Instrumentation & Measurement Magazine 24-4 - 40

Table 2 - Performance measurements of
different methods for Set2 image
Set 2
EN
SD
MIA
MIB
FF
FS
SSIM
Method1 Method2 Method 3 Proposed
4.4911
3.9298
61.4794
4.729
3.786
8.455
0.0593
0.6930
61.8743
4.5761
3.6951
8.2712
0.0533
0.6602
4.5187
62.2631
4.6436
3.7288
8.3724
0.0546
0.6899
4.9446
71.1855
4.8233
4.3100
9.1333
0.0281
0.7518
and that suggests better fusion. Higher values of the mutual information
of the source images with respect to the fused image,
and therefore the fusion factor of the proposed method, emphasize
that more data is present within the fused image of the
proposed strategy compared to other methods. Low fusion factor
of the merged image in the proposed strategy constitutes an
improved fusion algorithm compared with the other methods.
Higher values of structural similarity of the proposed method
shows more similarity to the source image, being superior to
other methods. All of these performance metrics of the proposed
technique depict the advantage of the proposed way, due
to its shift invariant property.
Table 3 - Performance measurements of
different methods for Set3 image
Set 3
EN
SD
MIA
MIB
FF
FS
SSIM
Method1 Method2 Method 3 Proposed
3.5101
4.635
73.3631
4.6979
4.3456
9.04
0.0195
0.7333
73.4889
4.6338
4.5178
9.1516
0.0063
0.7535
4.6439
74.113
4.7343
4.6916
9.4259
0.0023
0.7619
5.1100
75.5026
5.1853
4.888
10.0741
0.0147
0.7767
Table 4 - Performance measurements of
different methods for Set4 image
Set 4
EN
SD
MIA
MIB
FF
FS
SSIM
Method1 Method2 Method 3 Proposed
4.5315
6.3735
64.961
4.9116
4.8407
9.7523
0.0036
0.5335
65.0393
4.5201
4.6545
9.1746
0.0073
0.6042
6.4047
65.533
4.7145
4.747
9.4615
0.0017
0.596
6.7576
71.2123
4.9515
4.9199
9.8700
0.0016
0.6697
Conclusion
In this research, a hybrid clinical 2D image fusion dependent
on 2-Dimensional Double Density Wavelet Transform
with Principle Component Analysis is implemented. The 2DDDWT
decomposition of clinical objects such as CT and MRI
brain 2D images are estimated in terms of high and low wavelet
coefficient due to its shift invariant property, depending on
the congruency rule of phase and gabor coefficients. The 2DDDWT
frequency coefficients are fused by the application of
Principle Component Analysis (PCA) for the approximation
parameters and henceforth rule of maximum selection is being
applied for the coefficients to improve quality of the image
to be fused. Since PCA analysis is implemented by extracting
the principal components based on eigen values and eigen vectors,
no information is being lost, resulting in a better fused
image that represents spectral and spatial information and
performs better and outrates the existing methodologies. The
fusion metrics of the images prove that the proposed system
performance is better than existing systems. The fused images
obtained by the proposed method were more informative and
could be utilized for clinical analysis.
References
[1] G. Bhatnagar, Q. M. J. Wu, and Z. Liu, " Human visual system
inspired multi-modal medical image fusion framework, " Expert
Sys. Applications, vol. 40, no. 5, pp.1708-1720, 2013.
[2] V. A. Rani and S. Lalithakumari, " Recent medical image fusion
techniques: a review, " Indian J. Public Health Res. Dev., vol. 10, Jul.
2019.
Table 5 - Performance measurements of
different methods for Set5 image
Set 5
EN
SD
MIA
MIB
FF
FS
SSIM
40
Method1 Method2 Method 3 Proposed
6.7351
5.8200
63.236
5.3835
4.9991
10.3826
0.0185
0.661
63.3773
5.0792
5.5382
10.6173
0.0216
0.4637
6.7369
63.5402
5.2282
5.8484
11.0766
0.028
0.5091
6.8203
68.1776
5.6312
5.6201
11.2513
0.004
0.6690
[3] S. Goel, S. Budhiraja, A. Dhindsa and N. Mehta, " CT and MRI
image fusion using wiener filter in dual tree framework, " in Proc.
2nd Int. Conf. Telecommunication and Networks (TEL-NET), pp. 1-7,
2017.
[4] T. Akbarpour, M. Shamsi and S. Daneshvar, " Medical image
fusion using discrete wavelet transform and lifting scheme, " in
Proc. 22nd Iranian Conf. Biomedical Eng. (ICBME), pp. 293-298,
2015.
[5] L. Tao and Z. Qian, " An improved medical image fusion
algorithm based on wavelet transform, " in Proc. 7th
Natural Computation, pp. 76-78, 2011.
Int. Conf.
[6] Y. Yang, Y. Que, S. Huang and P. Lin, " Multimodal sensor medical
image fusion based on type-2 fuzzy logic in NSCT domain, " IEEE
Sensors J., vol. 16, no. 10, pp. 3735-3745, May 2016.
IEEE Instrumentation & Measurement Magazine
June 2021
Instrumentation & Measurement Magazine 24-4

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-4
