GIMI: A New Evaluation Index for 3D Multimodal Medical Image Fusion

Multimodal medical image fusion plays important roles in clinical applications. Existing indexes used to evaluate 2D medical image fusion algorithms are not suitable for 3D fusions. In this paper, a new evaluation index, which is named as GIMI, is proposed to evaluate and compare the quality of 3D medical image fusion algorithms. GIMI index is based on image volumes not slices. It captures spatial information through the combination of image intensity and gradient, where gradients are computed in 3D space to reconnect all separated image slices together. It treats image slices as a whole of volume to improve the consistency of evaluation. Quantitative and qualitative test results show that GIMI index is effective in evaluating 3D medical image fusions. Its evaluation is consistent with the visual perception of fused images.

[1]  Ashish Khare,et al.  3D medical image fusion using dual tree complex wavelet transform , 2015, 2015 International Conference on Control, Automation and Information Sciences (ICCAIS).

[2]  Jens H. Krüger,et al.  State of the Art in Transfer Functions for Direct Volume Rendering , 2016, Comput. Graph. Forum.

[3]  Pooran Singh Negi,et al.  3-D Discrete Shearlet Transform and Video Processing , 2012, IEEE Transactions on Image Processing.

[4]  C. Himabindu,et al.  DWT based medical image fusion with maximum local extrema , 2016, 2016 International Conference on Computer Communication and Informatics (ICCCI).

[5]  M. Hossny,et al.  Comments on 'Information measure for performance of image fusion' , 2008 .

[6]  V. Bhavana,et al.  A survey on multi - modality medical image fusion , 2016, 2016 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET).

[7]  Lei Zhang,et al.  Multi-Scale Fusion Algorithm Based on Structure Similarity Index Constraint for Infrared Polarization and Intensity Images , 2017, IEEE Access.

[8]  Ke Lu,et al.  An overview of multi-modal medical image fusion , 2016, Neurocomputing.

[9]  Bin Li,et al.  Multimodal medical volumetric data fusion using 3-D discrete shearlet transform and global-to-local rule. , 2014, IEEE transactions on bio-medical engineering.

[10]  Mohammad Haghighat,et al.  Fast-FMI: Non-reference image fusion metric , 2014, 2014 IEEE 8th International Conference on Application of Information and Communication Technologies (AICT).

[11]  Sumit Budhiraja,et al.  Multimodal medical image fusion using modified weight maps and guided filter , 2017, 2017 2nd International Conference on Telecommunication and Networks (TEL-NET).

[12]  Henk J. A. M. Heijmans,et al.  A new quality metric for image fusion , 2003, Proceedings 2003 International Conference on Image Processing (Cat. No.03CH37429).

[13]  Shutao Li,et al.  Pixel-level image fusion: A survey of the state of the art , 2017, Inf. Fusion.

[14]  David R. Bull,et al.  Contrast Sensitivity of the Wavelet, Dual Tree Complex Wavelet, Curvelet, and Steerable Pyramid Transforms , 2016, IEEE Transactions on Image Processing.