(θ ) Or (x, y ) = 1
π

• Back Projection :

π

∫ P (θ,z)dθ
0 ~

• Map :

O r (k x ,k y )= P (θ,k z ) where k z = kx cos(θ ) + k y sin(θ )

More Organized Proof of The Central Slice Theorem

The PSF associated with the simple Bach projection is: PSF ∴ Or (x, y ) = O(x, y ) ⊗
BF

=

1 r

1 x2 + y2

where Or (x, y ) = B{P (α,z)} and B = ∴ 1

π

∫ P (α x cos(α ) + y sin(α ))dα
1 0

R

Or (x, y ) 123 4 4

=

c

O(x, y ) 123

1 x2 + y2 1 24 4 3 1 k

c

Or (k x ,k y ) so

=

O(k x ,k y )

c

O(k x ,k y ) = k Or (k x ,k y )

More Organized Proof of The Central Slice Theorem

1. 2. P (α , z) =

∫∫ O(x, y )δ(x cos(α ) + y sin(α ) − z)dxdy

Equate the z− axis with a tilted reference frame x ' || z, y ' ⊥ z ∴ and x ' = x cos(α ) + y sin(α ) x = x ' cos(α ) − y ' sin(α ) y = x ' sin(α ) + y ' cos(α )

3.

Substitute # 2 into #1 and change integral to dx ' dy ' (still over all space) P (α , z) =

∫∫ O(x cos(α ) − y sin(α ), x sin(α ) + y cos(α ))δ(x − z)dx dy
' ' ' ' ' ' ' ' ' ' '

'

4.

Integrate along x ' and note that z is only a point along the x ' axis. P α, x ' =

( ) ∫∫ O(x cos(α ) − y sin(α ), x sin(α ) + y cos(α ))dy
) ∫∫ (
O x ' cos(α ) − y ' sin(α ), x ' sin(α ) + y ' cos(α ) e

5.

Fourier Transform along x ' p α , kx ' =
~

(

)

−ix ' k

x'

dx ' dy '

More Organized Proof of The Central Slice Theorem

6. Transform back to the (x, y ) coordinate system p α ,kx ' = 7.
~

(

) ∫∫ O(x, y )e

− i(x cos(α )+ y sin(α ))k

x'

dxdy

Define the tilted k− space coordinate system. k x = k x ' cos(α ) − ky ' sin(α ) k y = k x ' sin(α ) − k y ' cos(α )

8.

Rewrite # 6 as p α ,kx ' = p(α ,kz ) =
~
~

(

) ∫∫ O(x, y )e

    − i k ' cos(α )−k ' sin(α ) x −i k ' sin(α )+ k ' cos(α )y y y  x   x 

e

dxdy

k ' =0
y

∫∫ O(x, y
)e− ik xe−ik y dxdy
x y y

k ' =0
y

= F2D {O(x, y )} k ' =0

The Central Slice Theorem

Consider a 2-dimensional example of an emission imaging system. O(x,y) is the object function, describing the source distribution. The projection data, is the line integral along the projection direction.
P 0o, y =

( ) ∫ O(x, y )dx

The Central Slice Theorem can be seen as a consequence of the separability of a 2-D Fourier Transform.
o(kx ,k y )=
~

∫ O(x, y)e

−ikx x −iky y

e

dxdy

The 1-D Fourier Transform of the projection is,
p(ky ) =
~

∫ P (0 , y)e = ∫ O(x, y )e = ∫ O(x, y )e
o

−iky y

dy

−iky y

dxdy e dxdy

−iky y − i 0x

= o(0, ky )

~

The Central Slice Theorem
The one-dimensional Fourier transformation of a projection obtained at an angle J, is the same as the radical slice taken through the two-dimensional Fourier domain of the object at the same angle.
O(x, y )

Radon transform

P θ, y '

( )

2D FT

1D FT

o(kx ,k y )

~

θ

p θ,k y '

~

( )