TOMOGRAPHIC IMAGE RECONSTRUCTION OF FAN-BEAM PROJECTIONS WITH EQUIDISTANT DETECTORS USING PARTIALLY CONNECTED NEURAL NETWORKS



Learning and Nonlinear Models - Revista da Sociedade Brasileira de Redes Neurais, Vol. 1, No. 2, pp. 122-130, 2003
© Sociedade Brasileira de Redes Neurais

Figure 7: Zoom of part of the sinusoid, now showing the points located at gaps (in gray) used in the interpolation method.


It is clear from the equations (16) and (17) that the weights are complementary themselves,

ui=1 -vi

(16)


and both belong to the range [0,1].

The resulting single layer partially connected neural network to compute backprojection with fan-beam geometry and
interpolation is:

1K

f (r, φ) = yj =-∑[ ujkq (L5 J ,θk )+vjcq (Γ51 ,θ )]                                                                         (17)

K k=1

where the synaptic weights are given by:

vji = U-2(Γ 5 J- 5 )                                                                                                                  (18)

and

vii = U-2( 5 -L 5 J)                                                                                                                  (19)

It is worth mentioning that the backprojection algorithm using partially connected neural networks needs to run in two
distinct phases: the
building phase where the relevant weights are calculated and the feeding phase where the filtered projection
is presented at the input and the values propagate to the output according to the weights giving the reconstructed image. The
building phase involves the weights calculated by equations (14) and (15) for parallel beam, or equations (18) and (19) for fan-
beam with interpolation geometry. The network structure is very similar, the main difference resides on the weights values.
Network for parallel beam reconstruction with no interpolation have only unity weights while network with interpolation or
fan-beam geometry have non-integer weight values.

2.2 Filtering Network

The high-pass filtering process required for the inverse Radon transform amplifies high spatial frequencies. To avoid excessive
noise amplification at these frequencies a band-limited filter, called Ram-Lak filter is often used [10,11,14]. The filter impulse
response is:

1

∕4,

h ( n ) = -0,

-1

. nπ


2,


n=0

n even

n odd.


(20)


Only the first N values of n are considered, and the resulting FIR (finite impulse response) filter can be calculated by
convolving the filter impulse response with the input data, in this case the projection image
p(5,θ). This convolution operation
is implemented as a single layer neural network with linear activation function according to previous works [5,6].

q(n,θ) = N h(n - k) p(5k,θ),      n=1,..,N

k=1

127



More intriguing information

1. The Shepherd Sinfonia
2. Errors in recorded security prices and the turn-of-the year effect
3. Consumer Networks and Firm Reputation: A First Experimental Investigation
4. AGRICULTURAL TRADE IN THE URUGUAY ROUND: INTO FINAL BATTLE
5. The name is absent
6. Shifting Identities and Blurring Boundaries: The Emergence of Third Space Professionals in UK Higher Education
7. Imperfect competition and congestion in the City
8. Regional science policy and the growth of knowledge megacentres in bioscience clusters
9. The name is absent
10. The name is absent
11. Regional Intergration and Migration: An Economic Geography Model with Hetergenous Labour Force
12. The name is absent
13. Correlation Analysis of Financial Contagion: What One Should Know Before Running a Test
14. Macroeconomic Interdependence in a Two-Country DSGE Model under Diverging Interest-Rate Rules
15. The name is absent
16. Towards Teaching a Robot to Count Objects
17. The Role of Immigration in Sustaining the Social Security System: A Political Economy Approach
18. Infrastructure Investment in Network Industries: The Role of Incentive Regulation and Regulatory Independence
19. Skills, Partnerships and Tenancy in Sri Lankan Rice Farms
20. Handling the measurement error problem by means of panel data: Moment methods applied on firm data